Académique Documents
Professionnel Documents
Culture Documents
Michael E. Auer
Danilo G. Zutin Editors
Online Engineering
& Internet of
Things
Proceedings of the 14th International
Conference on Remote Engineering and
Virtual Instrumentation REV 2017, held
15–17 March 2017, Columbia University,
New York, USA
Lecture Notes in Networks and Systems
Volume 22
Series editor
Janusz Kacprzyk, Polish Academy of Sciences, Warsaw, Poland
e-mail: kacprzyk@ibspan.waw.pl
zamfira@unitbv.ro
The series “Lecture Notes in Networks and Systems” publishes the latest develop-
ments in Networks and Systems—quickly, informally and with high quality. Original
research reported in proceedings and post-proceedings represents the core of LNNS.
Volumes published in LNNS embrace all aspects and subfields of, as well as new
challenges in, Networks and Systems.
The series contains proceedings and edited volumes in systems and networks,
spanning the areas of Cyber-Physical Systems, Autonomous Systems, Sensor
Networks, Control Systems, Energy Systems, Automotive Systems, Biological
Systems, Vehicular Networking and Connected Vehicles, Aerospace Systems,
Automation, Manufacturing, Smart Grids, Nonlinear Systems, Power Systems,
Robotics, Social Systems, Economic Systems and other. Of particular value to both
the contributors and the readership are the short publication timeframe and the
world-wide distribution and exposure which enable both a wide and rapid
dissemination of research output.
The series covers the theory, applications, and perspectives on the state of the art
and future developments relevant to systems and networks, decision making,
control, complex processes and related areas, as embedded in the fields of
interdisciplinary and applied sciences, engineering, computer science, physics,
economics, social, and life sciences, as well as the paradigms and methodologies
behind them.
Advisory Board
Fernando Gomide, Department of Computer Engineering and Automation—DCA, School
of Electrical and Computer Engineering—FEEC, University of Campinas—UNICAMP,
São Paulo, Brazil
e-mail: gomide@dca.fee.unicamp.br
Okyay Kaynak, Department of Electrical and Electronic Engineering, Bogazici University,
Istanbul, Turkey
e-mail: okyay.kaynak@boun.edu.tr
Derong Liu, Department of Electrical and Computer Engineering, University of Illinois
at Chicago, Chicago, USA and Institute of Automation, Chinese Academy of Sciences,
Beijing, China
e-mail: derong@uic.edu
Witold Pedrycz, Department of Electrical and Computer Engineering, University of Alberta,
Alberta, Canada and Systems Research Institute, Polish Academy of Sciences, Warsaw,
Poland
e-mail: wpedrycz@ualberta.ca
Marios M. Polycarpou, KIOS Research Center for Intelligent Systems and Networks,
Department of Electrical and Computer Engineering, University of Cyprus, Nicosia, Cyprus
e-mail: mpolycar@ucy.ac.cy
Imre J. Rudas, Óbuda University, Budapest Hungary
e-mail: rudas@uni-obuda.hu
Jun Wang, Department of Computer Science, City University of Hong Kong
Kowloon, Hong Kong
e-mail: jwang.cs@cityu.edu.hk
zamfira@unitbv.ro
Michael E. Auer Danilo G. Zutin
•
Editors
123
zamfira@unitbv.ro
Editors
Michael E. Auer Danilo G. Zutin
Carinthia University of Applied Sciences Carinthia University of Applied Sciences
Villach Villach
Austria Austria
zamfira@unitbv.ro
Preface
zamfira@unitbv.ro
vi Preface
• Tarek M. Sobh
Senior Vice President for Graduate Studies and Research and Dean of the
School of Engineering, University of Bridgeport, USA
It was in 2004 when we started this conference series in Villach, Austria,
together with some visionary colleagues and friends from around the world. When
we started our REV endeavor, the Internet was just 10 years old! Since then, the
situation regarding Online Engineering and Virtual Instrumentation has radically
changed. Both are today typical working areas of most of the engineers, and are
inseparably connected with
• Internet of Things
• Cyber-physical Systems
• Collaborative Networks and Grids
• Cyber-cloud Technologies
• Service Architectures
to name only a few.
With our conference in 2004 (thirteen years ago), we tried to focus on the
upcoming use of the Internet for engineering tasks and the problems around it –
with big success as we can see today.
The following main themes have been discussed in detail:
• Online Engineering
• Cyber-physical Systems
• Internet of Things
• Industry 4.0
• Cyber-security
• M2M Concepts
• Virtual and Remote Laboratories
• Remote Process Visualization and Virtual Instrumentation
• Remote Control and Measurement Technologies
• Networking, Grid and Cloud Technologies
• Mixed-reality Environments
• Telerobotics and Telepresence, Coboter
• Collaborative Work in Virtual Environments
• Smart City, Smart Energy, Smart Buildings, Smart Homes
• Innovative Organizational and Educational Concepts
• Standards and Standardization Proposals
• Applications and Experiences
As submission types have been accepted:
• Full Paper, Short Paper
• Work in Progress, Poster
• Special Sessions
• Round Table Discussions, Workshops, Tutorials
zamfira@unitbv.ro
Preface vii
All contributions were subject to a double-blind review. The review process was
very competitive. We had to review nearly 300 submissions. A team of about 129
reviewers did this terrific job. My special thanks go to all of them.
Due to the time and conference schedule restrictions, we could finally accept
only the best 116 submissions for presentation. The conference had again more than
140 participants from 31 countries from all continents.
REV2018 will be held in Düsseldorf, Germany, and REV2019 in Bangalore,
India.
Michael E. Auer
REV General Chair
zamfira@unitbv.ro
Online Engineering & Internet
of Things – Proceedings of the 14th International
Conference on Remote Engineering
and Virtual Instrumentation (REV 2017)
Committees
General Chair
Michael E. Auer
Conference Co-chairs
ix
zamfira@unitbv.ro
x Online Engineering & Internet of Things
ASEE Liaison
IEEE Liaison
zamfira@unitbv.ro
Online Engineering & Internet of Things xi
zamfira@unitbv.ro
Contents
Internet of Things
Cloud-Based Industrial Control Services . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Reinhard Langmann and Michael Stiller
Wireless Development Boards to Connect the World . . . . . . . . . . . . . . . . 19
Pedro Plaza, Elio Sancristobal, German Carro, Manuel Castro,
and Elena Ruiz
CHS-GA: An Approach for Cluster Head Selection
Using Genetic Algorithm for WBANs . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Roopali Punj and Rakesh Kumar
Proposal IoT Architecture for Macro and Microscale Applied
in Assistive Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Carlos Solon S. Guimarães, Jr., Renato Ventura B. Henriques,
Carlos Eduardo Pereira, and Wagner da Silva Silveira
Using Industrial Internet of Things to Support Energy Efficiency
and Management: Case of PID Controller . . . . . . . . . . . . . . . . . . . . . . . . 44
Tom Wanyama
MODULARITY Applied to SMART HOME . . . . . . . . . . . . . . . . . . . . . . 56
Doru Ursuţiu, Andrei Neagu, Cornel Samoilă, and Vlad Jinga
Development of M.Eng. Programs with a Focus on Industry 4.0
and Smart Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Michael D. Justason, Dan Centea, and Lotfi Belkhir
Remote Acoustic Monitoring System for Noise Sensing . . . . . . . . . . . . . . 77
Unai Hernandez-Jayo, Rosa Ma Alsina-Pagès, Ignacio Angulo,
and Francesc Alías
xiii
zamfira@unitbv.ro
xiv Contents
zamfira@unitbv.ro
Contents xv
zamfira@unitbv.ro
xvi Contents
zamfira@unitbv.ro
Contents xvii
Games Engineering
Dinner Talk: A Language Learning Game Designed
for the Interactive Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 555
Jacqueline Schuldt, Stefan Sachse, and Lilianne Buckens
The Experimento Game: Enhancing a Players’ Learning Experience
by Embedding Moral Dilemmas in Serious Gaming Modules . . . . . . . . . 561
Jacqueline Schuldt, Stefan Sachse, Verena Hetsch, and Kevin John Moss
zamfira@unitbv.ro
xviii Contents
zamfira@unitbv.ro
Contents xix
Simulation
Augmented Reality-Based Interactive Simulation Application
in Double-Slit Experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 701
Tao Wang, Han Zhang, Xiaoru Xue, and Su Cai
Developing Metacognitive Skills for Training
on Information Security . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708
Jesus Cano, Roberto Hernandez, Rafael Pastor, Salvador Ros,
Llanos Tobarra, and Antonio Robles-Gomez
Optimization of the Power Flow in a Smart Home . . . . . . . . . . . . . . . . . 721
Linfeng Zhang and Xingguo Xiong
A Virtualized Computer Network for Salahaddin University
New Campus of HTTP Services Using OPNET Simulator . . . . . . . . . . . 731
Tarik A. Rashid and Ammar O. Barznji
Online Engineering
GIFT - An Integrated Development and Training System
for Finite State Machine Based Approaches . . . . . . . . . . . . . . . . . . . . . . . 743
Karsten Henke, Tobias Fäth, René Hutschenreuter,
and Heinz-Dietrich Wuttke
A Web-Based Tool for Biomedical Signal Management . . . . . . . . . . . . . . 758
S.D. Cano-Ortiz, R. Langmann, Y. Martinez-Cañete, L. Lombardia-Legra,
F. Herrero-Betancourt, and H. Jacques
Optimization of Practical Work for Programming Courses
in the Context of Distance Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . 764
Amadou Dahirou Gueye, Pape Mamadou Djidiack Faye,
and Claude Lishou
Enabling the Automatic Generation of User Interfaces
for Remote Laboratories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 778
Wissam Halimi, Christophe Salzmann, Hagop Jamkojian,
and Denis Gillet
A Practical Approach to Teaching Industry 4.0 Technologies . . . . . . . . . 794
Tom Wanyama, Ishwar Singh, and Dan Centea
Design of WEB Laboratory for Programming and Use
of an FPGA Device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 809
Nikola Jović and Milan Matijević
Remote Triggered Software Defined Radio Using GNU Radio . . . . . . . . 822
Jasveer Singh T. Jethra, Pavneet Singh, and Kunal Bidkar
zamfira@unitbv.ro
xx Contents
zamfira@unitbv.ro
Contents xxi
zamfira@unitbv.ro
Internet of Things
zamfira@unitbv.ro
Cloud-Based Industrial Control Services
The Next Generation PLC
1
Hochschule Duesseldorf University of Applied Sciences, Duesseldorf, Germany
langmann@ccad.eu
2
Fraunhofer Institute for Embedded Systems and Communication Technologies ESK,
Munich, Germany
michael.stiller@esk.fraunhofer.de
Abstract. The paper presents the concept and implementation for Cloud-based
Industrial Control Services (CICS) as a next generation PLC. As a distributed
service-oriented control system in the cloud, a CICS controller can replace the
traditional PLC for applications with uncritical timing in terms of Industry 4.0.
The CICS services are programmed to industry standards, pursuant to standard
IEC 61131-3, and executed in a CICS runtime in the cloud. This paper gives an
overview about the concept and implementation, discusses the results of appli‐
cation examples as well as the evaluation of the operability of a CICS controller.
1 Introduction
Industrial controls and, in particular, PLC controllers currently form an important techno‐
logical basis for the automation of industrial processes. Even in the age of Industry 4.0
(I40) and Industrial Internet, it can be assumed that these controllers will continue to be
required to a considerable extent for the production of tomorrow. However, the controllers
must fulfil a range of additional requirements resulting from the new production conditions.
When applying Industry 4.0 principles [1], high-quality networked production systems
result based on Cyber Physical Systems (CPS), also referred to as Cyber Physical Produc‐
tion Systems (CPPS). A series of I40 requirements are placed on the future controllers used
in these systems. Current PLC controllers cannot yet fulfil the majority of these require‐
ments or can only do so on a rudimentary basis or with extremely high expense.
Basic requirements of future and I40-able PLC controllers involve efficient networking
in an, at least partially, global network and the ability to provide control functions as
control services in this network. Here the IP network (IP – Internet Protocol) functions as
a global network in the version as Intranet or Internet with all associated standardised
Information and Communication (IC) technologies. Only in this way can the required inte‐
gration become part of a future I40 production landscape.
zamfira@unitbv.ro
4 R. Langmann and M. Stiller
The paper describes the concept and a prototype implementation for the new type of a
PLC controller in which the controller functions (programs) will be implemented as control
services in a cloud. The programming of this new PLC occurs as is usual in industry
pursuant to the standard IEC 61131-3. The R&D results described in the paper have been
provided since 2014 as part of the R&D project “Potential, structure and interfaces of
cloud-based industrial control services (CICS)”.
2 State-of-the-Art
Resulting from the historical development of PLC controllers, these have been devel‐
oped as proprietary device systems that are operated locally under real-time conditions.
If a networking of these controllers is necessary from a user viewpoint, proprietary TCP/
IP protocols or those standardised in the automation sector (Modbus TCP, Profinet etc.)
are used for this. The standard technologies widespread from the Internet and Web have
so far hardly played any role for PLC controllers.
For a number of years, however, a transformation has been under way, with PLC
manufacturers increasingly integrating IC technologies from the web world in their
systems, such as web server and HTML pages for diagnosis and configuration, in order
to adapt the controllers incrementally to the new requirements.
Four different approaches to make PLC controllers I40 compatible can essentially
be revealed from state-of the-art technologies. Generally, all work assumes that the
creation of the control programs must take place in accordance with the standard
IEC61131-3, i.e. controllers from the cloud are currently only accepted in the industry
if the engineering also follows the industry standard.
Most of the newer PLC controllers already contain a web server as well as special HTML
pages built into the device, enabling a browser-based configuration and diagnosis of the
controller. Process data or program variables form the control program and can also be
read, and sometimes also written, with restrictions. Access via a web browser occurs
through the HTTP protocol, which is query-based and therefore relatively slow. Exam‐
ples of this can be found in [2]. The solutions are proprietary and adapted to the relevant
controller.
zamfira@unitbv.ro
Cloud-Based Industrial Control Services 5
Current R&D work deals with the virtualisation of complete PLC controllers and their
outsourcing into a cloud. A scalable control platform for cyber-physical systems in
industrial productions is researched and realised in [10]. Such a control platform is
intended to provide scalable computing power that is automatically made available
depending on the complexity of the algorithms. The strict requirements of production
technology, such as real-time capability, availability and security should be met.
In [11], a cloud-based controller is presented, which also uses a virtual control system
in an IaaS cloud. The work of [12] also uses virtualised PLC controls in the cloud and
connects these to OPC UA-based automation devices using web technologies. Problems
with the virtualisation of PLC result especially from the fact that already available
manufacturer-specific PLCs are virtualised. These controllers, however, are closed
systems, which were originally not developed considering the aspect of web technolo‐
gies. Adjustments, modifications or extensions of these controllers by third parties are
hardly possible. Functionality cannot be resolved as services. The flexibility of virtual‐
isation is very limited.
In summary, it can be estimated that there are different solutions and efforts to equip
PLC controllers with additional functions in order to be able to use the controllers in an
Industry 4.0-type IP network. To this end, the known work already uses web technologies
zamfira@unitbv.ro
6 R. Langmann and M. Stiller
in part, in a manufacturer specific and/or limited way, and increasingly also use the service
principle as well as cloud structures as a new paradigm for the realisation of control func‐
tions. However, there are several deficits exist, which accordingly require additional
research.
3 Concept
Looking at a PLC as a CPS component, the traditional IEC 61131 control program
(CP) can be divided into three parts:
• basic functional program part (CP basic - CPb),
• a program part which performs superior, administrative and/or user interface func‐
tions (CP supervisory - CPs),
• critical part of the program regarding real-time and security (CP critical - CPc).
In order to evaluate the I40 capabilities of a CICS controller, this 3-part structuring
of the control program is used, among other things. Figure 1 shows the structure of such
a PLC.
zamfira@unitbv.ro
Cloud-Based Industrial Control Services 7
If the control system, as shown in Fig. 1, is used as the basis and modified as a result
of the increasing displacement of the control programs into a cloud as services, it leads
to the evolution of a PLC as CPS component “industrial control”, as shown in Fig. 2.
Three types of CPS components (Fig. 2) are produced according to the aforemen‐
tioned disassembly of the PLC program into three parts:
(a) The controller only implements the program components CPb and CPc. The tradi‐
tional runtime environment of a PLC is still required.
(b) For safety reasons, only the CPc program parts are implemented in the CPS compo‐
nent. The classic and manufacturer-specific PLC runtime machine is no longer
required. The implementation of the CPc could also be carried out with specific
embedded program parts (e.g. in C).
(c) The CPS component no longer contains a control part, but only sensors and actua‐
tors. All control programs are distributed in the network.
Service ability considers the ability of a control to utilise control functionalities (control
program parts) as services in the sense of cloud computing. According to Table 1, the
program parts CPb, CPs and CPc can be distributed unequally. In class C11, for example,
the uncritical and overlapping functionalities (CPs) are not located on the local control
platform, but distributed on other systems in the network (corresponds to a traditional,
distributed control system). However, part of the CPs could also be used as a service from
a cloud.
zamfira@unitbv.ro
8 R. Langmann and M. Stiller
By separating CICS-P, CICS-RT and CICS-R in a CICS controller and the principle
arbitrary distribution of the individual components in an IP network utilising cloud
technologies, both a change in the control program code (control algorithm) as well as
a change in the device configuration with, for example, the replacement of modules and
Plug & Work in real time are possible. CICS-R and CICS-P can be exchanged on-the-
fly during a program cycle.
For the identification of a viable CICS basic model, it is also necessary to show
possible solution variants for CICS, starting with the basic principles on the web, and
then to reflect these in the available web technologies.
In principle, two types of network computers are available in the Internet (Web) as
a world-wide computer network:
• Server computers that can provide and run IT entities (objects, services, programs).
• Client computers that can only run IT entities.
As a working principle on this server/client computer network, the client-server
principle, i.e. a client must first submit a request to run an IT entity on the server. This
means that application technology IT entities in the server cannot act on their own (self-
acting). The client is usually a web browser on the client computer.
zamfira@unitbv.ro
Cloud-Based Industrial Control Services 9
Fig. 4. Models for executing a web technology based functional system (block with black
font = only saved, block with white font = is being executed)
In all three cases, the functional system can be distributed over several servers (cloud)
or even several clients. If the control technology based CICS structure, shown in Fig. 3,
is adapted to the web based functional systems, seen in Fig. 4, two server and client
based CICS base models are obtained:
1. Server Mode (SM): The CICS runtime is statically linked to a fixed CPS component
in a configuration process. After the CICS control has been started via the client, the
CICS controller automatically connects to the associated CPS component via the IP
network and executes the control program.
2. Server-based Mixed Mode (SMM): Before starting the CICS runtime, a CICS router
is loaded from the server to the client. After the CICS runtime is started, this router
dynamically connects the CPS component with the CICS runtime on the server. The
process data from the automation device are now routed to the server via the client.
3. Client Mode (CM): CICS runtime and CICS routers are executed as an instance on
the client (Web browser). The client is an inherent part of the CICS control system
and is necessarily required for executing the control program. The server is no longer
required at the runtime.
zamfira@unitbv.ro
10 R. Langmann and M. Stiller
4. Client-based Mixed Mode (CMM): The control program runs in the CICS runtime
on the client, but the communication to the CPS component runs over a dynamically
reconfigurable CICS router in the server.
Figures 5 and 6 illustrate the four basic models of a CICS control.
Fig. 5. Component structure and communication paths for server-based CICS solutions (1) –
Server Mode (SM); (2) – Server-based Mixed Mode (SMM)
Fig. 6. Component structure and communication paths for client-based CICS solutions (1) –
Client-base Mixed Mode (CMM) (2) – Client Mode (CM)
zamfira@unitbv.ro
Cloud-Based Industrial Control Services 11
or Web world and generally also are stored there in databases, it makes sense to use data
models for the modelling of the CICS service architecture. Figure 7 shows the CICS
structure, seen in Fig. 3, as an Entity Relationship Diagram (ERD).
Fig. 7. CICS services of a CICS controller according Fig. 3 presented as an ERD diagram
According Fig. 7 a CICS controller is realised with two services: Runtime service
and Router service. Both CICS services are built according to the principle of web-
oriented automation services [13].
1
IIoT – Industrial Internet of Things.
zamfira@unitbv.ro
12 R. Langmann and M. Stiller
The digital and analogue inputs and outputs of a device, connected via the channel
interface, are routed to absolute I/O program addresses and transferred to the CICS-RT
via a CICS block channel. The routing rules (interconnection matrix) are defined via a
CICS-R XML file. The I/O process data is transmitted via the bidirectional CICS block
channel as a string between the CICS router and the CICS runtime.
4 Implementation
Within the context of the CICS project, there were two prototype implementations for
a CICS controller:
• CICS controller in the Client Mode (CM – Fig. 6-2)
• CICS controller in the Server-based Mixed Mode (SMM – Fig. 5-2)
zamfira@unitbv.ro
Cloud-Based Industrial Control Services 13
The CICS controller is implemented as a JavaScript object, which is loaded onto the
client from a cloud (CICS Cloud). Communication between the CICS controller instance
and the automation devices takes place via a universal gateway as a web connector. The
gateway is implemented in an embedded system (pure.box of Wiesemann & Theis) as
a Device Gateway for Modbus TCP and for a proprietary TCP protocol. WebSocket is
used as a web protocol [14].
The control programs for the CM prototype were created with the industry-standard
programming system PC WORX (Phoenix Contact) in the language IL and exported as
PLCopen XML programs for execution in the CICS runtime. The execution is performed
via a JavaScript-based IL interpreter.
In the case of the SMM solution, the CICS runtime is executed in the server (cloud) as
an instance and the CICS router in the client (browser) as an instance (Fig. 10).
Here, as well, a direct process data communication takes place only between the
client and devices. Between CICS-R and CICS-RT, there is a special bidirectional CICS
zamfira@unitbv.ro
14 R. Langmann and M. Stiller
block channel for the transmission of I/O images. The process data are transmitted by
this channel as Strings over WebSocket. The CICS runtime is operated via an HMI proxy
on the client. In terms of technical implementation, the CICS-SMM controller is a
distributed elaborate solution. However, process data connection to the devices can also
be performed locally and the CICS-RT can use the full performance of the server
anyway. A dynamic re-configuration is easy and possible.
5 Application
The CICS components (CICS-RT and CICS-R) are constructed as services according
to the WOAS principle [15] and are instantiated as JavaScript objects using uniform and
consistent methods. They can therefore be used in an available IoT platform or as stand-
alone web pages. The freely available IIoT platform WOAS was used for the following
application examples (http://woas.ccad.eu). The applications therefore did not have to
be programmed, but instead were configured in the IIoT platform in a browser-based
EDIT mode with little effort.
An application example for a CICS-CM and for a CICS-SMM described hereafter.
The CICS-CM was successfully tested at a processing and testing station and presented
at the SPS/IPC/Drives automation fair in Nuremberg (Germany) in 2015 (Fig. 11).
Fig. 11. Application example of a CICS controller in client mode (CM); (a) processing and test
station; (b) I/O modules, IoT gateway and switch; (c) HMI in the web browser for operating/
visualisation of the CICS-CM
zamfira@unitbv.ro
Cloud-Based Industrial Control Services 15
The PLC program works like a cyclical state machine that waits for the presence of
a piece in position 1 of the table and then performs some actions in other positions; it
returns to position 1 and repeats the process. The dedicated PLC program has around
250 lines and uses about 60 variables. Some of the advantages of the CM solution are:
• The server is no longer required for the runtime. The server/cloud only serves the
purpose of storing the CICS services, including the control programs. Faults or fail‐
ures of the server have no effect on the CICS control system at runtime.
• The process data communication between the devices and the CICS controller can
be limited to the local network when the client is also located on this network.
• Depending on the performance class of the client, several CICS control instances can
run on a client and thus control different CPS components (devices).
• The quality and reliability of communication between the CICS controller and
connected devices can be monitored by the client.
• Generally, any client (PC, tablet, smartphone) can function as a CICS controller.
Two CICS controller instances control the two processing and test stations. Another
CICS instance is responsible for coordinating the robot with the two stations. Same as
in the test example for the CICS-CM, the connection of the respective CICS router
instances to the devices of the two stations takes place via a universal gateway as a web
connector. The Modbus TCP interface of the robot is connected to the Internet via
zamfira@unitbv.ro
16 R. Langmann and M. Stiller
6 Evaluation
A CICS control system uses IP networks for data transmission, regardless of the solution
variation. From the perspective of an automation technician, these networks are a priori
neither reliable nor deterministic and are not within the jurisdiction of the respective
technical automation solution. Extensive time measurements for different communica‐
tion structures were therefore performed for both CICS prototypes.
A practice-oriented method was chosen for the time measurements, which allows
direct statements about the reaction time of a CICS controller [17]. With respect to the
real-time capability, the following general statements can thus be made:
• With a CICS CM solution, response times of about 80…120 ms at a 95% probability
can be achieved by a standard Internet connection.
• With a CICS SMM solution, the reaction times likewise with a probability of 95%
are about 100 ms.
If the CICS controller is operated only in the Intranet, response times of under 40 ms
can certainly be achieved. Altogether, the statement can be made that technical processes
with process times of >150 ms (simple assembly process, temperature and mixing
processes, climate and energy processes, etc.) can already be performed successfully
from the cloud by means of a CICS.
6.2 Operability
The operability of a CICS control system is understood to mean the characteristics which
are important for a conventional PLC: reliability, data security and machine safety. In
this regard, new challenges arise that need to be studied and solved for the practical use
of CICS control systems.
For the sake of maintaining reliability, the two realised prototypes were investigated
in more detail by way of example. In this process, monitoring the network quality, inte‐
grating a ping-pong channel as well as a local operation for the process data transfer to
the device played a role.
Although reliability problems can be recognised with the methods described, they
cannot be eliminated. However, the CICS controller can at least be brought into a safe
state in problem situations.
7 Summary
Using the CICS concept, a new type of industrial control was developed and tested that
allows for the complete detachment of control function and associated equipment to
zamfira@unitbv.ro
Cloud-Based Industrial Control Services 17
References
1. Kagermann, H., Wahlster, W., Helbig, J.: Recommendations for implementing the strategic
initiative INDUSTRIE 4.0. In: Research Union: Business - Science, April 2013
2. Evans, Z.: Web Server Technology in Automation. Siemens Industry (2011)
3. Klindt, C.J., Baker, R.B.: Interface to a programmable logic controller. US 6853867 B1,
8 February 2005
4. Bosch-Rexroth: WebConnector: Connect simply automation and web environment (in
German). Product News, p. 50 (2015)
5. Beckhoff automation GmbH: From sensor to IT Enterprise - Big Data & analytics in the
cloud (2015). ftp://beckhoff.com/Software/embPC-Control/Solution/Demo-IoT/Flyer-
IoT-Sensor_to_Cloud.pdf
6. EU project: SOCRADES 2006–2009. http://www.socrades.net
7. Colombo, A.W., et al.: Industrial Cloud-Based Cyber-Physical Systems. Springer,
Switzerland (2014)
8. Microsoft: Introducing DPWS (2015). https://msdn.microsoft.com/en-us/library/dd170125.aspx
9. Mathes, M., et al.: SOAP4PLC: web services for programable logic controller. In: 17th
Euromicro International Conference on Parallel, Distributed and Network-Based Processing,
Proceedings, pp. 210–219 (2009)
10. ISW of the University Stuttgart: Industrial cloud-based control platform for the production
with Cyber-Physical Systems (piCASSO - in German). BMBF project, Stuttgart (2013).
http://www.projekt-picasso.de
zamfira@unitbv.ro
18 R. Langmann and M. Stiller
11. Grischan, E.: UACloud-Based Automation (in German). atp edn., 3/2015, pp. 28–32
12. Schmitt, J.: UACloud-Enabled Automation Systems using OPC UA. atp edn., 7–8/2014,
pp. 34–40
13. Competence Center Automation Düsseldorf (CCAD): WOAD – Interface description (in
German). – R&D document, Düsseldorf, 09 April 2014
14. Braß, M.: Development and test of an Industry 4.0 gateway for the Internet access to Modbus
TCP and proprietary TCP protocols. (in German). – Documentation of project work, CCAD,
25 January 2016
15. Langmann, R., Meyer, L.: Architecture of a web-oriented automation system. In: 18th IEEE
International Conference on Emerging Technologies & Factory Automation, ETFA 2013,
10–13 September 2013, Cagliari, Italine, Proceedings
16. Node-RED (2016). http://nodered.org
17. Langmann, R.: An interface for CPS-based automation devices (in German). In: Proceedings
of AALE 2014, pp. 133–142, DIV-Verlag, Munich (2014)
zamfira@unitbv.ro
Wireless Development Boards to Connect
the World
1 Introduction
The aim of this paper is to present some of the current development boards which can
be used to deploy IoT (Internet of Things) applications within an STEM (Science,
Technology, Engineering and Mathematics) educational environment.
Nowadays, there are a wide range of development boards which can be classified in
several ways. According to [1], the development platforms can be categorized in four
groups:
• Based on Microcontrollers,
• based on Microprocessors,
• based on FPGA (Field Programmable Gate Array), and
• Hybrid Development Platforms.
The IoT (Internet of Things) movement has impacted on the traditional Develop-
ment Boards. There are lots of IoT based applications which are being developed in
many different fields [2]. Some examples are [3] in Emergency Medical Services, [4] in
Cloud Computing and [5] in Remote Educational laboratories.
The Arduino WiFi Shield provides to the Arduino board a wirelessly internet
connection [6]. It cannot work in a stand-alone mode. Hence, this board requires a
microcontroller to interact with other elements.
The Yún Shield easily brings the Yún features to Arduino and Genuino boards. It is
a good choice for IoT projects using wireless connection to access the internet [6].
Genuino MKR1000 is a powerful board that combines the functionality of the Zero
and the Wi-Fi Shield. It is the ideal solution for makers wanting to design IoT projects
with minimal previous experience in networking [6].
This board is an Open Source Firmware and development kit that helps the IoT
product prototyping within a few Lua script lines [7].
The Onion Omega is a Hardware development platform with built-in WiFi and a
full Linux Operating System [8].
There are a vast variety of software development platforms which can be used as
core of IoT applications. Some of them are cost-effective platforms such as the men-
tioned in this paper. Furthermore, these platforms can be easily included with the aim
of elaborating robotic educational activities where the proactive learning is empowered
through experiments in the real world.
This paper is divided in four sections. Section 2 presents the analyzed IoT devel-
opment boards. Section 3 compares all of them. The last section summarizes the
achieved conclusions after the performed investigation.
zamfira@unitbv.ro
Wireless Development Boards to Connect the World 21
• There is an onboard micro-SD card slot, which can be used to store files for serving
over the network.
• The board mechanical data are: Length: 63.2 mm and width: 53.5 mm.
• The Arduino WiFi Shield board cost is 69.00 € [6].
In the same way than Arduino WiFi Shield, the Yún Shield extends the Arduino
board with the power of a Linux based system which enables advanced network
connections and applications. This board presents the following characteristics:
• Yún Web Panel and the ‘‘YunFirstConfig’’ sketch can be used to connect through
WiFi or wired network (Ethernet) in a simple way.
• The Shield preferences and sketch uploading can be performed directly from the
attached Arduino/Genuino board.
• The board mechanical data are: Length: 68.6 mm and width: 53.3 mm.
• The Genuino Yún Shield board cost is 39.90 € [6].
Genuino MKR1000 has been designed to offer a practical and cost effective
solution for projects which require Wi-Fi connectivity. This board presents the fol-
lowing characteristics:
• It is based on the Atmel ATSAMW25 SoC (System on Chip).
• This processor is part of the SmartConnect family of Atmel Wireless devices.
SmartConnect family is specifically designed for IoT projects.
• The ATSAMW25 includes also a single 1 1 stream PCB (Printed Circuit Board)
Antenna.
• The board includes a Li-Po charging circuit which allows the use of a Li-Po battery
as external power. Additionally, a 5 V external power supply is allowed. Internally,
the MKR1000 switches automatically from both supply sources.
• The board mechanical data are: Length: 65.0 mm and width: 25.0 mm.
• The Genuino MKR1000 board cost is 31.99 € [6].
zamfira@unitbv.ro
22 P. Plaza et al.
• High level programming languages such as Python, Javascript, PHP can be used.
• The Onion Omega is fully integrated with the Onion Cloud with the aim of creating
Internet of Things applications.
• It is Open Source. The processor is the Qualcomm Atheros AR9331 SoC.
• The board mechanical data are: Length: 42.7 mm and width: 26.4 mm.
• The Onion Omega board cost is 19.99 $ [10]. Using [11] for currency conversion
from United States dollars to Euros the board cost is 17.94 €.
3 Discussion
Along the previous sections seven IoT development platforms have been analyzed with
the aim of knowing about the Built-in Hardware, the Programmer Interface, the con-
nection possibilities and the Developer Community which is behind the corresponding
board.
There are two types of Arduino/Genuino IoT boards: Shield boards and the
full-integrated boards. The Shield boards require an additional microcontroller in order to
interact with other elements such as sensors or actuators which are widely used in robotic
education. On the other hand, NodeMCU and Onion Omega can be used in a stand-alone
mode. Table 1 summarizes the microcontroller and processor for each board.
With the aim of powering the boards, it is important to know what is voltage level
for each one in order to adapt levels from the board to other connected devices. Table 2
lists the IoT development boards and the voltage for the power supply and the input
and output port interfaces.
Other important characteristic for the development is the available memory. The
presented boards include different kind of memory resources, Table 3 compiles which
type of memory – volatile and non-volatile - is available and how much memory can be
used.
In common applications, IoT boards should interface with other devices using a
wired connection. These communications usually are performed using a serial inter-
face. Furthermore, digital and analog ports are used in order to read sensor values or
interact with some kind of actuators. Table 4 summarizes the serial and port interfaces
which can be performed for each IoT board.
zamfira@unitbv.ro
Wireless Development Boards to Connect the World 23
Table 2. IoT development board power supply and port interface voltage level.
IoT development board Power supply Input/Output voltage
Arduino WiFi Shield 5 V externally There is not Input/Output
port interface
Genuino Yún Shield 3.3 V There is not Input/Output
port interface
Genuino MKR1000 5 V or Li-Po single cell, 3.7 V, 3.3 V
700 mAh minimum
NodeMCU ESP8266 5 V from USB or 3.3 V from VIN 3.3 V
Onion Omega 5 V from USB or 3.3 V from VIN 3.3 V
zamfira@unitbv.ro
24 P. Plaza et al.
When there are Hardware and Software developments, the software required for
programming the board and the used language are very important elements. Table 5
states the IoT development board programming software and the used language for
programming them.
Other important specification for any application is the size restriction. None of the
analyzed boards are very large, the larger one has a length of 68.6 mm and a width of
53.3 mm - Genuino Yún Shield – and the smallest one is sized with a length of
38.0 mm and a width of 25.0 mm - NodeMCU ESP8266. Table 6 compares the IoT
development boards dimensions.
Moreover, the board cost is compared too. Due to the project funding, the cost is an
important aspect which should be have in mind. As it can be seen, none of them are
especially expensive, the most expensive board costs 69.00 € - Arduino WiFi Shield –
and the cheapest one costs 7.95 € - NodeMCU ESP8266. All of them are affordable for
majority of projects. Table 7 specifies the IoT development boards cost.
Furthermore, other important aspect is getting support during the development,
communities are very useful. Traditionally, manufacturers provided some telephone or
some e-mail for this purpose. Nowadays, most of manufacturers include a forum with
the aim of providing support to their customers. These forums are built by company’s
zamfira@unitbv.ro
Wireless Development Boards to Connect the World 25
experts and the customers and all of them form a community for the corresponding
product.
Arduino WiFi Shield, Arduino Yún Shield and Arduino MKR1000 are supported
by the Arduino community [12]. NodeMCU ESP8266 is supported by two commu-
nities: NodeMCU community [13] and ESP8266 community [14]. Onion Omega is
supported by the Onion community [15].
Finally, when a development is started it is important to get references about things
what other people has made and the way they are using the development boards.
Arduino projects are widely extended to scientific community, professionals and
hobbyists [16] presents a body area network, for acquiring data related to body position
and some simple movements based on a WiFi Shield stacked on an Arduino ChipKIT
MAX32 [17] describes a low-cost Wi-Fi sensor network based on ESP8266. Enable
monitoring of heart pulse sensor data on the cloud with ESP8266 is detailed in [18].
Using one of the described IoT development boards, mobile robots can be used as
remote laboratory in order to teach computer science in a similar way that is described
in [19, 20]. It can be used as a guide for using IoT as a TEL (Technology Enhanced
Learning) tool.
4 Conclusions
The result of this work shows the results of the analyzed IoT development boards that
can be introduced easily in classrooms within a STEM context. Performed educational
activities using the mentioned platforms are also considered. Hence, recommendations
are included with the aim of easing the inclusion of one of them in a classroom.
Any of this IoT development boards can be used in classrooms or remotely in order
to provide an easy way with the aim of including robotics within a STEM context.
Additionally, they allow making homemade applications framed in a DIY (Do It
Yourself) context.
Presented analysis is part of the state of the art of a doctoral thesis; a novel approach
to collaborative robotic educational tool is being developed. An Open Hardware plat-
form that can be used in classrooms with aim of developing educational programs
related to robotics in a collaborative environment which promotes innovation and
motivation for students during the learning process 1. The platform which is being
developed presents wirelessly connections such as Bluetooth a WiFi as enhancements 2.
The Wireless connection is provided by a WiFi development board which is integrated
as part of the collaborative robotic educational tool. The doctoral thesis is being carried
zamfira@unitbv.ro
26 P. Plaza et al.
out in the Engineering Industrial School of UNED (Spanish University for Distance
Education) and the Electrical and Computer Engineering Department (DIEEC).
Acknowledgment. The authors acknowledge the support provided by the Engineering Indus-
trial School of UNED, the Doctorate School of UNED, and the “Techno-Museum: Discovering
the ICTs for Humanity” (IEEE Foundation Grant #2011-118LMF).
And the partial support of the eMadrid project (Investigación y Desarrollo de Tecnologías
Educativas en la Comunidad de Madrid) - S2013/ICE-2715, IoT4SMEs project (Internet of
Things for European Small and Medium Enterprises), Erasmus+ Strategic Partnership nº
2016-1-IT01-KA202-005561), and PILAR project (Platform Integration of Laboratories based on
the Architecture of visiR), Erasmus+ Strategic Partnership nº 2016-1-ES01-KA203-025327.
And to the Education Innovation Project (PIE) of UNED, GID2016-17-1, “Prácticas remotas
de electrónica en la UNED, Europa y Latinoamérica con Visir - PR-VISIR”, from the Academic
and Quality Vicerectorate and the IUED (Instituto Universitario de Educación a Distancia) of the
UNED.
References
1. Plaza, P., Sancristobal, E., Fernandez, G., Castrom, M., Pérez, C.: Collaborative robotic
educational tool based on programmable logic and Arduino. In: 2016 Technologies Applied
to Electronics Teaching (TAEE), Seville, pp. 1–8 (2016)
2. Merino, P.P., Ruiz, E.S., Fernandez, G.C., Gil, M.C.: A wireless robotic educational
platform approach. In: 2016 13th International Conference on Remote Engineering and
Virtual Instrumentation (REV), Madrid, pp. 145–152 (2016)
3. Xu, B., Xu, L.D., Cai, H., Xie, C., Hu, J., Bu, F.: Ubiquitous data accessing method in
IoT-Based information system for emergency medical services. IEEE Trans. Ind. Inform.
10(2), 1578–1586 (2014)
4. Nastic, S., Sehic, S., Vogler, M., Truong, H.-L., Dustdar, S.: PatRICIA – a novel
programming model for IoT applications on cloud platforms. In: Service-Oriented
Computing and Applications (SOCA)
5. Fernandez, G.C., Ruiz, E.S., Gil, M.C., Perez, F.M.: From RGB led laboratory to servomotor
control with websockets and IoT as educational tool. In: 2015 12th International Conference
on Remote Engineering and Virtual Instrumentation (REV), pp. 32–36, 25–27 February
2015
6. Arduino. https://www.arduino.cc/. Accessed 21 Nov 2016
7. NodeMcu ESP8266. http://www.nodemcu.com/. Accessed 21 Nov 2016
8. Onion Omega. https://wiki.onion.io/Get-Started. Accessed 21 Nov 2016
9. NodeMCU v2 - Lua based ESP8266. http://www.exp-tech.de/nodemcu-v2-lua-based-
esp8266. Accessed 21 Nov 2016
10. Onion Omega. https://onion.io/store/. Accessed 21 Nov 2016
11. Currency conversion. http://www.x-rates.com/calculator/?from=USD&to=EUR. Accessed
21 Nov 2016
12. Arduino forum. https://forum.arduino.cc/. Accessed 21 Nov 2016
13. NodeMCU forum. https://www.hackster.io/nodemcu. Accessed 21 November 2016
14. ESP8266 forum. http://www.esp8266.com/viewforum.php?f=17. Accessed 21 Nov 2016
15. Onion community. https://community.onion.io/. Accessed 21 Nov 2016
zamfira@unitbv.ro
Wireless Development Boards to Connect the World 27
16. Orha, I., Oniga, S.: Study regarding the optimal sensors placement on the body for human
activity recognition. In: 2014 IEEE 20th International Symposium for Design and
Technology in Electronic Packaging (SIITME), Bucharest, pp. 203–206 (2014)
17. Thaker, T.: ESP8266 based implementation of wireless sensor network with Linux based
web-server. In: 2016 Symposium on Colossal Data Analysis and Networking (CDAN),
Indore, pp. 1–5 (2016)
18. Škraba, A., Koložvari, A., Kofjač, D., Stojanović, R., Stanovov, V., Semenkin, E.:
Streaming pulse data to the cloud with bluetooth LE or NODEMCU ESP8266. In: 2016 5th
Mediterranean Conference on Embedded Computing (MECO), Bar, pp. 428–431 (2016)
19. Lopes, M., Gomes, I., Trindade, R., Silva, A., Lima, A.C.: Web environment for
programming and control of mobile robot in a remote laboratory. IEEE Trans. Learn.
Technol. PP(99), 1–1
20. Charlton, P., Avramides, K.: Knowledge construction in computer science and engineering
when learning through making. IEEE Trans. Learn. Technol. PP(99), 1–1
zamfira@unitbv.ro
CHS-GA: An Approach for Cluster Head
Selection Using Genetic Algorithm for WBANs
1 Introduction
The recent technological advancements have witnessed vast expansion of WSN
applications in many fields such as Power system applications, Disaster emer-
gency response, Healthcare applications, Air pollution monitoring, Structural
monitoring, Urban temperature monitoring, Precipitation monitoring, Water
pipeline monitoring, Ubiquitous geo-sensing, Commercial asset tracking, Urban
Internet and many more [1]. Sensor nodes are capable of sensing, processing
and transmitting physical, biological and environmental factors such as sound,
temperature and motion. WBANs, an extension of WSNs, consists of low-power,
intelligent, minute, lightweight sensor nodes to monitor the human body func-
tionalities, physiological parameters, physical activities and environmental con-
ditions of the patients. WBANs are useful in many health related application
c Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6 3
zamfira@unitbv.ro
CHS-GA: An Approach for Cluster Head Selection 29
2 Related Work
Many clustering protocols have been developed for WSNs in the recent past
[6–8]. Low Energy Adaptive Clustering Hierarchy (LEACH) [6], is a single hop
clustering algorithm. It reduces energy consumption but the number of dead
nodes increase with the increasing number of nodes which ultimately affects the
network lifetime. In [7], a cluster based routing protocol, Energy Aware Clus-
tering Algorithm (EADC), has been developed to solve the problem of imbal-
anced energy usage at CH by constructing equal sized clusters enhancing network
lifetime. But an overhead occurs in sending large number of control messages.
zamfira@unitbv.ro
30 R. Punj and R. Kumar
In [8], the authors have proposed An Unequal Multi-Hop Balanced Immune Clus-
tering Protocol (UMBIC). It partitions the network into cluster sand constructs
optimum cluster heads and routing tree amongst them. But it works only for
static sensor nodes.
The algorithms proposed for WSNs are not suitable for WBANs as the lat-
ter have some specific properties [3,9]. Therefore, some algorithms have been
proposed for WBANs [10–12]. In [11], routing protocol Anybody for body area
networks has been proposed. It is a self-organized multi-hop routing protocol
and is better than the previously proposed algorithm LEACH [6] in terms of
constant number of clusters with the increasing number of nodes but it does not
take into account the residual energy. In [10], authors had proposed a Hybrid
Indirect Transmission (HIT) algorithm for data gathering that makes use of two
or more clusters and multiple multi-hop indirect transmissions. The authors had
focused on energy consumption and network delay. But the residual energy of the
sensor nodes is not taken into consideration which increases the number of dead
nodes per round. In [12] cluster based epidemic control through smartphone-
based body area networks has been proposed. It is efficient in densely populated
area and closer social interaction zone. It is efficient than traditional epidemic
control methods in dynamic data collection and numerical assumption about
social interaction. But missing data poses a serious threat to any data collection
application.
But the already existing algorithms do not consider residual energy of the sen-
sor nodes which ultimately affects the network lifetime and energy consumption,
so there is need for an efficient cluster head selection algorithm that enhances
the network lifetime and minimizes energy consumption.
In this section, we explain the prerequisites of our problem and describe the
proposed algorithm. In this paper, we propose an efficient cluster head selec-
tion algorithm in WBANs for optimizing energy consumption, load balancing
and enhancing network lifetime. The process of cluster head selection is repre-
sented using the properties of Genetic Algorithm (GA). GA follows the process
of natural evolution and evaluates the fitness of an individual. The fitness value
depends upon the parameters specific to the application. It is multi-objective
optimization criteria for (i) load balancing at CH, (ii) reducing energy consump-
tion and (iii) enhancing network lifetime.
As it has been proved that cluster head selection is an NP-Hard problem [13],
it requires random and optimization techniques to select CH. Thus, we choose
GA for cluster head selection because of its properties such as evolutionary,
convergence and global optimum solution. GA is based on a search procedure
that uses random choice to guide search through a parameter space. GA mainly
require value of the objective function associated with the particular problem in
hand. GA is basically used for optimizing parameters to approach some global
zamfira@unitbv.ro
CHS-GA: An Approach for Cluster Head Selection 31
a. Initialization: In
√ initialization block, CHS-GA randomly selects ten nodes in
the range 2 to n from initial population to explore the genetic diversity in
the search space. Each cluster head serves equal number of cluster members to
achieve fairness and load balancing. Each sensor node has associated with it
the sense radius, Rs and residual energy, Re. The new selected node must be
within the sensing range of the previous node so that the cluster is not widely
dispersed. Thus, saving transmission time. The residual energy of sensor nodes
must satisfy a particular threshold, ET h so that it can be a candidate for
cluster head. The nodes that satisfy the fitness function, are selected as cluster
members. Thus, as an output, a cluster is obtained with 10 cluster members.
b. Fitness Function: Fitness function block is used to determine the quality of
individuals obtained as an output from the initialization block. Each node is
n
evaluated on the basis of the fitness function represented by f = i=1 Rsi Rei
subject to Rsi ≤ Rsi+1 and Rei ≥ ET h where Re is the residual energy, Rs is
the sense radius and ET h is the threshold for residual energy. The best node
is elected as the cluster head. The selected cluster head advertises a message
to the cluster members and maintains a routing table for communication
purposes. As an output, a cluster head is obtained which communicates with
the cluster members using the routing table.
c. Genetic Operators: Genetic Operators are used to select the next generation
on the basis of the previous generation. Selection Criteria, a genetic operator,
zamfira@unitbv.ro
32 R. Punj and R. Kumar
Fitness
evaluation If all
of each sensor Yes
START End
node nodes
added in covered ?
Cluster
Random
Selection No
No
of Sensor Best node
Nodes is elected Next Ge-
as CH neartion
is selected
using
Selection
and
Rsi >
Mutation
Rsi+1
operators
&&
Rei >
ET h ?
Yes
Nodes are
added to
Cluster
4 Complexity Analysis
This section presents the complexity analysis of proposed algorithm CHS-GA.
In this paper, we had analyzed the computational complexity of the proposed
algorithm with respect to time complexity, overhead and fault tolerance.
Lemma 1. The total time to find cluster heads required to cover the whole net-
work is inversely proportional to n and the overhead is O(n).
zamfira@unitbv.ro
CHS-GA: An Approach for Cluster Head Selection 33
Algorithm 2. CHS-GA
1: Repeat ∀n
2: Total Population, Nt = n1 , n2 , n3 , . . . nt
3: while i = 1 to 10 do
4: ni ← Random Selection (Ni )
5: if (Rsi ≤ Rsi+1 && Rei ≥ ET h ) then
6: N c ← Nc n i
7: i=i+1
8: end if
9: Each node maintains a Routing Table (RT) consisting of Node Id, Residual
Energy and Next Hop
10: end while
11: NCH = max(Nc )
12: NCH floods a message to rest of the nodes in cluster
13: a = Multi-Hop nodes for NCH
14: for (a = 1; a ≤ k; a + +) do
15: calc D(ai , ai+1 )
16: end for
17: Nk = nodes are selected using RWS with min(D) for max(k) = 10
18: NextCluster = Nk
19: for (Nk = 1; Nk <= 10; Nk + +) do
20: Goto step 2
21: end for
22: until all the sensor nodes are included in any of the cluster
Proof. It depends upon the number of total sensor nodes (N ) and cluster mem-
bers (n) in the network. These values are pre-defined or user specific, therefore,
the optimal number of cluster heads required to cover the whole network is known
beforehand. The constraints to include any node in the cluster are checked glob-
ally that reduces the overhead of checking conditions at each level. The nodes
that do not satisfy the constraints, i.e., outliers are declared as dead nodes. The
worst case occurs when n is small. It will lead to less number of cluster members
in a cluster and hence more number of cluster head in the network. If n is large
then the physical size of cluster will increase which denies the meaning of clus-
tering. The best case is when n has an optimal value that balances the load at
cluster head as well as manages the physical size of cluster. Also, load balancing
helps in reducing energy consumption at CH.
After CH selection, it advertises control messages to all cluster members
for initiating communication process. CH and all cluster members maintain a
routing table with 3 entries, i.e, node id, residual energy and next hop values.
Since the control messages are transmitted only once, we neglect this overhead.
Worst case occurs when the selected cluster head dies and for re-election of the
cluster head control messages are transmitted again.
zamfira@unitbv.ro
34 R. Punj and R. Kumar
Proof. Cluster members send data to cluster head as soon as the cluster head is
elected. Routing table is updated dynamically. CH is a powerful node and there
is less probability of its failure and being declared as a dead node. CHS-GA is
fault tolerant because when the cluster head fails, all the nodes will be prevented
from sending their data to the dead cluster head. Cluster head keeps track of the
node which is second to it in terms of residual energy with the help of routing
table. Before it completely shreds its energy, it sends a control message consisting
of its residual energy to the second best node and informs the node to act as
cluster head. The newly formed cluster head advertises a control message to all
the cluster members for further data communication. Thus, enhancing network
lifetime.
As AnyBody [11] has constant number of cluster heads and serves unequal
number of cluster members, cluster head shreds its energy soon. This leads to
increase in energy consumption and decrease in network lifetime. Whereas in
the proposed algorithm, cluster head serves equal number of cluster members
which balances load at cluster head. Thus, enhances network lifetime and reduces
energy consumption. As HIT [10] uses chaining of cluster heads for transmitting
data from cluster head to sink, network delay increases which ultimately affects
the total transmission time. Also, it does not take into account the residual
energy of nodes which increases the number of dead nodes affecting network
lifetime. HIT has no mechanism if cluster head dies. Whereas the proposed
algorithm considers the residual energy of sensor nodes. Also, in worst case if
cluster head dies it notifies the cluster members for future communication. Thus,
the proposed algorithm CHS-GA outperforms the existing techniques in terms
of energy consumption, network lifetime and fault tolerance.
An efficient cluster head selection algorithm has been proposed to improve load
balancing and network lifetime at cluster head in WBANs. It uses genetic heuris-
tics for cluster head selection and create equal-sized clusters to balance load at
cluster head. The proposed algorithm checks for nodes which are in vicinity of
the randomly selected nodes so that the nodes lie within each other sensing
radius. The proposed algorithm guarantees for selecting a node as cluster head
with highest residual energy and to find optimal set of cluster heads for the com-
plete network coverage in a time inversely proportional to total number of sensor
nodes. Thus, balancing load at the CH which leads to reduced energy consump-
tion and enhanced network lifetime. Following this line of research, in future this
algorithm can be tested on the moving sensor nodes taking into consideration
the fluctuating distance between the sensor nodes and the base station. Security
mechanism can also be included at the cluster data for secure data transmission
between the source and sink.
zamfira@unitbv.ro
CHS-GA: An Approach for Cluster Head Selection 35
References
1. Rashid, B., Rehmani, M.H.: Applications of wireless sensor networks for urban
areas: a survey. J. Netw. Comput. Appl. 60, 192–219 (2016)
2. Misra, S., Chatterjee, S.: Social choice considerations in cloud-assisted WBAN
architecture for post-disaster healthcare: data aggregation and channelization. Inf.
Sci. 284, 95–117 (2014)
3. Movassaghi, S., Abolhasan, M., Lipman, J.: A review of routing protocols in wire-
less body area networks. J. Netw. 8(3), 559–575 (2013)
4. Hruschka, E.R., Campello, R.J., Freitas, A.A., de Carvalho, A.C.P.L.F.: A survey
of evolutionary algorithms for clustering. IEEE Trans. Syst. Man Cybern. Part C
(Appl. Rev.) 39(2), 133–155 (2009)
5. Gajjar, S., Sarkar, M., Dasgupta, K.: FAMACRO: fuzzy and ant colony optimiza-
tion based MAC/routing cross-layer protocol for wireless sensor networks. Procedia
Comput. Sci. 46, 1014–1021 (2015)
6. Heinzelman, W.B., Chandrakasan, A.P., Balakrishnan, H.: An application-specific
protocol architecture for wireless microsensor networks. IEEE Trans. Wireless
Commun. 1(4), 660–670 (2002)
7. Yu, J., Qi, Y., Wang, G., Gu, X.: A cluster-based routing protocol for wireless sen-
sor networks with nonuniform node distribution. AEU Int. J. Electron. Commun.
66(1), 54–61 (2012)
8. Sabor, N., Abo Zahhad, M., Sasaki, S., Ahmed, S.M.: An unequal multi-hop bal-
anced immune clustering protocol for wireless sensor networks. Appl. Soft Comput.
43, 372–389 (2016)
9. Movassaghi, S., Abolhasan, M., Lipman, J., Smith, D., Jamalipour, A.: Wireless
body area networks: a survey. IEEE Commun. Surv. Tutorials 16(3), 1658–1686
(2014)
10. Culpepper, B.J., Dung, L., Moh, M.: Design and analysis of hybrid indirect trans-
missions (HIT) for data gathering in wireless micro sensor networks. ACM SIG-
MOBILE Mob. Comput. Commun. Rev. 8(1), 61–83 (2004)
11. Watteyne, T., AugéBlum, I., Dohler, M., Barthel, D.: Anybody: a self-organization
protocol for body area networks. In: Proceedings of the ICST 2nd International
Conference on Body Area Networks, pp. 1–6, Florence, Italy (2007)
12. Zhang, Z., Wang, H., Wang, C., Fang, H.: Cluster-based epidemic control through
smartphone-based body area networks. IEEE Trans. Parallel Distrib. Syst. 26(3),
681–690 (2015)
13. Chatterjee, M., Das, S.K., Turgut, D.: WCA: a weighted clustering algorithm for
mobile ad hoc networks. Cluster Comput. 5(2), 193–204 (2002)
14. Goldberg, D.E.: Genetic Algorithms in Search, Optimization and Machine Learn-
ing, 8th edn. Pearson Education, London (1989)
zamfira@unitbv.ro
Proposal IoT Architecture for Macro
and Microscale Applied in Assistive Technology
1 Introduction
Objects around us have been connected for decades. Devices like TV remote
controls and garage door openers have been part of our domestic landscape for
generations. Industrial applications of these technologies-for example, through
remote monitoring and control of production-are also nothing new. In fact,
even the phrase “Internet of Things” or the abbreviation IoT is not a recent
invention [1].
However, recent developments in both networks and devices are enabling
much greater range of connected devices and IoT functionalities. Today, the
phrase “Internet of Things” refers to the world of smart connected objects and
devices. All of this is made possible by the miniaturization of electronic devices,
accompanied by a huge increase in the availability of internet connectivity. The
potential applications of this new IoT are virtually unlimited, and they have the
ability to greatly improve the quality of life of people. Devices allow a user to
change his or her thermostat remotely, dim or increase the intensity of lights,
control door locks, activate alarm systems, etc. While these applications certainly
add a level of fun and convenience for all users, the applications take on a whole
new level of importance when used by persons with disabilities and older adults.
c Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6 4
zamfira@unitbv.ro
Proposal IoT Architecture for Macro and Microscale 37
The implementation of IoT (intelligent network, smart home, smart city, per-
sonalized wearables) can be conceptualized as an ecosystem or scenarios - from
a technical point of view (focusing on norms, protocols or skills) and from a
social perspective (Analysis of social relationships or use cases), case studies
should be considered as user-oriented IoT implementations. Notably, however,
the smart home is a microscale ecosystem, while an smart city is a macroscale.
In both cases, questions about architecture models for implementing Internet
infrastructure of things should be analyzed for each application scenario [4].
The proposed IoT architecture is Service-Oriented Architecture (SOA) for the
construction of software solutions that use as their main element units of devel-
opment called services, which are self-described elements, platform agnostics, that
perform functions and that can range from simple requests to complex processes [5].
The tiered model of the service-oriented architecture provides services consumed
by people or other organizations to execute their activities, enabling the composi-
tion of new services and processes, Fig. 1 shows the conceptual model of the pro-
posed architecture.
The interactions between these components are search, publish, and interac-
tion operations. The service provider represents the layer that hosts the service
by allowing clients to access the service. The service provider provides the service
and is responsible for posting the description of the service it provides. The ser-
vice requestor is the application you are looking for, invoking an interaction with
the service, that is, requesting the execution of a service. Consumers search for
services on the registration server and retrieve information related to the com-
munication interface for services during the development phase or during client
zamfira@unitbv.ro
38 C.S.S. Guimarães Jr. et al.
execution [6]. Tools of middlewares and frameworks are being researched, at first
we will use the Robot Operating System (ROS) as the main framework for the
application server, it provides the services you would expect from an operating
system, including hardware abstraction, low-level device control, implementation
of commonly-used functionality, message-passing between processes, and pack-
age management [4]. Secondary frameworks and midlewares will also be used to
develop scenarios and case studies [8,9]. Future multi-agent layers will be added
in the architecture.
Smart homes refer to the integration of technology and information about the
home network for a better quality of life. Since intelligent machines are equipped
with an automatic environmental control system and with various devices such
as automatic sensors control devices, actuator devices and safety devices [4].
Initially, as a starting point we are using the openHAB framework for some
smart home services [7]. OpenHAB is a software for integration of different sys-
tems and technologies of residential automation into a single solution, and can
act as a central system. This feature makes it interoperable, since openHAB
can communicate with devices that use protocols such as Z-wave, KNX, xPL,
Enocean, MQTT, etc. Being free software, written in Java, it works on top of
any device that can run a JVM. OpenHAB has a web server integrated into its
user interface. Figure 2 shows the OpenHAB architecture.
zamfira@unitbv.ro
Proposal IoT Architecture for Macro and Microscale 39
Smart city can be defined as the use of information and communication technolo-
gies to detect, analyze and integrate as key information of the central systems
in the execution of cities. At the same time, the smart city can make a smart
response to different types of needs, including daily subsistence, environmental
protection, public safety, city services, accessibility, industrial and commercial
activities. In short, “smart city” is the real approach of “smart planet” apply
to the specific region, achieving the informational and integrated management
of cities. It can also be said to be an effective integration of intelligent planning
ideas, intelligent building modes, intelligent management methods, and intelli-
gent development approaches [4].
To test for Smart City, this API is used from Google Maps and OpenGTS [8].
The GPS module sends the global positioning information using a communica-
tion protocol, such as NMEA 0183 protocol In this case, the protocol is based on
American Standard Code for Information Interchange (ASCII) and outputted
serially to the controller that transfers data over a GSM connection to GGSN
(Gateway GPRS Support Node) mobile operator providing the data to a remote
server over a TCP (Transmission Control Protocol), as shown in Fig. 3.
A digital city, refers to remote sensing, global positioning system (GPS),
geographic information systems (GIS) and other space information technologies
as the main means, building the digital city’s geographical information structure,
platform construction Of urban geographic information for the public service.
zamfira@unitbv.ro
40 C.S.S. Guimarães Jr. et al.
Fig. 4. Prototypes of cables for the embedded system. Source: Design and Graphic
Expression (DEG) - UFRGS.
The case study has been developed based on Silva [9], it is an electronic
system to support mobility to replace sight by sound and Vibration [10]. There
are many configurations that can be defined for a design of an electronic walk-
ing stick, Fig. 5 presents the conceptual model numbered with the deployment
diagram Smart Stick.
zamfira@unitbv.ro
Proposal IoT Architecture for Macro and Microscale 41
Fig. 5. Conceptual model with deployment diagram smart stick. Source: Author.
The case study partially describes the design of a Smart Stick for Telemetry
and Telecontrol of an Embedded System applied to the macro and micro navi-
gation of the visually impaired. The conceptual model shows that the embedded
micro navigation system (surrounding environment) is integrated into the Stick,
while the macro navigation (Telemetry and remote control) is adapted to the
visually impaired (board computer wearable or mobile phone). This separation
takes the weight of the electronic cane and also divides the processing.
For this case study, a motorized wheelchair will be used that are assembled with
modules, sensors and controllers [10]. In this way, a chair can connect with the
scenarios and interact with the environments. The information is made available
to an application server for which the system uses the services appropriate for the
needs of the users and working conditions [11]. Figure 6 presents the deployment
diagram Smart Wheelchair.
zamfira@unitbv.ro
42 C.S.S. Guimarães Jr. et al.
Wheelchair users can enjoy of intelligent control that allow the diversion of
obstacles and avoid irregular terrain, executed automatically, providing greater
safety and comfort for macro and micro navigation. The goal of the instrumenta-
tion is to develop a smart, low-cost wheelchair that, guided by a sensor network,
can avoid obstacles and prevent mistaken actions regardless of user actions.
Fig. 7. Partial screen prototypes for the smart home and smart city scenarios. Source:
Author and Wagner Silveira.
The system back-end allows administrator access to all content and function-
ality. A central panel is developed with shortcuts to the most common options
and is divided by menus that allow to manage contents and functionalities. To
access with administrator profile must be informed in the authentication screen
an identifier referring to the administration, the system will provide features
related to this profile. The system must allow access to all configurations refer-
ring to the smart stick and smart wheelchair, in the reporting area can be gener-
ated statistics referring to the quantity of devices available by city and generate
detailed report [6].
zamfira@unitbv.ro
Proposal IoT Architecture for Macro and Microscale 43
5 Conclusions
The project is in the process of heuristics and has been gaining space for the
development of an IoT architecture with scenarios of applications connected to
AT to assist in the orientation and mobility of people with disabilities. The
project is in the research, modeling and definition phase of hardware and soft-
ware devices, prediction for the first tests for the second half of 2017, with
partial system. Future work includes the improvement of the project as a whole,
development of embedded hardware, new services for assistive technologies and
multi-agent systems for adaptive scenarios.
Acknowledgement. The authors would like to thank to CAPES, this work has been
funded by the research project PROCAD Assistive Technologies.
References
1. G3ict: Internet of Things: New Promises for Persons with Disabilities (2015).
http://g3ict.org/resource center/publications and reports/p/productCategory
books/subCat 2/id 335. Accessed 7 Sept 2016
2. Domingo, M.C.: An overview of the Internet of Things for people with disabilities.
J. Netw. Comput. Appl. 35, 584–596 (2015)
3. Gubbi, J., Buyya, R., Marusic, S., Palaniswami, M.: Internet of Things (IoT): a
vision, architectural elements, and future directions. J. Fut. Gener. Comput. Syst.
29, 1646–1658 (2013)
4. Robot Operating System: Open-source collection of software frameworks for robot
development (2016). http://www.ros.org/. Accessed 11 Nov 2016
5. Vermesan, O., Friess, P.: Internet of Things. Converging Technologies for Smart
Environments and Integrated Ecosystems, p. 17363. European Commission,
Belgium (2013)
6. Erl, T.: Service-oriented architecture. In: Concepts, Technology, and Design, pp.
83–280. Indianapolis, Indiana (2005)
7. Botta, A., Donato, W., Donato, W., et al.: Integration of cloud computing and
Internet of Things: a survey. J. Fut. Gener. Comput. Syst. 56, 684–700 (2016)
8. openHAB: A vendor and technology agnostic open source automation software for
your home (2016). http://www.openhab.org/features/architecture.html. Accessed
16 Sept 2016
9. Guimarães, C.S.S., Pereira, C.E., Henriques, R.V.B.: Telemetry and remote an
embedded system applied in macro and micro blind navigation. Paper presented
at the 11th international conference on remote engineering and virtual instrumen-
tation, Porto, Portugal, pp. 424–433, February 2014
10. Silva, R.F.L.: Integrated product design to urban design: long Bengal electronics.
Dissertation, Federal University of Santa Catarina, Brazil (2009)
11. Lee, E.A., Seshia, S.A.: Introduction to Embedded Systems. A Cyber Physical
Systems Approach, pp. 93–370. University of California, Berkeley (2011)
12. Marques, P.J.: Proposal of a wheelchair position determination system in an intel-
ligent environment. Dissertation, Federal University of Rio Grande do Sul, Brazil
(2014)
13. Open GPS Tracking System: Open-Source GPS Tracking System - OpenGTS
(2016). http://www.opengts.org/. Accessed 11 Aug 2016
zamfira@unitbv.ro
Using Industrial Internet of Things to Support Energy
Efficiency and Management: Case of PID Controller
Tom Wanyama ✉
( )
1 Introduction
The Proportional Integral and Derivative (PID) is the most used automatic controller of
industrial processes today. This controller requires its parameters to be adjusted
according to the nature of the process. This adapting of the controller to the process is
called controller tuning. The focus of tuning is usually minimizing the error between
the desired value of the process variable (PV) and the setpoint (SP). However, it is
generally agreed in literature that most PID controllers are not properly tuned which
affects the performance as well as the energy consumption of the controlled system. In
traditional manufacturing systems, PID controllers as well as their associated data are
usually separated from the energy monitoring and control systems, making it difficult
to relate the controller performance parameters and the process energy consumption.
But modern industrial network technologies through the paradigm of Industrial Internet
of Things (IIoT) make all of the information throughout manufacturing facilities acces‐
sible to those who need it, whenever they need it, wherever they are [1]. This makes it
zamfira@unitbv.ro
Using Industrial Internet of Things to Support Energy Efficiency and Management 45
possible to integrate process control and energy monitoring information into single
business automation applications such as Microsoft Excel, enabling real-time associa‐
tion of PID performance and process energy consumption.
In this paper we present an IIoT that integrates industrial process control, energy
monitoring data, and utility electricity pricing information using industrial networking
technologies. The industrial process component of the IIoT is based on PID control of
a pilot scale heat exchanger using a Micrologix PLC that has Ethernet IP communication
capability. The energy consumption of the industrial process is monitored by the energy
component of the IIoT using an IEC61850 SEL751A relay. The data from the PLC and
the relay is sent to DataHub OPC client from where it is accessed by the business appli‐
cation; in this case Microsoft Excel. In addition, this paper describes how control and
energy data is processed in a single Microsoft Excel file, in real-time, showing the effect
of PID settings on the energy consumption of the controlled system. Providing such
information to machine operators and plant managers ensures that they know the impact
of the way they operation their PID controllers on energy consumption.
The rest of this paper is arranged as follows: Sect. 2 covers the background of data
access for the PID controlled pilot heat exchanger. In Sect. 3 we present a case study
and Sect. 4 deals with the testing and results of the case study. Section 5 covers the
conclusion.
2 Background
Industrial networks are the backbone of Industrial Internet of Things (IIoT). In fact, IIoT
is the integration of sensors, industrial controllers and computers, cloud computing
systems, big data technologies, and advanced data analytics systems using network
(industrial networks) and web technologies. The development of IIoT is based on the
philosophy that smart machines are better than humans at accurately and consistently
capturing and communicating data, and at analyzing that data to generate actionable
information. This information should enable companies to pick up on inefficiencies and
problems sooner, saving time and money and supporting business intelligence efforts.
In manufacturing specifically, IIoT holds great potential for quality control, sustainable
and green practices, supply chain traceability and overall supply chain efficiency.
Although IIoT has the potential to improve the overall industrial supply chain efficiency,
the focus of this paper is the use of IIoT to show the effects of PID tuning and process
control on energy efficiency of industrial systems.
Industrial networks are used in many industrial domains including but not limited to
manufacturing, electricity generation, transmission and distribution, food processing,
transportation, water distribution and waste management, oil and gas production [5].
Each industrial domain has its own slightly different networking requirements, leading
to differences in the associated network protocols. This creates a problem of integrating
zamfira@unitbv.ro
46 T. Wanyama
data from different domain, since devices with different network protocols cannot
communicate with each other. The Open Platform Communication OPC is the solution
to this problem. Figure 1 shows that OPC defines the standard for the interface between
industrial data servers and clients. The clients can be Human Machine Interface appli‐
cations, MES or ERPs. In addition, the figure shows that servers have specific network
protocol drivers that enable them to communicate with industry specific controllers [3].
As the manufacturing sector changes its business paradigm from “maximum gain for
minimum capital” to “maximum value from minimum resources”, energy efficiency is
becoming one of the most important forms of “alternative energy”. Therefore there is
need to focus on energy saving and optimization from design to management of manu‐
facturing processes [4]. Energy saving and optimization through operational and
management techniques is like a marathon, rather than a sprint, with savings measured
in hour-to-hour and day-to-day increments. What enables energy optimization is the
continuous seeking of answers to the questions such as:
• When and why did a machine exceed typical energy draw?
• Why did equipment changeover cause startup surges?
• Why did component change extend the production cycle into a peak-draw period?
Therefore, energy as well as its quality has to be continuously monitored down to
the manufacturing and process lines.
zamfira@unitbv.ro
Using Industrial Internet of Things to Support Energy Efficiency and Management 47
The ability to integrate energy and manufacturing process data facilitated by IIoT
brings about a search for answers to a new set of questions that can increase energy
efficiency and optimization. Such questions include:
• How does energy consumption change with process control strategy?
• What happed to energy consumption when controller (e.g. PID Kp, Ki and Kd) param‐
eters were changed?
• What is the difference in energy consumption during dynamic state and stead state?
• How do system dynamics affect energy quality and consumption?
Moreover, IIoT enables the posting and automatic updating of the energy pricing
information, including time-of-use. This constantly reminds plant operators and
managers the importance of production timing to the overall cost of energy. In general,
IIoT supports the integration of energy efficiency performance criteria into production
management systems such as Manufacturing Execution Systems (MES) and Enterprise
Resource Planning (ERP) applications as an enabler of energy efficient manufacturing
processes [2].
zamfira@unitbv.ro
48 T. Wanyama
Where:
Kp is the proportional gain, Kp = KC Controller gain (unit-less).
Ti is the reset time (seconds-1).
Td is the rate time (seconds).
Two hundred seventy six tuning rules have been identified for ideal PID controller
structure described by Eq. 2 [8], and each of these rules has difference energy cost
performance. The dependence of energy efficiency performance of controlled systems
on the PID architecture, and on the tuning rules, increases the importance of monitoring
energy performance of PID controllers in real time.
3 Case Study
This section presents a model IIoT system that uses a PID to control a pilot scale heat
exchanger. The PID controller is deployed on a Micrologix PLC and the energy supply
of the system is monitored using a SEL751A relay. The system data is collected into an
Excel file over Ethernet, using OPC technology.
Our pilot scale heat exchanger unit shown in Fig. 3 is heated using an ON-OFF controlled
blow dryer (see Fig. 4). A PID controlled fan is used to cool the unit to the desired
temperature that supports the transfer of heat to air flowing through a copper pipe that
is inside the exchanger chamber. Incoming air into the chamber is at room temperature,
while outgoing air is at preset temperature. The focus of this paper is the control of
temperature inside of the heater exchanger chamber.
zamfira@unitbv.ro
Using Industrial Internet of Things to Support Energy Efficiency and Management 49
Figure 4 shows the process diagram of our pilot heat exchanger system. The setpoint
(SP) Ts (t) is the desired temperature of the exchanger chamber, and the manipulated
variable (MV) f (t) is the 0–5 V analog input to the KT-5194 DC Motor PID speed
controller. But in this system, the controller is used in open loop PWM control mode,
were the 0–5 V input signal determines the value of the 0–24 V (10 A-maximum) PWM
zamfira@unitbv.ro
50 T. Wanyama
output. The 0–24 V power supply to the fan DC motor is the control variable (CV) c(t)
of our PID loop.
The temperature inside the heater transfer chamber is measured using an RTD probe
whose resistance varies from 100 Ω at 0 °C to 220 Ω at 300 °C. The RTD signal is fed
in a signal conditioner that produces a proportional 0–10 V analog signal. The signal
conditioner output is the input to a Micrologix 1400 PLC that has an ADC the converts
the analog signal to 0–4095 digital variable. This variable is scaled to produce the actual
temperature of the chamber, which is the process variable (PV) Ta (t) of the PID loop [6].
The output of our PID controlled is given by Eq. 2.
[ ( )]
1 t d Ta (t)
Output = Kc e(t) + ∫ e(𝜏)d𝜏 + Td + bias2 (2)
Ti 0 dt
We set the feedforward bias to zero, and the process variable (PV) Ta (t) is equal to
the setpoint (SP) Ts (t) plus the process error e(t) for and indirect process such as cooling.
Since the setpoint is a constant, the derivative of the process variable is equal to the
derivative of the process error (see Eq. 3). With bias = 0,
( ) ( )
d Ta (t) d Ts (t) + e(t) de(t)
= = , if Ts (t) = Constant3 (3)
dt dt dt
Substituting Eq. 3 into Eq. 2 result into our PID output being given by an equation
similar to Eq. 1.
zamfira@unitbv.ro
Using Industrial Internet of Things to Support Energy Efficiency and Management 51
The process and energy data of our model IIoT system is accessed using KEPServer
OPC server. The server has multiple drivers including Ethernet IP and IEC61850 MMS.
These drivers are configured as channels to deliver the associated data to OPC clients.
This is necessary because OPC servers usually do not poses advanced data access
features such as HMI, alarms and event handling, data logging and historian, and process
data tunneling and bridging. It is OPC clients that are normally utilized to provide these
features. In our model IIoT system, we use OPC DataHub [6] to access data from
KEPServer and provide a Human Machine Interface (HMI) for the system, as well as
Dynamic Data Exchange (DDE) to a Microsoft Excel file.
Microsoft Excel has powerful features that allow the user to directly query databases
and websites. Our model IIoT system uses the Web query feature of Excel to retrieve
refreshable information that is stored on the electricity utility company’s web site (see
Fig. 5). The pricing data is extracted from the information using Macros programmed
in Visual Basic. Then the date is analyzed and integrated with process and energy data
using the tools in Excel.
4.1 Testing
During testing, the pilot scale heat exchange was set to maintain an operating temper‐
ature of 32 oC in the heat exchanger chamber. This temperature was to be achieved by
heating the chamber with a blow dryer, while cooling it with a PID controlled fan. Test
zamfira@unitbv.ro
52 T. Wanyama
settings such PID parameters (Kp, Ki and Kd), temperature setpoint, and PID mode were
done through the HMI shown in Fig. 6. In addition, the HMI provided the means for
monitoring the following performance measures of the system: System supply voltage,
fan controlled voltage, power consumption, and actual heat exchanger temperature.
Fig. 6. HMI of the heat exchange system with PID in manual mode
The heat exchanger was tested with PID mode set to manual for a period of 5 min.
Its power consumption was sampled every 30 s and sent to a Microsoft Excel sheet in
real time, through Cogent DataHub OPC client. Thereafter, the heat exchanger was
tested with the PID mode set to automatic, and its power consumption was logged and
stored in an Excel sheet. In both test cases, the trapezoidal rule was used to calculate the
energy consumption of the heat exchanger.
4.2 Results
Figure 6 shows the HMI of the heat exchanger with the PID set to manual, while Fig. 7
shows the HMI with PID in automatic mode.
Figure 8 shows the power consumption of the 24 V DC, 1.7 A fan motor when
controlled by PID in manual and in automatic modes.
zamfira@unitbv.ro
Using Industrial Internet of Things to Support Energy Efficiency and Management 53
Fig. 7. HMI of the heat exchange system with PID in automatic mode
zamfira@unitbv.ro
54 T. Wanyama
Figure 9 show the real time display of the energy data of the heat exchanger and
energy cost in a Microsoft Excel sheet. This is made possible by the use of IIoT tech‐
nologies.
Fig. 9. Real time display of heat exchanger energy data in excel sheet
The trend chart in Figs. 6 and 7 show that the PID automatic mode provides smooth
system control than the manual mode. Moreover, Fig. 8 shows that the power consump‐
tion of the motor under manual model has higher variance that when the PID is in auto‐
mation mode. This is expected since the manual model in essentially ON-OFF control.
Not that ON-OFF switching (control) of machinery such as Heating, Ventilation and
Air Condition equipment courses power quality issue.
The area of the graphs in Fig. 8 represent the energy consumed by the motor in 5 min.
Our calculations show that the area under the manual mode graph is equivalent to
1.5 × 10−3 kWh, while the area under the automatic model graph is equivalent to
0.833 × 10−3 kWh. This means that automatic PID loop is over 40% more efficient than
the manual loop; leading to over 40% reduction in energy cost.
zamfira@unitbv.ro
Using Industrial Internet of Things to Support Energy Efficiency and Management 55
5 Conclusion
In this paper we present an IIoT that integrates industrial process control, energy moni‐
toring data and utility electricity pricing information using industrial networking tech‐
nologies. The industrial process component of the IIoT is based on PID control of a pilot
scale heat exchanger using an Ethernet IP enabled Micrologix PLC. The energy
consumption of the industrial process is monitored by the energy component of the IIoT
that has an IEC61850 SEL751A relay. Furthermore, this paper describes how the IIoT
enables the processing of control and energy data in a single Microsoft Excel file, in
real-time, showing the effect of PID settings on the energy cost of the controlled system.
Providing such information to machine operators and plant managers ensures that they
know the impact of the way they operation their PID controllers on energy consumption.
References
1. Bunse, B., Kagermann, H., Wahlster, W.: Industry 4.0: Smart Manufacturing for the Future,
Germany Trade and Invest, Berlin, German, July 2014. http://www.gtai.de/GTAI/Content/EN/
Invest/_SharedDocs/Downloads/GTAI/Brochures/Industries/industrie4.0-smart-
manufacturing-for-the-future-en.pdf. Available as of 12 April 2015
2. Bunse, K., Vodicka, M.: Managing energy efficiency in manufacturing processes –
implementing energy performance in production information technology systems. In: Berleur,
J., Hercheui, M.D., Hilty, L.M. (eds.) What Kind of Information Society? Governance,
Virtuality, Surveillance, Sustainability, Resilience. IFIP Advances in Information and
Communication Technology, vol. 328, pp. 260–268. Springer, Heidelberg (2010). doi:
10.1007/978-3-642-15479-9_25
3. Burke, T.J.: OPC Unified Architecture Interoperability for Industry 4.0 and the Internet of
Things, OPC Foundation. https://opcfoundation.org/wp-content/uploads/2016/05/OPC-UA-
Interoperability-For-Industrie4-and-IoT-EN-v5.pdf. Available as of November 2016
4. European Communities, ICT and Energy Efficiency: The Case for Manufacturing, Office for
Official Publications of the European Communities, Luxembourg (2009). ISBN
978-92-79-11306-2
5. Galloway, B., Hancke, G.P.: Introduction to industrial control networks. IEEE Commun. Surv.
Tutorials 15(2), 860–880 (2013). Second Quarter
6. Kafuko, M., Wanyama, T.: Integrated hands-on and remote PID tuning laboratory. In:
Proceeding of the Canadian Engineering Education Association Conference, Hamilton,
Ontario, Canada, June 2015
7. Likins, M.: PID Tuning Improves Process Efficiency, Yokogawa Corp. of America. http://
www.yokogawa.com/technical-library/resources/media-publications/pid-tuning-improves-
process-efficiency/. Available as of November 2016
8. O’Dwyer, A.: Reducing energy costs by optimizing controller tuning. In: Proceedings of the
2nd International Conference on Renewable Energy in Maritime Island Climates, Dublin,
Ireland, pp. 253–258, April 2006
zamfira@unitbv.ro
MODULARITY Applied to SMART HOME
From Research to Education
1
Transylvania University of Braşov - AOSR Academy, Braşov, Romania
udoru@unitbv.ro
2
Transylvania University of Braşov, Braşov, Romania
andrei.c.neagu@gmail.com
3
Transylvania University of Braşov - ASTR Academy, Braşov, Romania
csam@unitbv.ro
4
Transylvania University of Braşov - Benchmark Electronic, Braşov, Romania
jingavlad@yahoo.com
1 Introduction
Energy usage and its resulting impacts on our environment became one of the major
concerns humanity is facing today. Depletion of fossil fuels, the impacts on the envi‐
ronment from mining those fuels, and the spectre of global warming exacerbated by
burning them are critical reasons for us to become more responsible for the energy we
consume.
As one report of The Intergovernmental Panel on Climate Change shows, the indus‐
trial activities that our modern civilization depends upon have raised atmospheric carbon
dioxide levels from 280 parts per million to 379 parts per million in the last 150 years.
The panel also concluded there’s a better than 90% probability that human-produced
greenhouse gases such as carbon dioxide, methane and nitrous oxide have caused much
of the observed increase in Earth’s temperatures over the past 50 years. According to
this report the rate of increase in global warming due to these gases is very likely to be
unprecedented within the past 10,000 years or more [1].
Generating energy requires precious natural resources, for instance coal, oil or gas
while reducing energy consumption has lots of benefits – we can save money and help
zamfira@unitbv.ro
MODULARITY Applied to SMART HOME 57
protect and preserve our environment. Therefore, using less energy helps us to preserve
these resources and make them last longer in the future.
In the light of these facts above, we believe energy must be in future a concern of
all citizens and new tools must be created in support of our environment, starting even
from every household.
It must be noted that energy service demand may also reflect changes in the level of
comfort and lifestyle requirements of households. Specific energy consumption is
defined as the energy required to maintain a particular level of energy service in house‐
holds. It is a modelled alternative to energy intensity, and takes account of changes in
demand for individual energy services (such as level of household comfort or hot water
use), and helps to remove the impact of higher and lower external temperatures on energy
use.
In this paper we focused on energy consumption in Belgium and Netherlands.
According to VEA Flemish Energy Agency, the average energy consumption of a person
per day is 50 kWh and from that 71% represents heating [2].
Let`s take into consideration that most companies involved in reducing energy
consumption and environmental pollution try to minimize energy consumption by
raising the efficiency of their systems and improving the buildings (heating devices
manufacturers such as Daikin or Viessmann, campaigns like the US Solar Decathlon),
or increasing user’s control over their systems (Google’s Nest, Smappee, smart metering
systems of Daikin and Veissmann). The disadvantage of these systems is high cost and
significant changes in the construction of the building (Fig. 1).
30000
25000
20000
kWh/year
15000
10000
5000
0
Brussels region Flemish region Walloon region Belgium
natural gas 16480 20712 18733 19644
fuel oil 16604 27902 26176 26489
wood 20202 22099 21350
coal 19110 22582 21803
Propane 21234 15946 18608
Fig. 1. Average total energy consumption (kWh/year, dwelling) per principal energy source per
dwelling per region and for Belgium (survey results) [2]
Figure 2 describes the heating losses starting from the boiler till the end user. As we
can see there are three main factors that are involved in the heating process: the first two
ones are the Boiler and the Building, components that most of the companies mentioned
zamfira@unitbv.ro
58 D. Ursuţiu et al.
above are dealing with; the 3rd one, and we might say the most unpredictable one is the
third one, people’s behavior. In this research we will focus to provide a technological
solution, an energy monitoring system which can engage people in a more responsible
way to save energy in the cope of lowering energy costs.
To address the above issues, monitoring the habitant’s behaviour and informing them
about their own way of using energy, this paper describes such a monitoring system
designed to validate, collect and connect the users with the status of the building. During
our study we compared different ways of building a wireless modular system (here we
focused on the wireless technologies available on the market) and with the information
collected we had identified and validate how the flow of input data collected can
contribute in habitants perception of using energy.
Regarding the hardware, our focus is to provide a reliable and also cheap solution
for first validation stage. As to the modular configuration and the limits of the system.
During our research we evaluate several scenarios and validate our assumptions by case
studies conducted in two buildings.
Our aim during the research is to find the best solution for creating and testing a moni‐
toring system dedicated to student house owners. The focus of the system is on the
heating structure and heating losses, which is the most expensive cost for our market.
Because our modular system should not affect the building structure, heating pump
structure and should be easy to install, the following are required:
– a wireless communication structure
– collecting information of air flow in each room.
zamfira@unitbv.ro
MODULARITY Applied to SMART HOME 59
In order to be able to validate the requirements mentioned above and also to validate
the principals of creating a modular system for this purpose we defined several hardware
requirements for the alfa product: low power wireless communication protocol, two
temperature sensors and one contact sensor.
Figure 4 presents the complete structure of the alfa system we are proposing. The flux
of information is marked with the blue arrow and is composed by:
1. The “Gateway” described above - the link between the building and the cloud
(formed by Xbee module, the main logic board);
2. The “Sensor” – 2 temperature sensors, 1 contact sensor, the link board and the Xbee
module;
3. The “Cloud” – database;
4. User interface – website.
A full descriptions of the elements mentioned above and also the role of each of
them, please refer to the full paper “Evaluating the reliability and scalability of a wireless
energy monitoring system in buildings” [3].
In order validate the system capabilities and purpose we created an end to end data
flow concept. And so the working pattern of the sensors to the users interface is described
in Fig. 4. This process has 4 steps:
1. Starting with the sensors, every 30 s we collect samples from the two temperature
sensors (TMP36) and the contact sensor. Through the Xbee Explorer the data is
zamfira@unitbv.ro
60 D. Ursuţiu et al.
streamed to the Xbee module. The Xbee module packs data in frames and renders
the information ready for the transmission.
2. The frame is send wirelessly to The Gateway. Acting as a coordinator, the Xbee
module receives the frame and forwards it to the Arduino Ethernet Board through
the Arduino Xbee Shield.
3. The Arduino board unpacks the frame and pushes the raw data through the Ethernet
port to the Carriots database to be stored.
4. After each week the data is interrelated and processed manually. The result then is
placed on a website, using charts and easy to understand description.
Figure 5 Presents the first part of the System – “The Sensor”, module placed in the
rooms, the transmission node.
zamfira@unitbv.ro
MODULARITY Applied to SMART HOME 61
In order to validate the system and the impact that it may have to the habitants we
installed it in two different buildings. We collected data during for several weeks by not
informing the habitants about the system presence and purpose and after that we inform,
let them challenge each other in ways of saving energy.
During one month of collecting and stored data from each building, with the scope
to present to the house owner, in an easy to understand way, the result of energy
consumption in each room we developed a website including some charts and few
recommendations on how to save energy was developed for that purpose.
Based on the input from our three sensors we managed not to just inform the owner
on the consumption but also to evaluate the habits of the residents. Figure 6 illustrates
three days consumption by data collected from one of the experimental rooms (Room4).
In this case we can clearly see that the room average temperature is above the normal
comfort zone of 21 °C while open window usage is not energy efficient. The tenant, in
most of the situations, turns the heater at the maximum and opens the window.
Fig. 6. Data collected from one of the devices during three days (red- heater temperature, green
– room temperature, blue – window open/close).
In order to have a clear view of those habits and based on the average of the inputs,
we created a profile of each user. For that we made an average of temperatures collected
and movement open/closed of the window based on the time line.
Figure 7 describes user profile. Furthermore, providing to users tips on how to be
more environmental friendly, we can clearly see the improvements in energy usage of
the users, represented by the green line. Even if we did not get the same involvement
for each tenant, as Error! Reference source not found. shows, it is obvious that, by
concentrating more on those tips we can get better results for the future developments.
zamfira@unitbv.ro
62 D. Ursuţiu et al.
In order to have a better understanding of the Fig. 7 the red line represents the usage
habit during the first three days of testing and the green one in the last days, after
providing the tips. Subtracting the two, in Fig. 6 we can see the improvement in user
habits. This result was obtained in one room, Room 4.
5 Evaluation Methodology
To evaluate system coverage and redundancy we applied several key methods to simu‐
late and test it in a real case scenario. First we used the WHIPP tool to simulate the
coverage of the system focusing on the reception sensitivity. By using this tool we were
able to have a first understanding of the distance that might be between the devices. The
next step is represented by a RSSI measurement conducted in a real environment. The
test was made using X-CTU software, allowing us to understand which links are reliable
and where an extra device is needed. The last step is a mash routing redundancy. In that
phase we evaluated the system capability of establishing a new link in case of power
loss. There are several key methods that we applied to test and also to determine the
maximum range of our system.
zamfira@unitbv.ro
MODULARITY Applied to SMART HOME 63
In our case we used WHIPP tool to define the exposure limitation of the system in
the tested buildings. We started by creating a plan of the building. For that purpose we
used the preset material such as dry-wall, wood doors and the xbee sensors JN516x with
the transmitting power of 3 dBm. The elevation of the sensors, for all the simulation,
was set to 1.5 m. Taking into consideration that the position of the devices is strictly
related to the heating modules in the building we focused on prediction coverage simu‐
lation.
Received signal strength indicator (RSSI) is the signal strength level of a wireless device
measured in dBm of the last received packet [5]. The main idea behind the RSS system
is that the detected signal strength value reduces within the distance travelled. In free
space, the RSS degrades with the square of the distance from the sender [6]. Using the
Friis transmission equation, the ratio of the received power Pr (dB) to the transmission
power Pt (dB) can be expressed as:
( )2
𝜆
Pr = Pt × Gt × Gr
4𝜋d
where, Gt(dB), Gr (dB) are gain of transmitter and gain of receiver respectively, λ is the
wavelength, and d (m) is the distance between the sender and receiver. It can be seen
that the larger the wavelength of the propagating wave is, the less susceptible it is to
path loss. The received signal strength is converted to RSSI which can be defined as the
ratio of the received power Pr (dB) to the reference power PRef (dB).
zamfira@unitbv.ro
64 D. Ursuţiu et al.
Pr
RSSI = 10log
PRef
Using X-CTU software, one of the XBee modules is configured as Coordinator while
the other as a Router. After pairing with the Coordinator, the Router starts transmitting.
Once the Coordinator has received the data packets successfully, it sends back an
acknowledgment (ACK). To obtain the RSSI value the software takes the average of
100 RSSI results from 100 packets of 32 bytes each. Hence, the RSSI value is measured
after sending 100 packets of 32 bytes each, and then the average is used to generate
RSSI.
The distance between the Routers and the Coordinator was variable. We applied
different case scenarios depending of the building where we tested the system. As
presented in Figure Building A and Figure Building B we conducted static tests by
placing the coordinator setup in the main hallway of each floor. The average distance
between floors in case of Building A is 2.7 m and for Building B 2.3 m. For each floor
setup we tested the link to each device in the system, one by one, in order to determine
the relationship between RSSI values and the distances.
A ZigBee mesh network configuration is done automatically and flawlessly by the XBee
devices. The Coordinator starts a ZigBee network, and other devices then join the
network by sending association requests. As we described in the second chapter, ZigBee
networks are considered self-forming networks due to their ability of self-routing.
After forming the mesh network, to relay the message from one device to another,
the most optimized path is selected. However, if one of the routers becomes damaged
or otherwise unable to communicate due to power loss, the network can select an alter‐
native route.
One of the most important characteristics of ZigBee mesh networking has been its
self-healing capacity, the ability to create alternative paths when one node fails or a
connection is lost through mesh routing [7]. In order to test this attribute of mesh routing,
we observed the elapsed time between the elimination of one path and the search and
formation of another. To perform the test, we captured messages received on Coordi‐
nator from the network. Then we powered off one by one each Router until the link was
disconnected. This experiment determines where a repetitive (Router) device is needed
in order to guarantee the redundancy of the system.
Taking into consideration the results of the mesh redundancy we observed that without
the device from room 2, the network cannot communicate with the device from room
1. In order to compare the results of the WIHPP tool with the ones conducted in real
situation we simulated the scenario A without the device from room 2, Fig. 9b.
zamfira@unitbv.ro
MODULARITY Applied to SMART HOME 65
Below, in Fig. 9 are presented the two scenarios: figure (a), with the device from
room 2 and picture (b) without it. There is a change in color between the two simulations.
The first one has a green – yellow color but the second one a darker green, marked with
the red oval. According to the tool this color represents a rage of sensitivity between
−44 to −50 dBm and in real situation test it will show a loss of connection.
6 Experiment Conclusions
As it was shown in this paper, the performance of the proposed system, different
scenarios were undertaken in order to measure the performance of the monitoring system
network, with a focus on the mesh routing redundancy and the RSSI level of the Xbee
modules, comparing the results with a software simulation. The results showed that the
performance of the network highly depends on distance range between devices and
indoor environment. A reliable communication is two floor distance (15 m) or the
communication can be lost or impossible to set.
As a general conclusion of the design and testing phases of the proposed alpha
monitoring system, the study shows the performance of the system as a tool to monitor
and optimize energy consumption. After one month of collecting data for the proposed
system, by installing the system in student houses, we managed to define habits of the
building users. This helped to obtain a profile for each room and the informing approach
on the students lead us to validate the system’s purpose to save energy. By using the
monitoring system we managed to obtain savings between 8% and 21% and to make the
first steps in creating a standard regarding the energy habits of the room user [3].
This result bring even more value to the scope of saving energy as the cost of using
the system is one small investment for an indefinite period of time. It comes with a
relatively simple structure and usage, having a user friendly interface and a low cost,
while the energy savings on the long term can bring significant reduction on energy cost
for the user and protection of the environment in terms of conservation of natural
resources needed to produce energy.
zamfira@unitbv.ro
66 D. Ursuţiu et al.
7 Future Development
We are looking to explore the possibilities of using different technology like the
PSoC Analog Coprocessor provided by Cypress as a layer of processing multiple sensors
inputs. PSoC Analog Coprocessor integrates programmable analog blocks, including a
new Universal Analog Block (UAB), which can be configured with GUI-based software
components. By using this technology we aim to upgrade the ways we can design custom
analog front ends for sensor interfaces.
Fig. 11. CY8CKIT-048 PSoC from Cypress and CK-USB-04A from IQRF
zamfira@unitbv.ro
MODULARITY Applied to SMART HOME 67
More than that, due to our strong collaboration with Microrisc and to their complex
structure of collecting, cloud storage and visual online platform, we intend to integrate
into our system the IQRF transition technology. Till now the IQRF technology help us
in our need of scaling the range of the system and on top of that, form our first test we
were able to reduce to 20% the energy consumption of each node. As presented in
Fig. 11 we started our test by using CY8CKIT-048 PSoC from Cypress and CK-
USB-04A from IQRF.
Regarding the user approach we want to improve the web platform in order to give
better access to each student. We also want to create a friendly way to send real time
messages to students with possible actions. In order to protect the concept we will look
for a solution to move all the processing algorithms from the hardware. Extra features
will be also implemented in the same device, like CO2 and humidity sensors. This
information will add even more value to our system providing quality of the air status
in an indoor environment.
References
zamfira@unitbv.ro
Development of M.Eng. Programs with a Focus
on Industry 4.0 and Smart Systems
zamfira@unitbv.ro
Development of M.Eng. Programs with a Focus on Industry 4.0 and Smart Systems 69
zamfira@unitbv.ro
70 M.D. Justason et al.
2 Motivation
The School of Engineering Practice and Technology is well positioned to prepare grad‐
uates for employment in the manufacturing sector. The School’s location in Hamilton,
Ontario is central to Canada’s manufacturing industry.
Producing university graduates with skills that are immediately applicable is a chal‐
lenge in many industries with rapidly-changing technology, and this is especially true
in the manufacturing industry [1]. It is particularly evident in the area of automotive
engineering [2]. SEPT has already implemented undergraduate programs that address
this challenge; namely Automotive and Vehicle Technology, Process Automation Tech‐
nology, and Manufacturing Engineering Technology. Industry 4.0-based Master’s
Programs will provide students with a continuing pathway to graduate-level programs.
It is also intended to facilitate the pathway to professional licensure for graduates of the
aforementioned undergraduate programs, but is also open to graduates of traditional
undergraduate engineering programs as well as international students. The pathways
created by the new M.Eng. programs are shown in Fig. 1.
Experience
References
Professional PracƟce ExaminaƟon
PROFESSIONAL LICENSURE
Fig. 1. Pathways
zamfira@unitbv.ro
Development of M.Eng. Programs with a Focus on Industry 4.0 and Smart Systems 71
amount of work experience required for licensure. Typically, the M.Eng. program will
count as 1-year of the required 4-years of work experience.
The new M.Eng. Programs within the School of Engineering Practice and Tech‐
nology have content and delivery methods consistent with the Vision of the new school.
The School’s Vision can be characterized by the following elements: industry-driven,
hands-on, case-studies, in course projects, advanced methods and technologies, inno‐
vative teaching methods, sustainability, community-focused, professional development,
communications, management, design, problem-solving, and integration of professional
and technical skills.
With this Vision in mind the motivation for introducing these three new focus areas
to the Master of Engineering in Manufacturing Engineering (MEME) can be organized
into four main areas:
1. Opportunities for Students
2. Opportunities for Faculty
3. Opportunities for Partners
4. Opportunities for the Faculty of Engineering.
The creation of the new M.Eng. programs also provides the School with an oppor‐
tunity to educate students in areas that are complementary to the technical aspects of
Industry 4.0. Successful Industry 4.0 implementation involves aspects of a business
‘outside’ the functions related directly to the manufacturing process. Business consid‐
erations such as human resource management, accounting and finance, strategy, culture,
and leadership all play a role in the successful implementation of Industry 4.0. This
supports the idea of a T-Shaped graduate, with a broad knowledge of business that is
outside their specific technical area [3]. This concept is particularly important in the area
of human resource management and supply-chain management [4, 5].
zamfira@unitbv.ro
72 M.D. Justason et al.
• A clear pathway into a Master’s program will offer Bachelor of Technology graduates
an improved pathway to professional licensure.
• Completing an M.Eng. degree offers Bachelor of Technology graduates the oppor‐
tunity to participate in the Ritual of the Calling of the Engineer (the ‘Iron-Ring’).
• The new focus areas will remain accessible to graduates of more traditional Bachelor
of Engineering programs, both from McMaster and elsewhere.
These new focus areas within the M.Eng. Programs in the W Booth School of Engi‐
neering Practice and Technology create the following opportunities for the School’s
partners:
• Feeder colleges to the DCP programs can offer students a direct pathway from college
through to an M.Eng. degree, and ultimately to professional licensure
• Community Partners (Companies, Organizations, and Government):
– Richer engagement with groups of undergraduate and graduate students
– Working with a broader spectrum of potential co-op or full-time employees
prescreened through project engagement.
zamfira@unitbv.ro
Development of M.Eng. Programs with a Focus on Industry 4.0 and Smart Systems 73
3 Methodology
The activities described in this section were carried out by a SEPT committee called the
M.Eng. Task Force. This group of six SEPT faculty members, plus 1 representative from
the School of Graduate Studies, carried out monthly meetings from approximately
mid-2015 to mid-2016. This section outlines the committee’s activities.
The first step in the development of the new M.Eng. focus areas was to engage students
and alumni in market research. The results of a survey that included responses from 354
B.Tech. students, 342 B.Eng. students, and 146 alumni are shown below. A study of
competing programs at nearby Universities was also completed.
• B.Tech. Students–50% indicated a desire to pursue an M.Eng. degree
• B.Eng. Students–more than 80% indicated a strong interest in an M.Eng. degree
• Alumni–approximately 66% of McMaster Engineering Alumni living within a 1-hour
commute of McMaster indicated a strong interest in an M.Eng. degree
• M.Eng. programs at other Ontario Universities are popular; even ‘over-subscribed’
• Based on historical enrollment numbers in the existing M.Eng. programs, there is
typically a large demand from international students (>60% of existing M.Eng.
students are international students).
zamfira@unitbv.ro
74 M.D. Justason et al.
Based on the results of the market research, it was evident that the demand for M.Eng.
programs was strong among all target groups. This market research encouraged the
M.Eng. Task Force to continue its activities.
3.2 Implementation
To facilitate a January 2017 implementation of the new Master of Engineering in Manu‐
facturing Engineering (MEME) focus areas, the new focus areas needed to be structured
within the framework of the existing program. It was not possible to seek approval for
a completely new program structure as this could take up to 2-years. The existing
framework for the MEME program was as follows:
• Students can take up to two graduate level courses from the Mechanical, Materials,
and Chemical Engineering Departments.
• Each student must complete a project at a manufacturing company plus six graduate
level courses (or 8-graduate level courses without a project).
• Courses from Departments other than the three ‘approved’ Departments (Mechan‐
ical, Materials, and Chemical) must be approved on a case by case basis.
• All other courses must be taken within SEPT.
The details of the implementation suggested that it was possible to offer the new
MEME focus-areas for students starting the program in January 2017 provided they
elected to complete the 6-course plus project option. The 8-course option would need
to be implemented in the Fall of 2017 due to the requirement to create and seek approval
for additional (new) courses within SEPT.
zamfira@unitbv.ro
Development of M.Eng. Programs with a Focus on Industry 4.0 and Smart Systems 75
The launch of a fourth Industry 4.0 focus area is targeted for September 2018. This focus
area would be in ‘Digital Solutions’.
A second M.Eng. theme-area tentatively referred to as ‘Smart Cities’ is also targeted
for September 2018. Specializations may include: Civil Infrastructure, Biotechnology,
and Power and Energy.
zamfira@unitbv.ro
76 M.D. Justason et al.
References
1. Schuh, G., Gartzen, T., Rodenhauser, T.M.A.: Promoting work-based learning through
industry 4.0. In: The 5th Conference on Learning Factories 2015, Bochum (2015)
2. Riel, A., Tichkiewitch, S., Stolfa, S., Kreiner, C., Messnarz, R., Rodic, M.: Industry-academia
cooperation to empower automotive engineering designers. In: 26th CIRP Design Conference,
Stockholm (2016)
3. Schumacher, A., Erol, S., Sihn, W.: A maturity model for assessing industry 4.0 readiness and
maturity of manufacturing enterprises. In: Changeable, Agile, Reconfigurable and Virtual
Production, Stockholm (2016)
4. Hecklau, F., Galeitzke, M., Flachs. S., Kohl, H.: Holistic approach for human resource
management in Industry 4.0. In: 6th CLF - 6th CIRP Conference on Learning Factories (2016)
5. Huxtablea, J., Schaefera, D.: On servitization of the manufacturing industry in the UK. In:
Changeable, Agile, Reconfigurable and Virtual Production, Bath (2016)
6. Stock, T., Seliger, G.: Opportunities of sustainable manufacturing in industry 4.0. In: 13th
Global Conference on Sustainable Manufacturing- Decoupling Growth from Resource Use,
Ho Chi Minh City (2016)
7. Faller, C., Feldmuller, D.: Industry 4.0 learning factory for regional SMEs. In: The 5th
Conference on Learning Factories 2015, Bochum (2015)
zamfira@unitbv.ro
Remote Acoustic Monitoring System
for Noise Sensing
1 Introduction
According to the European Commission, a large majority of European Citizens
are living in urban environments [1], being approximately in 2010 the 50.5% of
the world’s population [2]. This trend, far from diminishing, in increasing year
by year. The United Nations Population Division reported in 1990 that in this
decade there were ten “mega-cities” with 10 million inhabitants or more. In 2014,
the number of mega-cities was 28 (representing about 12% of the world’s urban
dwellers). By 2030, the world is projected to have 41 mega-cities with 10 million
inhabitants or more [3].
From the perspective of the emerging economies, mega-cities will become
the largest markets for consuming new technology products, due to the need of
the authorities to apply the so-called ICT (Information and Communications
Technologies) to deal with problems related to the economy, buildings, mobility,
energy, citizens, planning and governance of the cities. It is in this scenario where
the concept of smart cities has been growing up for the last ten years in order
to provide solutions to the new challenges posed by these urban areas.
c Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6 8
zamfira@unitbv.ro
78 U. Hernandez-Jayo et al.
The idea of smart cities comprises a wide range of control and actuators
systems aimed to improve the habitability and perception that citizens have of
cities. A smart city covers many of these systems, ranging from applications
that facilitate the governance of cities and encourage citizens’ participation to
services specifically focused on improving their quality of life. Among these sys-
tems, we can highlight those using ICT to improve the environment of the city,
but not only from an air quality perspective, but also to control the noise levels
of the city. It is assumed by the World Health Organization that environmen-
tal noise has emerged as the leading environmental nuisance triggering one of
the most common public complaints in many Member States of the European
Union. The European Union tries to face the problem of environmental noise
with international laws and directives (as the European Noise Directive [4]) on
the assessment and management of environmental noise [5].
In this context, the work in progress presented in this paper shows the ICT-
based approach that the University of Deusto and the Ramon Llull University are
developing jointly in the frame of the Aristos Campus Mundus initiative, with
the goal of obtaining real time information about the equivalent noise levels
(Leq ) of a city. For that purpose, the developed tool will be able to monitor
remotely noise levels at different points of the city using the public bus system
as a mobile sensors network.
The paper is structured as follows. In Sect. 2, the related work regarding
mobile acoustic monitoring is detailed, and in Sect. 3, its challenges are pointed
out. In Sect. 4, the hardware approach proposed to address the problem of mobile
acoustic monitoring is explained, and in Sect. 5 the first acoustic signal processing
algorithms developed to face this challenge are described. In Sect. 6, we detail
the expected outcomes of the collaborative project, and, finally, in Sect. 7 we
focus on the conclusions and future work.
Traditional noise measurements in cities have been mainly carried out by pro-
fessionals that record and analyze the data in a certain location typically using
certified sound level meters. This approach allows reliable analyses but is costly,
hardly scalable and difficult to follow rapid changes of urban environments.
To address these drawbacks, in the last decade, several approaches of systems
focused on the monitoring of environmental noise have been proposed (see [6]
and references therein). Their main goal has been oriented to the development
of small equipments assuring the reliability of the acoustic measurements. More-
over, these systems have been designed to allow scalability by reducing the
cost of the hardware (i.e., low-cost acoustic sensors) and improving the net-
work data communication in order to tailor a noise map by means of mobile
acoustic monitoring.
One of the first experiences in mobile acoustic monitoring is detailed in [7],
where a mobile sensing unit (MSU) associated with a Global Positioning System
(GPS) is used to perform acoustic monitoring in Seoul (South Korea). With a
zamfira@unitbv.ro
Remote Acoustic Monitoring System for Noise Sensing 79
reduced set of sensors, the Seoul ubiquitous sensing project conducted a wide
range of tests across several city locations. The MSU were even installed in cars
and buses moving around the city by following repetitive circuits. These nodes
measured temperature and humidity, and also noise level. However, no signal
processing is included for the latter (or at least detailed) neither in the sensors
nor in the network hub.
In [8], the system is based on an array of sensors carried by a vehicle driving
along the streets of the city to acquire measurements from different locations.
The goal of that piece of research is estimating the locations and the power of
the stationary noise sources of the locations of interest. For this purpose, the
data gathered by the array is post-processed before plotting the several sources
in the noise map, but no details are given about the vehicle noise treatment.
Dekoninck et al. [9] focus on the study of low density roads, including both
mobile and fixed noise monitoring platforms. The proposal is based on perform-
ing the mobile measurements by bicycle, which provides a new view on the local
variability of noise and air pollution based on computing the differences of mea-
surements along road segments [10]. This proposal is easily applicable to any
other cities to monitor both noise and air pollution at the cost of having enough
bicycle riders.
zamfira@unitbv.ro
80 U. Hernandez-Jayo et al.
events from real data using coupled matrix factorization of spectral representa-
tions and class annotations. Finally, in [16], the authors exploit deep learning
methods to detect acoustic events by means of using the spectro-temporal local-
ity. For more references and details about these approaches, the reader is referred
to [13].
To conclude, the most challenging issue for the problem at hand is appropri-
ately integrating the bus and surrounding traffic noise levels to the noise map
due to their similarity so as to compute the Leq value correctly. This increases
the complexity of the separation system significantly, which can be addressed
thanks to having the reference signal, that is, considering as input the noise gen-
erated by the bus for the sound source identification and subsequent integration
in the noise map computation.
4 Hardware Approach
The hardware system designed to deploy the ubiquitous acoustic sensor net-
work to remotely monitor the noise pollution of a city is based on three main
subsystems:
zamfira@unitbv.ro
Remote Acoustic Monitoring System for Noise Sensing 81
The current hardware prototype developed for the remote acoustic monitor-
ing system is shown in Fig. 1, which represents the scope of the whole system,
showing the following set of subsystems:
– FRDM-KL25Z embedded system
– CMA-4544PF-W omnidirectional capsule microphone with auto gain control
based on the MAX9814 amplifier
– ESP8266 WiFi communications module
– Adafruit FONA 808, which is and all-in-one mobile communication interface
plus a GPS module
zamfira@unitbv.ro
82 U. Hernandez-Jayo et al.
ClassificaƟon
Audio MFCC
Signal Coefficients
Windowing Filter bank Log
FFT IDCT Select
filtering |·|
After obtaining the MFCC of the input audio signal, they are input to a
machine learning algorithm for their classification (see Fig. 2). Although a myriad
algorithms can be found in the literature for this purpose, initially we consider
Fisher Linear Discriminant algorithm [18] following the work described in [11].
This is due to the fact that this method will give us the coefficient of participation
of each type of noise in the overall noise picture, in our case, the bus engine noise
and the road traffic noise.
We plan to test this approach using real data measurements in a bus driving
its route, recording audio samples from the traffic noise surrounding the vehi-
cle. Moreover, in order to have a general picture of the vehicle own noise, we
also plan to collect audio samples of the several specific driving sounds (brake,
engine, throttle, etc.) that occur on the bus route. Depending on the obtained
results, other feature extraction methods and machine learning algorithms could
be considered to solve the challenge in the future.
6 Expected Outcomes
From the mobile audio processing algorithms, we expect to identify the con-
tribution of the bus noise to the total traffic noise. This value will be used to
balance the noise contribution of the vehicle itself, in order to compensate the
short distance to the measuring equipment. This way, the final value of Leq cor-
responding to the traffic noise can be evaluated with the suitable contribution
of the bus noise in order to obtain reliable real-time noise maps from the routes
of the selected bus lines.
To develop a robust mobile audio processing approach, it is necessary to
properly configure and program the hardware and the firmware of the elec-
tronic embedded system deployed on the buses. This will require a good balance
between the accuracy of the signal acquisition, the number of collected samples
zamfira@unitbv.ro
Remote Acoustic Monitoring System for Noise Sensing 83
and the time needed to process them. In a first approach, our design will do
it on board, sending the results of the signal processing block to the central
server. Then, these data will be integrated and displayed as a noise map of the
city through the web-based GIS. If the performance of the embedded system
is not good enough to obtain an accurate noise map, then we will consider the
possibility of conducting the signal processing on the web server, using a more
powerful processor at the cost of increasing the communications cost.
The goal of the research under development is to obtain a remote acoustic mon-
itoring system that allows to know in real time the city noise impact. For this
purpose, the acoustic sensor (a microphone) is connected to a mobile embedded
device with signal processing and data transmission capabilities. This embedded
system is designed to be installed on a public bus, so urban noise is captured
along the route travelled by the bus. Therefore, the larger number of sensors
deployed in buses lines, the more detailed information will be obtained to gen-
erate the noise map of the city. Finally, the collected information is sent to a
central server that runs a web-based GIS application designed to display the
collected noise levels in real-time.
After validating the viability of the ICT-based approach, the research team is
planning a set of field operational tests to evaluate the accuracy and reliability
of the proposed system. These tests are designed in order to: (i) characterize
the natural noises of the bus inside and outside: engine, breaks, throttle, etc.;
(ii) improve the processing of the acquired noise by applying the necessary error
corrections according to the values obtained in the first tests and weighting the
bus noise appropriately.
In these terms, the proposed system is intended to allow making a signifi-
cant step from current noise maps of cities, which are updated with a certain
periodicity (usually, every five years) through fixed measure points in the city
by following the European Noise Directive.
References
1. European Union Transport Themes, Clean Transport and Urban Mobil-
ity. http://ec.europa.eu/transport/themes/urban/urban mobility/index en.htm.
Accessed Nov 2016
zamfira@unitbv.ro
84 U. Hernandez-Jayo et al.
zamfira@unitbv.ro
Testing Security of Embedded Software Through Virtual
Processor Instrumentation
Abstract. More and more functionality that demands remote access on a vehicle
is integrated into modern cars. Fleet management, infotainment, updates-over-
the-air and the upcoming functionality for autonomous driving need gateways
that enable a car-2-x communication. Misuse is a threat. Consequently, security
mechanisms play an increasing important role. But how can we show and prove
the effectiveness of these security functions?
Therefore, in this paper we will show an approach to test security aspects,
based on virtual instrumentation. The approach is to use a framework that
executes the application under development on a virtual model of the target micro
controller. An interception library generates scenarios systematically, whereas
the effects on registers and memory are monitored. We are intercepting the
running software at vulnerable functions and variables to detect potential
malfunctions. This will detect security vulnerabilities of all internal failure even
if no malicious behavior at the interfaces occur.
1 Motivation
Within the last decade mobility has undergone major changes. One is the advent of data
exchange between cars and infrastructure. Instead of a vehicle being a standalone
mechanical device it has been transformed to a mobile platform with extensive electronic
sensors and computing power. Nowadays within a car a large amount of data is available
in form of sensor data, representing the state of the vehicle as well as its understanding
of the surrounding. By exchanging such information with others, new concepts for
efficient driving, optimizing traffic flow (see Kramer in [1]), and new comfort functions
become possible.
On the other hand many new threats are generated. The increasingly technology
allows the transmission of large amounts of data in real time to transmit states for diag‐
nosis and software updates over the air will be possible. I.e. the EE topology will get
accessible via an air interface. Therefore the vehicles may offer new attacking surfaces,
as some examples already show today.
It has already been shown how modern cars can be attacked and controlled without
having physical access to these vehicle [2, 3]. These attacks allow the manipulation of
car’s brakes and driver assistance systems or remote eavesdropping on conversations
zamfira@unitbv.ro
86 A. Lauber and E. Sax
held within the car. They are just a few examples of possible attacks. With an even more
connected car an even broader attack vector might be created.
To secure the vehicles against possible attacks security mechanism needs to be
implemented which has been a research focus within the last couple of years. Unfortu‐
nately not all attack surfaces can be closed by integrating security mechanism. Attack
vectors can also arise through sloppy implementations and inexperienced programming.
To overcome these issues functional testing and testing for security weaknesses are
necessary.
This paper is structured as follows: First we give a short overview of state of the art
security testing in Sect. 2. Afterwards we categorize the attacks on systems and point
out the important test cases in Sect. 3. Thereafter the virtual instrumentation and
processor interception is explained in Sect. 4. The interception lead to the security testing
framework shown in Sect. 5. At the end we conclude with a summary and give an outlook
for future work.
In theoretical security analyzes one must distinguish between high-level design analyzes
and detailed analyzes. In design analysis, protocols, interfaces, and specifications are
analyzed by reviewers to find and resolve systematic vulnerabilities such as bad encryp‐
tion or short keys. While only a theoretical description of the system has to be present
during design analyzes, one needs explicit knowledge about the implementation of the
algorithms for the detailed analyzes.
Theoretical security analyzes cannot detect any implementation errors due to misin‐
terpretation of the specification or errors from third-party software. To protect the system
against this type of error, Bayer recommends in [4] secure software development stand‐
ards. These can be achieved by means of the various standards and coding guidelines.
Even if errors are reduced, errors by specification or errors in third-party software, can
only be found by explicit tests of the functions in the overall system.
For static code analysis the source code is automatically analyzed by means of formal
criteria, to identify volatility errors. Static code analysis can identify implementation
errors, but functional errors or design errors cannot be found by this analysis. In addition,
Knechtel described in [5] that these kind of analyzes are unreliable. He suggests the use
of explicit code reviews for sensitive functions. Another option to find weaknesses is
the system test on the real platform or a hardware prototype.
zamfira@unitbv.ro
Testing Security of Embedded Software 87
Fuzzing is a technique that has been used for some time to test software in IP networks.
To do this, the implementations are subjected to an unexpected, invalid, or random input,
with the hope that the target will react unexpectedly to identify new vulnerabilities. The
responses to such attacks range from strange behavior of interfaces, unspecified behavior
of the system to complete crashes.
As a rule, the fuzzing can be divided into three steps. First, the input data is generated.
This can either be structured according to the specification or completely random. The
data are then fed into the system interfaces and the output is monitored. As a last step,
the recorded behavior must be analyzed by experienced programmers in order to identify
potential weaknesses. Disadvantage of fuzzing is that only the interfaces of the system
are monitored. Faulty states within the system cannot be detected.
While the above tests can be automated, the penetration test is a test method using human
testers. These tests attempt to exploit known vulnerabilities and gain access to the
system. The appropriate approach is usually based on years of experience by human
testers who perform these tests. An example of typical penetration tests is exploiting
undocumented debugging interfaces to gain access to buses and internal signals, but also
by opening the controller and directly accessing the silicon, the testers are looking for
information on possible attack vectors, according to Bayer in [4]. The knowledge
provided for the testers usually range from no information, access to the specification
to all information about the source code. Therefore, these tests can be used for black
box, gray box and white box tests.
The state of the art security testing can usually only be automated for independent
functions without the interaction of all functions in the complete system. Knechtel writes
zamfira@unitbv.ro
88 A. Lauber and E. Sax
in [5] that attacks are rarely due to weaknesses of individual keys or algorithms but rather
by weaknesses of the entire system. I.e. for security testing of the overall system,
including third party software, the overall system needs to be present. Further the internal
state of this system needs to be monitored in addition to the external interfaces. This
leads us to use virtual instrumentation of a processor running the software under test.
Finally, it should be noted that all practical security tests cannot guarantee complete
coverage. Therefore a compromise between test effort, time and completeness must be
made. I.e. practical security test serve only as a complement to theoretical security
analyzes and the consideration of security in the design phase.
3 Categorization of Attacks
As Radzyskewycz writes in [7], it is not a question of whether systems are attacked, but
when. Therefore it is important to implement security mechanisms according to the state
of the art. In addition, Wheatley in [8] describes that 44% of all attacks will be done over
known vulnerabilities.
The Symantec cooperation describes in [9] the loss or theft of passwords, incidental
ties and insider knowledge as other important causes of intrusion into systems. Only a
very small part of the attacks on systems are carried out by unknown vulnerabilities at
the time of attack. A distribution of the given causes for attacks is shown in Fig. 1.
Especially due to the large number of attacks with known vulnerabilities, it is impor‐
tant to design new software in such a way that known vulnerabilities are no longer
present. To ensure this, a software must be regularly tested against known vulnerabilities
during the development cycle. This must include all known security gaps, because the
old wisdom from project management is even more important in the field of security:
“The later a problem is detected, the higher is the cost to fix it.”
In the PC world, vulnerabilities are stored in a database of the MITRE Cooperation
on behalf of U. S. Department of Homeland Security. This Common Vulnerabilities and
Exposures Database [10] saves all known security gaps in existing applications. By the
year 2016, about 100,000 attacks on various systems were recorded in this database. In
addition, the MITRE Cooperation provides a database for the overview of all known
zamfira@unitbv.ro
Testing Security of Embedded Software 89
vulnerabilities in the Common Weakness Enumeration Database (CWE) [11]. There are
currently about 1,000 different vulnerabilities in this database. In the CWE, the weak
spots are divided into different categories. A classification of the attacks from the year
2015 leads to the distribution shown in Fig. 2. The most common attacks that are listed
in the CWE are so called Denail of Service (DoS) attacks. These attacks make 33% of
all known attacks on today’s systems. The goal is to get the attacked system to crash
and thus destroy the functionality of the system.
More important than DoS attacks are attacks where an attacker can gain control over
the entire system. Buffer overflows with 22% and code execution with 24% have a special
significance. With so called buffer overflows (or overflows), memory areas are written
with too long data sets to overwrite the following data records in the memory, thereby
manipulating the contents of these variables. For an overview of attacks by buffer over‐
flows, see Foster in [12].
The principle of buffer overflows is also used in code execution. Whereupon not only
variables are overwritten, but the return address is set to a malicious, injected code of a
function. This not only generates influence of the behavior by changing the variables
but also take control over the entire system and execute malicious code.
Reason for the above attacks is usually a badly implemented software. Especially
the consistent check of the memory area for dynamic variables can prevent overflow in
most cases. However, due to time and memory space requirements in embedded systems,
this is often omitted. One reason for DoS attacks is often the division by zero, which is
not uniformly specified in microcontrollers and can therefore lead to different behavior
or even program termination.
In addition to the above, there is often undefined behavior in software development
when dereferencing so called null pointers that do not point to any memory, using
memory or objects after executing “free”, or read access to unallocated memory. In most
cases, the aforementioned problems can be avoided by means of consistent queries in
the programming, but the queries are rarely implemented for runtime and memory
reasons.
Not all problems can be found by individual testing of the independent modules.
Security weaknesses most often arise from the interaction of different modules and
therefor the overall system needs to be tested as a whole.
zamfira@unitbv.ro
90 A. Lauber and E. Sax
Instruction set simulators like Open Virtual Platforms (OVP) [13] can be used to model
a processor with the corresponding peripherals and run the cross-compiled application.
Running the cross-compiled application inside an instruction set simulator gives the
same behavior as on the target platform. The virtual prototyping of embedded systems
for OVP is described by Werner in [14]. Werner also compares OVP with other platforms
for virtual prototyping for embedded systems. He further explains in [15] the usage of
OVP for debugging cross-compiled applications to build a virtual test environment.
The Imperas binary interception tool as defined in [16] can cause the simulation to
stop the application and run the interception library at any point in time. This includes
among others the interception of the virtual platform before each instruction is morphed,
specific instructions are executed and a specific address range is read or written.
The interception technology is usually used for verification, analysis and profiling,
including detection of memory corruption, deadlocks, data races or memory usage. As
Imperas Software Limited writes in [16] this is especially useful when many different
data scenarios should be executed.
With the binary interception tool we can use our own library to examine the state of
the internal registers, instructions, memory, and other periphery. Furthermore, a replace‐
ment of the simulated behavior with a behavior defined in the interception library is
possible. This means if the interception library detects a specific behavior during simu‐
lator the corresponding instruction is either replaced or extended by the one defined in
the library.
The advantages of using the novel framework with interception libraries compared
to other debug interfaces is that no additional code needs to be inserted in the application
and no special access to the processor is needed. I.e. no resource overhead in the appli‐
cation and no additional instructions are executed. The application will be cross-
compiled as running on the real hardware platform without any additionalities. Another
advantage is that all parts of the interception technology will run in parallel to the simu‐
lation of the virtual platform.
As mentioned above we need to monitor the memory in order to find overflows and
the instructions to find zero divisions. Both can be done by running an interception
library in parallel to the main application.
An overview of the test framework can be found in Fig. 3. The platform for the virtual
processor will be described in a platform model file as described by Werner in [14]. The
virtual processor will consist of a processor with registers and memory for heap and
stack, local memory for code and variables, as well as peripherals. The interception
library will have direct access to these registers, memory, instructions, and peripherals.
The location of the variables inside the registers and memory will be configured in a
configuration file. Further, this file holds information about the intercepted instructions
and functions. We are generating this file with information of the source files from the
application. Therefore the source files are parsed and variables will be detected. The
supervision of instructions will be done during run time with the disassembled applica‐
tion code, searching for divisions and illegal memory access.
zamfira@unitbv.ro
Testing Security of Embedded Software 91
With the interception library [16] we can monitor all instructions on assembler level and
check each of them if we need to observe the corresponding instruction. The behavior
of the interception depends on the instruction. Our novel approach will intercept only
potential vulnerabilities and directly executed all other instructions. E.g. we look at the
different assembler instructions, if we find a division (either udiv for unsigned or sdiv
for signed division) the corresponding registers will be checked for zero division. If the
denominator is zero the execution of this command will be stopped and an error message
will be displayed. If the instruction is not a division the interception library will not be
executed.
To find potential vulnerabilities by zero divisions the observation of instructions can
be implemented as static interceptions, because the instructions are well known at
compile time and will be constant for all applications.
The same concept can be done with the memory. Each memory access (read and write)
will be monitored and an error message will be displayed if data is written to the wrong
memory range. The address range of the variables are stored in the interception config‐
uration (see Fig. 3) and access to these ranges will be observed. If a write access across
the variable boarders occur (buffer overflow) an error message will be displayed.
For the heap and stack monitoring we need the memory tracing, because the local and
dynamic variables will be stored at the local memory. Further we need the function
tracing to trigger the interception whenever a function is called and new variables are
stored on stack or heap.
The local variables will be pushed to the stack, therefore the instruction monitoring
needs to add the variables to the dynamic monitoring, thus the interception library knows
zamfira@unitbv.ro
92 A. Lauber and E. Sax
the address and range of the new variables. The same will be done for dynamically
allocated or deallocated memory inside the heap. This memory is usually allocated or
deallocated with malloc and free. Another observer will find write access to the function
return address to detect illegal code executions.
Both the dynamic memory observation and the observation of the return address
needs to be done during runtime. Therefore a dynamic part of the interception library is
necessary, that can be extended by the simulation.
The security testing is based on the cross-compiled code for the target platform. I.e. the
instructions order and the behavior is the same as on the real platform, since no further
or different optimization of other compilers will be done. Further the Executable and
Linkable Format (elf) file that is used for the testing can be flashed to the target device
without any additional changes. Current state of the art tests (see above) are focusing
on the source code without compiler optimization.
Even if the testing framework can check the source code using a static analysis before
cross-compiling and running the application on the virtual processor, we are not focusing
on this, since static code analysis is state of the art. This novel approach can even be
used to run the compiled application without having access to the source code of the
application. I.e. black box tests for security can be executed. But nevertheless informa‐
tion about functions and variables are needed in order to build the configuration file.
In the next step, after static code analysis, the application is checked for variables
and functions. The static variables and functions will be added to the interception
configuration. With this information the interception library is build and passed to the
instruction set simulator. If the defined interceptions occur the simulator will stop the
execution of the application and run the functions provided by the interception library.
The Imperas instruction set simulator is used to run the defined test cases and the
interception library. For this step a model of the target platform is needed. This should
include all necessary processors, memories, registers, and peripherals (see above). The
interception library will stop the running cross-compiled application in the simulator at
every predefined interception. Further if new memory is dynamically allocated the
interception library will be extended to observe this memory area as well. After deal‐
location of the memory the corresponding entry in the interception library will be
deleted.
At last the results of the simulation and the test process will be shown for documen‐
tation. The total workflow of the virtual instrumentation for security testing can be seen
in Fig. 4.
zamfira@unitbv.ro
Testing Security of Embedded Software 93
In this paper we analyzed the different security weaknesses and derived from CWE that
the most common ones are buffer overflow, code execution and division by zero.
According to this knowledge we did a conceptual design for a security test framework
based on virtual instrumentation. We build an interception library that monitors the
memory and the instructions and reports security weaknesses if they occur.
Future work will investigate the concepts to automate the tests using virtual plat‐
forms. Further the memory observer for the variables and the test cases should be gener‐
ated automatically. The interception library should be used to generate test cases
according to the interfaces and a weakness database (CWE). These test cases will based
on the information of variables (including dynamic variables) and functions from the
application.
Another work will be done to use the framework for black-box-testing. This means
without any knowledge of the source code. Especially the observation of dynamic vari‐
ables have to be researched.
Protection of Shared memories and multi-processor systems can be tested as well.
The virtual framework will be extended for the usage of multi-processor systems in the
future.
zamfira@unitbv.ro
94 A. Lauber and E. Sax
References
1. Kramer, J., Hillenbrand, M., Müller-Glaser, K.D., Sax, E.: Connected efficiency–a paradigm
to evaluate energy efficiency in tactical vehicle-environments. In: Bargende, M., Reuss, H.C.,
Wiedemann, J. (eds.) 16. Internationales Stuttgarter Symposium. Proceedings, pp. 1451–
1463. Springer, Wiesbaden (2016). doi:10.1007/978-3-658-13255-2_107
2. Koscher, K., et al.: Experimental security analysis of a modern automobile. In: IEEE
Symposium on Security and Privacy, pp. 447–462 (2010)
3. Checkoway, S., et al.: Comprehensive experimental analyses of automotive attack surfaces.
In: USINEX Security Symposium (2011)
4. Bayer, S., Enderle, T., Oka, D.-K., Wolf, M.: Automotive security testing—the digital crash
test. In: Langheim, J. (ed.) Energy Consumption and Autonomous Driving. LNM, pp. 13–22.
Springer, Cham (2016). doi:10.1007/978-3-319-19818-7_2
5. Knechtel, H.: Methoden zur Umsetzung von Datensicherheit und Datenschutz im vernetzten
Steuergerät. ATZ Elektronik 10(1), 26–31 (2015)
6. Spillner, A., Linz, T.: Basiswissen Softwaretest: Aus- und Weiterbildung zum Certified
Tester; Foundation Level nach ISTQB-Standard, 4th edn. dpunkt.verlag (2010)
7. Radzkewycz, T.: Automotive networks can benefit from security. In: Connected Vehicle
Journal: Designing for Next-Generation Connected and Autonomous Vehicles (2016)
8. Wheatley, M.: Known vulnerabilities cause 44 percent of all data breaches. http://
siliconangle.com/blog/2016/01/12/known-vulnerabilities-cause-44-percent-of-all-data-
breaches/. Accessed 31 Oct 2016
9. Symantec Corporation: Internet Security Threat Report. 2013 Trends, vol. 19 (2014)
10. MITRE Corporation: Common Vulnerabilities and Exposures (CVE). https://cve.mitre.org/.
Accessed 31 Oct 2016
11. MITRE Corporation: Common Weakness Enumeration (CWE). https://cwe.mitre.org/.
Accessed 31 Oct 2016
12. Foster, J.C., Osipov, V., Bhalla, N.: Buffer Overflow Attacks: Detect, Exploit, Prevent.
Syngress Publishing Inc., Rockland (2005)
13. Imperas Software Limited: Open Virtual Platforms: The source of Fast Processor Models &
Platforms. http://www.ovpworld.org/. Accessed 15 Dec 2016
14. Werner, S., et al.: Cloud-based design and virtual prototyping environment for embedded
systems. Int. J. Online Eng. (IJOE) 12(9), 52–60 (2016)
15. Werner, S., Lauber, A., Becker, J., Sax, E.: Cloud-based remote virtual prototyping platform
for embedded control applications: cloud-based infrastructure for large-scale embedded
hardware-related programming laboratories. In: Proceedings of 2016 13th International
Conference on Remote Engineering and Virtual Instrumentation (REV). IEEE (2016)
16. Imperas Software Limited: Imperas Binary Interception Technology: User Guide, no. V1.5.3
(2016)
zamfira@unitbv.ro
Virtual and Remote Laboratories
zamfira@unitbv.ro
LABCONM: A Remote Lab for Metal Forming
Area
Florianópolis, Brazil
lucasboeira@ifsc.edu.br, lucasccasagrande@gmail.com,
{vilson.gruber,roderval.marcelino}@ufsc.br,
schaefer@ufrgs.br
1 Introduction
Conceptual learning in the metal forming area is complex and difficult because it
involves knowledge of symbols, equations, theories, principles and procedures. This
complexity is natural, because basically knowledge is an abstraction of reality, result of
experiments, analyzes, studies, and standards.
It is believed that observation, interaction, practice, and experimentation are edu-
cational practices that can complement and enhance the learning process in engineer-
ing, making theory more meaningful for the student. Without interaction, students are
passive and the learning process becomes slower [1]. Doubtless, experimentation
establishes a relationship between practice and theory [2]. For this reason, experiments
are essential in the teaching process, especially in engineering and experimental sci-
ences [3].
A solution to provide experimentation and interaction as a way to aid learning has
been the development of experiments and remote laboratories [4]. With this new
concept of laboratory, it is possible to share the same equipment between users [3] and
geographically distant universities.
Considering the wide range of advantages and possibilities, research on remote
laboratories has grown increasingly in recent years. This trend aims to take advantage
of new resources and technologies to improve the technological education. Many of the
recent publications on remote laboratory are related to various engineering areas, such
as: “electrical and electronics”, “telecommunications” [5], “basic physics” [6], “bio-
medicine” [7], “automation/robotics/classic control” [8, 9], being rare to find something
related to the study of mechanical proprieties of materials, as in [10–13] and even rarer
on plastic deformation, as in [14].
Despite current advances in Information and Communication Technologies (ICTs),
there are few publications about remote laboratories that were developed aiming the
metal forming area and the representation of phenomena linked to the plastic defor-
mation of a metal. It is believed that some factors are responsible for this current
situation, such as: (a) Maintenance of mechanical parts from the remote experiment;
(b) Metal specimens production; (c) Creation of a replacement mechanism for the metal
specimens; (d) Use of higher precision and cost sensors; (e) Use of higher strength and
greater precision actuators; (f) High cost of maintenance. Despite these obstacles, the
challenge of developing a remote laboratory in the metal forming area means a big step
for research in remote experimentation, then generating knowledge and innovation in
the teaching process in engineering courses that are related to this area.
This paper represents a continuation of the outcomes that have been already pub-
lished about the LABCONM (Metal Forming Online Lab) research project. The initial
results were presented in [15, 16]. The LABCONM is an educational laboratory that
provides remote access to one remote educational compression testing machine. This
laboratory was developed specifically to help in the teaching/learning process of metal
flow curves in the metal forming area.
Instead of others publications, this paper describes the LABCONM with more
details and presents the new outcomes based in applications of this lab with Engi-
neering students. This paper aims to describe the LABCONM, that is an educational
laboratory that provides remote access to one remote educational compression testing
machine.
To plan the production of a piece that will be formed, it is necessary to know basic
common aspects of metal forming. Without these aspects, it is not possible to predict or
calculate overall costs, amount of metal necessary, the total energy used in the process,
maximum strength, or total time used in manufacturing process. The flow curve is one
of these fundamentals aspects of metal forming, which in its complexity carries the-
oretical and practical elements, including concepts, equations, symbols and procedures
which are indispensable for the designer.
Flow curves are one of the main ways to analyze the mechanical behavior of a
metal and are critical to design pieces using metal forming. These curves are defined
with data from the plastic zone (above the yield stress) of the stress-strain diagram and
zamfira@unitbv.ro
LABCONM: A Remote Lab for Metal Forming Area 99
they can express the mechanical behavior of metal during the plastic phase. Defor-
mation is a permanent change in the geometric shape of a metal that occurs after
applying stress (r) with intensity above the yield point. The deformation is visible
because it changes the height, length, and the depth of a metal.
Mechanical tests of strength/compression/torsion are made on metal specimens to
obtain data from stress and deformation of a metal. However, when the designer wants
to do calculations and simulations of the metal forming process, the flow curve is not
appropriate, and as a consequence, it is necessary to represent the same data in an
Eq. (1), as it follows:
kf ¼ C:un ð1Þ
3 LABCONM Description
The LABCONM is an educational laboratory that was created to support the theoretical
learning of metal forming with remote experiments. As shown in Fig. 1, LABCONM is
distributed in four main parts connected to the internet, which are:
• Learning Management System (LMS);
• Remote Experiments;
• Web Server;
• Device with internet access for the student;
The LABCONM central portal is a Learning Management System (LMS), which is
actually a web page. Remote experiments are physical experiments that are connected
to the Web Server and are managed by the LMS. It is important to describe that, despite
the “Student Device” is not an item developed for the laboratory, without it the
LABCONM would not be “complete” or “formed”.
In this version of LABCONM was implemented only one remote experiment
(Remote experiment 1), called remote educational compression testing machine
(MDTEC).
zamfira@unitbv.ro
100 L.B. Michels et al.
zamfira@unitbv.ro
LABCONM: A Remote Lab for Metal Forming Area 101
The main features of this equipment are: storage, positioning, measurement, com-
pression and disposal of metal specimens. All these features and processes are auto-
mated and remotely operated through preprogrammed commands in the minicomputer
Raspberry Pi and managed and operated by the virtual control panel in the access page.
In MDTEC, the operating speed vf depends on machine characteristics and the
resistance that the machine meets during the test, being impossible to test with strain
_ constant. The MDTEC performs the compression of the specimen and stores
rate ðuÞ
about 8 readings per second (strength and height variation of the metal specimen).
In MDTEC, there is a series of essential elements for its operation (see Fig. 3),
which can be classified into three mechanisms (Accordingly to its function):
zamfira@unitbv.ro
102 L.B. Michels et al.
zamfira@unitbv.ro
LABCONM: A Remote Lab for Metal Forming Area 103
• Schedule: This menu link serves to define the user’s time in the experimentation
page.
• Use: This menu displays a page with a summary of the student’s situation. It is a
report of the mandatory learning tasks and online exams proposed.
• Activities: On this web page there are several activities with videos, texts and
images related to the MDTEC experiment and the flow curves.
• Experiments: In the experiments menu the users will find submenus to access the
control panel of the experiments. The control panel is where the students
perform/control/monitor the experimentation process. Currently there is only the
panel of MDTEC (see Fig. 5).
• Scores: In this item, questionnaires are used by students to evaluate the laboratory
or to send the requested activities in the Activities Menu;
zamfira@unitbv.ro
104 L.B. Michels et al.
zamfira@unitbv.ro
LABCONM: A Remote Lab for Metal Forming Area 105
4 Methodology
Two different type of analysis were defined to validate the laboratory considering the
teaching approach and the operation of the laboratory. In the technical analysis, the
researchers conducted 20 remotely operated tests, where it was verified quality and
repeatability of the data to demonstrate metal flow. In the pedagogical analysis, two
classes, which were attending Metal Forming course, participated. During four weeks,
LABCONM was available for one of these two groups to use the remote laboratory and
to complete the proposed activities related to the experiment. After this period, both
groups attended a calculus exam related to the course. The group that had the
opportunity to access the remote laboratory also completed a satisfaction questionnaire,
where the questions were about various aspects of the laboratory.
zamfira@unitbv.ro
106 L.B. Michels et al.
(b) How easy was it to use and operate the LABCONM? (c) How was the experi-
mentation process? (d) How objective and understandable were the proposed ques-
tions? (e) How the activities helped to understand the content of the Mechanical
Conformation course? The answers were presented in the Fig. 7.
The mean value between all the answers presented in the Fig. 7 was excellent,
ranging between 8.8 and 9.33. The maximum standard deviation reached between the
five questions was 1.9 in objectivity and understanding of the activities, which
achieved the lowest average as well. This point was expected somehow, because it was
presented to the students a new environment with a lot of information. The idea was
that they should try the whole process of interpretation alone and ask questions via
email, thus stimulating the exchange of information. It is believed that this difficulty
was not an issue for the students, because they did not enter in this merit in the
questionnaire for complaints and suggestions.
The most important question to highlight with respect to the Fig. 7 was the average
score of 9.07 obtained by the item related to “How the activities helped to understand
the content of Metal Forming course”. This score confirms once again the focus of the
remote laboratory for metal forming area to aid in learning process of Metal Forming
theory and it offers good arguments to validate the LABCONM as an educational
laboratory.
zamfira@unitbv.ro
LABCONM: A Remote Lab for Metal Forming Area 107
correctness of the answer, being them: unsatisfactory, satisfactory and excellent. Based
on this criteria, the students were grouped into three subgroups for map the range of
scores, mean values and standard variation, resulting in the data presented in the
Table 1. Through this table, it can be observed that the group that accessed the
LABCONM had an average score of 7.7, which is 7% higher than the average grade for
the group that did not access the remote laboratory. It is important to explain that the
class who did not access the LABCONM is formed by a group of students with better
performance than the group that accessed the LABCOM, as shown in Fig. 8. This
figure shows the result of the groups in the calculus question in the first semester exam.
From this figure, it is possible to conclude that 17% students did not achieve an average
score (above 50%).
65%
56%
28%
17% 17% 17%
However, by the Table 1, where it is possible to verify the results of the second
exam that was applied after the use of the remote laboratory, it is possible to conclude
that none of the students from the group that accessed the LABCONM had an
unsatisfactory performance.
On the other hand, in the group that did not access the LABCONM, 17% of
students in this group had some problem to solve this kind of mathematical question.
zamfira@unitbv.ro
108 L.B. Michels et al.
Therefore, there is an influence of laboratory especially for those students who have
more difficulties in theoretical learning.
Therefore, in a general analysis, it may be affirmed that the class with access to the
laboratory had a larger group of students in the satisfactory and excellent level.
5 Final Considerations
The development of remote laboratory in the metal forming area has some difficulties,
as financial, technology, and maintenance, which are obstacles for the expansion of this
field of study. However, in this article an innovative remote laboratory was described,
which has a Remote Educational Compression Testing Machine capable of performing
real remotely compression tests on metal specimens.
The LABCONM validation showed good quality and repeatability in the MDTEC
results, and as a consequence, it is possible to conclude that the machine can be used
for the generation of flow curves in an educational way. Furthermore, considering the
educational nature of experimental activity proposed, it was verified by the students’
opinion that the laboratory presents a potential to improve learning in the Metal
Forming course. This fact was presented more clearly in the exam scores of the stu-
dents group that had access to LABCONM, because despite of the difficulties that they
have in calculus, none of them obtained an unsatisfactory result. Therefore, based on
these analyzes, it is considered that the LABCONM is a great advance in the field of
remote engineering and virtual instruments and to support the learning in the metal
forming area.
References
1. Fabregas, E., et al.: Development a remote laboratory for engineering education. Comput.
Educ. 57 (2011)
2. Pimentel, A.: A Teoria da Aprendizagem Experiencial como alicerce de estudos sobre
Desenvolvimento Profissional. Estudos de Pscicologia, 159–168 (2007)
3. Ikhlef, A., et al.: Online temperature control system. In: International Conference on Interactive
Mobile Communication Technologies and Learning (IMCL). [S.l.]: [s.n.], pp. 75–78 (2014)
4. Valls, M.G., Val, P.B.: Usage of DDS data-centric middleware for remote monitoring and
control laboratories. IEEE Trans. Indus. Inf. 9(1), 567–574 (2013). Fevereiro
5. Vlasov, I., et al.: Global navigation satellite systems (GNSS) remote laboratory at BMSTU.
In: 2013 2nd Experiment@ International Conference (exp.at 2013), Exp.at. 2013, Coimbra,
pp. 64–67 (2013)
6. Ožvoldová, M., Špiláková, P., Tkac, L.: Archimedes´ principle remote experiment. In: 11th
International Conference on Remote Engineering and Virtual Instrumentation (REV),
Porto-Portugal: [s.n.] (2014)
7. Barros, C., et al.: Remote physiological signals acquisition: didactic experimnets. In: 11th
International Conference on Remote Engineering and Virtual Instrumentation (REV),
Porto-Portugal: [s.n.] (2014)
zamfira@unitbv.ro
LABCONM: A Remote Lab for Metal Forming Area 109
8. Ghorbel, H., et al.: Remote laboratory for control process practical course in eScience
project. In: International Conference on Interactive Mobile Communication Technologies
and Learning (IMCL), Thessaloniki, Greece: [s.n.] (2014)
9. Ayodele, K.P., Inyang, I.A., Kehinde, L.O.: An iLab for teaching advanced logic concepts
with hardware descriptive languages. IEEE Trans. Educ. 58(4), 262–268 (2015)
10. Restivo, M.T., et al.: A Remote Laboratory in Engineering Measurement, vol. 56, no. 12,
pp. 4836–4843, December 2009
11. Marcelino, R., et al.: Extended immersive learning environment: a hybrid remote/virtual
laboratory. Int. J. Online Eng. (IJOE) 6, 46–51 (2010)
12. Michels, L.B., et al.: Using remote experimentation for study on engineering concepts
through a Didactic press. In: 2nd Experiment@ International Conference - Exp’at, Coimbra:
[s.n.], pp. 191–193 (2013)
13. Nasri, I., Ennetta, R.: Determination of resonance frequency and estimation of damping ratio
for forced Vibrations modules using remote lab. In: International Conference on Interactive
Mobile Communication Technologies and Learning (IMCL), Thessaloniki, Greece: [s.n.]
(2014)
14. Terkowsky, C., et al.: Developing tele-operated laboratories for manucfacturing engineering
education. In: International Conference on Remote Engineering and Virtual Instrumentation,
REV2010, Stockholm, pp. 60–70 (2010)
15. Michels, L.B., et al.: Educational compression testing machine for teleoperated teaching of
the metals flow curves. In: Exp’at 2015, Ponta Delgada: [s.n.] (2015)
16. Michels, L.B., et al.: Remote compression test machine for experimental teaching of
mechanical forming. Int. J. Online Eng. 12(04), 20–22 (2016)
zamfira@unitbv.ro
A Virtual Proctor with Biometric Authentication
for Facilitating Distance Education
Abstract. The lack of efficient and reliable proctoring for tests, examinations
and laboratory exercises is slowing down the adoption of distance education. At
present, the most popular solution is to arrange for proctors to supervise the
students through a surveillance camera system. This method exhibits two short‐
comings. The cost for setting up the surveillance system is high and the proctoring
process is laborious and tedious. In order to overcome these shortcomings, some
proctoring software that identifies and monitors student behavior during educa‐
tional activities has been developed. However, these software solutions exhibit
certain limitations: (i) They impose more severe restrictions on the students than
a human proctor would. The students have to sit upright and remain directly in
front of their webcams at all times. (ii) The reliability of these software systems
highly depends on the initial conditions under which the educational activity is
started. For example, changes in the lighting conditions can cause erroneous
results.
In order to improve the usability and to overcome the shortcomings of the
existing remote proctoring methods, a virtual proctor (VP) with biometric authen‐
tication and facial tracking functionality is proposed here. In this paper, a two-
stage approach (facial detection and facial recognition) for designing the VP is
introduced. Then, an innovative method to crop out the face region from images
based on facial detection is presented. After that, in order to render the usage of
the VP more comfortable to the students, in addition to an eigenface-based facial
recognition algorithm, a modified facial recognition method based on a real-time
stereo matching algorithm is employed to track the students’ movements. Then,
the VP identifies suspicious student behaviors that may represent cheating
attempts. By employing a combination of eigenface-based facial recognition and
real-time stereo matching, the students can move forward, backward, left, right
and can rotate their head in a larger range. In addition, the modified algorithm
used here is reliable to changes of lighting, thus decreasing the possibility of false
identification of suspicious behaviors.
zamfira@unitbv.ro
A Virtual Proctor with Biometric Authentication 111
1 Introduction
The distance education market keeps growing rapidly [1]. While several research threads
(i.e. on real-time creation of virtual environments for virtual laboratories [2, 3], augmen‐
tation of virtual laboratories [4], creation of smart sensor networks [5], etc.) have
contributed to the continued adoption of distance education approaches, the lack of
efficient and reliable proctoring is slowing this adoption process down.
At present, the most popular solutions in distance education for monitoring an
experiment or an examination are human proctors. Human proctors used in distance
education can be teaching assistants, instructors, laboratory administrators and faculty
members. There are also companies that provide the service of monitoring examinations
from a distance (e.g. ProctorU [6]). In most remote proctoring cases, the students take
examinations and perform experiments on a computer and the proctor(s) watch(es) them
from another computer through video cameras. The human proctors must monitor a
screen throughout the entire process. The basic requirement for operating a remote
proctor is that it needs a remote surveillance camera system mounted at the student’s
site. The advantage of this method is that it is similar to traditional classroom education
and therefore provides fewer challenges than technology-assisted methods such as VPs.
However, this method also has two shortcomings. One disadvantage is that the opera‐
tional costs are high since such proctoring services currently usually charge over $60
per student per course while the cost for setting up the surveillance system is high.
Another disadvantage is that the proctoring process is laborious and tedious.
With the further development of computer vision technology, VPs appeared. VPs
are integrated software-hardware solutions that have the potential to contribute to
bringing academic integrity to distance education. They were enabled by the prolifera‐
tion of the high-speed Internet and advanced computer peripherals. They first perform
the authentication of the students by scanning either their faces [7] or their fingerprints
[8]. Then, a camera monitors the environment and/or a microphone records the sounds
within it. Virtual proctoring software used in distance education includes Remote
Proctor Pro [9], Instant-InTM Proctor [10], Proctortrack [7], Proctorfree [9] and
Securexam Remote Proctor [11]. Virtual proctoring has three advantages over human
proctors. First is its low fixed cost. The students only need to set up a webcam and install
the VP software which then performs the authentication of the student and the proctoring
of the educational activity. Typically, the cost of the VP (including webcam, microphone
and software kit) will not exceed $15 per student. The second advantage of virtual proc‐
toring lies in its convenience. There is no need for human proctors, and thus the educa‐
tional activity to be proctored can take place at anytime and anywhere. The third
advantage is in the accurate authentication of the students. The utilization of biometric
technologies enables the accurate recognition of the students, thus ensuring a reliable
authentication [10]. It should be noted that virtual proctoring is still evolving and current
systems are often attracting complaints from the students, mostly because of two short‐
comings exhibited by these systems. First, they impose more severe restrictions on the
students than a human proctor would. The students have to sit upright and remain directly
in front of their webcams at all times. Second, the reliability of the VP highly depends
on the initial conditions under which the proctoring is started. For example, changes in
zamfira@unitbv.ro
112 Z. Zhang et al.
the lighting conditions can cause mistakes in the verification of suspicious behav‐
iors [12].
In order to improve the usability and to overcome the shortcomings of the existing
remote proctoring methods, a VP with biometric authentication and facial tracking
functionality is proposed here. This VP is designed to authenticate the students and
capture suspicious behaviors based on facial recognition and facial tracking. The work‐
flow of this VP is depicted in Fig. 1 and is composed of two main parts: authentication
and supervision. In addition, there is a database which stores the enrolled students’ face
templates indexed with their campus ID. When they use this VP, the students are first
required to scan their face using a webcam. Second, the scanned frame is processed, and
the part of the frame containing the face is cropped out. Third, the face is then compared
with the face template that was stored in the face database and could be retrieved by the
index of student’s campus ID. If the face and the template match, the student is authen‐
ticated and can continue the educational activity. Otherwise, the student is logged out.
After authentication, the frame used for authentication is stored in a newly allocated
memory address and is taken as the new template. Then, the subsequent matching is
based on this new template instead of the template stored in the face database. Then, the
educational activities are monitored by the webcam. During the monitoring period, the
VP samples the live video of the student with a sample rate of 30 frames per second. If
the mismatching percentage between the sampled face and the new template exceeds a
pre-configured threshold value, a suspicious behavior is identified and a video clip is
recorded, which is then used for further verification by the instructor of the examinations
or experiments.
zamfira@unitbv.ro
A Virtual Proctor with Biometric Authentication 113
was divided into two stages. The first stage is the detection of the face in a sampled
frame of the student’s live video. Once a face has been detected, the area of the detected
face is cropped out from that frame and used for the following facial recognition. The
second stage is the recognition of the detected face. The cropped face area is compared
with the face template as illustrated in Fig. 1. Because the cropped face area is much
smaller than the whole image area, the computational cost of recognizing the face is
reduced considerably.
Below, the reason why facial recognition was selected instead of other biometric
methods is explained first. Following that, various facial detection algorithms are
discussed, including the algorithm selected here and the method employed to crop out
the face area. Subsequently, a modified facial recognition algorithm based on stereo
matching is introduced that overcomes some of the shortcomings of other facial recog‐
nition algorithms based on template matching or eigenfaces. Finally, the results of some
benchmarks are presented to confirm that the proposed facial recognition algorithm is
reliable.
zamfira@unitbv.ro
114 Z. Zhang et al.
zamfira@unitbv.ro
A Virtual Proctor with Biometric Authentication 115
The process described above only finds an estimated area of the faces. It renders the
recognition process difficult since the information provided by the face area is insuffi‐
cient for the subsequent facial recognition. In fact, the estimated area of the faces results
in the loss of the entire background and part of the outline of the face. In order to
compensate for the loss of information and to facilitate the following facial recognition
process, a modification for the facial detection algorithms based on the localization of
the mouth and eyes was implemented. First, the coordinates of the eyes are set as
E1(x1, y1) and E2(x2, y2), and the coordinates of the mouth are set as M(x3, y3). All
coordinates represent the center of the area of the eyes and mouth. The cross product of
vectors E⃖⃖⃖⃖⃖⃖⃖
⃗ ⃖⃖⃖⃖⃖⃖⃖⃗
1 E2 and E1 M is positive if M is above the line E1E2. Based on the golden ratio
of the human face [30], the face area forms a golden rectangle with the eyes at its
midpoint. Then, the vertical ratio equals the distance between the pupils and the mouth
in relation to the distance of the hairline to the chin, i.e. E1 M/H C = 0.36. The horizontal
ratio equals the distance between the pupils in relation to the width of the face, i.e.
E1 E2/L R = 0.46 (see Fig. 3 [30, 31]). The obtained rectangle should be L R × H C, but
the rectangle actually used to facilitate the face recognition is increased by 10 pixels in
each direction in order to avoid loss of the facial information. The part of the code based
on the cascade of OpenCV is listed in Fig. 4.
zamfira@unitbv.ro
116 Z. Zhang et al.
where R is the resulting error, T is the template, I is the target image, (x, y) are the
coordinates of the image in pixels, and (x′, y′) are the coordinates of the template in
pixels.
Different error functions can be specified depending on the prevailing conditions.
After comparison between the template and the target image, the best matches can be
found as global minima or maxima [36].
The eigenface method is an efficient approach for recognizing a face [37]. A high
recognition rate can be achieved with a low dimension d of the eigenvector space since
the recognition rates are stable when the dimension of the eigenvector space equals 8.
zamfira@unitbv.ro
A Virtual Proctor with Biometric Authentication 117
In order to identify the eigenvectors, a principal component analysis was used to find
the directions with the greatest variances of the components of a given dataset. These
variances are called principal components (and are also the eigenvalues associated with
the eigenvectors used in the eigenfaces). Then, a high-dimensional dataset is described
by such a series of correlated variables. The algorithm can be described as follows.
Let X = {x1 , x2 , … , xn }, xi ∈ Rd be a random vector wherein xi are the observations.
The expected value μ of the observations is:
1∑
n
𝜇= x (2)
n i=1 i
1∑
n
S= (x − 𝜇)(xi − 𝜇)T (3)
n i=1 i
Svi = 𝜆i vi , i = 1, 2, … , n (4)
If there are k principal components and the corresponding eigenvectors are labelled
in descending order based on the values of the principal components, then the k principal
components of the observed vector x are given by:
The reconstruction of the eigenvectors from the principal component analysis (PCA)
is given by:
x = Wy + 𝜇 (6)
Following the procedures outlined above, three more steps realize facial recognition.
In the first step, all training samples are projected into the PCA subspace composed of
eigenvectors. In the second step, the query image (i.e. the target image that will be
identified) is projected into the PCA subspace. In the last step, the nearest neighbor
between the projected training images (i.e. the former training samples) and the projected
query image is determined.
Template matching is reliable and simple when the contexts of the images are
constrained. The eigenface method is designed based on a generalization of the faces,
and therefore, it is robust and accurate under conditions that vary only mildly.
Unfortunately, both the template matching method and the eigenface method are
vulnerable to changes in the environment and pose variations of the face. Therefore, the
method of facial recognition employed here is a modification of stereo matching based
on the facial detection results.
zamfira@unitbv.ro
118 Z. Zhang et al.
d(x, y) = ‖ ′ ‖
‖I(x, y) − I (x, y)‖ (7)
Then, the sum of the absolute intensity differences (SAD) of the intensity in the
template and captured image is:
∑
W
∑
W
SAD(d) = ‖I(x + i, y + j) − I ′ (x + i + d, y + j)‖
‖ ‖ (8)
y=−W x=−W
If the SAD is used directly to obtain the disparity maps (which refer to the apparent
pixel difference or motion between a pair of stereo images), the noise in the disparity
maps is very large since the signal-to-noise ratio is too low. In order to optimize the
stereo matching accuracy, a box filter based on the cross-correlation in the window areas
(of size 2W × 2W) around the landmarks is used:
∑
W
∑
W
SC (x, y, d) = [I(x + i, y + j) − I ′ (x + i + d, y + j)] (9)
j=−W i=−W
zamfira@unitbv.ro
A Virtual Proctor with Biometric Authentication 119
∑ ∑
W+1
W
SC (x, y + 1, d) = [I(x + i, y + 1 + j) − I ′ (x + i + d, y + 1 + j)] (10)
j=−W−1 i=−W
∑
W
[ ]
AC (x, y, d, j) = (I(x + i, y + 1 + j) − I ′ (x + i, y + 1 + j − d) (11)
i=−W
Here, d is the possible floating of the actual location of the captured image (i.e. the
lag in the definition of the cross-correlation).
In Fig. 5, the comparison of the disparity maps obtained with SAD and SAD with
window filter is depicted. It can be seen that the noise can be reduced efficiently with
the window filter.
After obtaining the disparity between the template image and the captured image,
the facial recognition of the users can be performed.
In order to test the proposed method used to recognize the students, a benchmark
analysis was conducted which compares the proposed method with the template
matching and eigenface methods. In this benchmark, “the Sheffield (previously UMIST)
face database” [38, 39] was used to test the performance of different methods with
respect to the pose translation. 20 sample sets from this database were tested from
different view points (0 to 90 in 10 intervals). The frontal view was defined as the 0
viewpoint, and it was taken as the template. The test results are shown in Fig. 6. It is
zamfira@unitbv.ro
120 Z. Zhang et al.
seen that the stereo matching method provides for a higher reliability than the other two
methods when the pose translation is large.
Fig. 6. Template matching, eigenfaces and stereo matching about pose translation
In order to test the reliability of the proposed method under variable illumination
conditions, a benchmark analysis with the “extended Yale face database B” [40, 41] was
conducted. The frontal faces of 15 individuals under 50 different illumination conditions
were examined. The results of the test are illustrated in Fig. 7. The eigenface method is
better than the template matching method, the reliability of which is critically affected
Fig. 7. Template matching, eigenfaces and stereo matching under various illumination conditions
zamfira@unitbv.ro
A Virtual Proctor with Biometric Authentication 121
by the illumination conditions. In contrast, the stereo matching method ensures a high
reliability even under poor illumination conditions.
As mentioned above, the final implementation of the stereo matching method focuses
on three areas: two eye-centered areas and one mouth-centered area. Decreasing the
sizes of the areas used for stereo matching does not only increase the efficiency of the
stereo matching algorithm but also improves the reliability of the facial recognition. In
addition, ‘C++ accelerated massive parallelism’ [42] was used in the implementation
of the stereo matching algorithm, and therefore the speed of the execution of the program
was increased significantly.
It is difficult to define all possible suspicious behaviors that may represent cheating
attempts. Therefore, a small set of such behaviors was defined in the pilot implementa‐
tion of the proposed VP.
In order to understand this definition, a coordinate system was chosen as depicted in
Fig. 8. First, rotations of the head about the Z axis are considered normal. Suspicious
behaviors of “rotating head” correspond to rotations about either the X or Y axes.
Second, “moving relative to webcam” corresponding to a translation along the X, Y or
Z directions is also suspicious. The suspicious behavior of “rotating head” is judged by
the matching percentage between the captured frame and the template stored in the face
database. The essence of this method is that the face in frontal view is taken as the facial
recognition object. Thus, rotations about the X and Y axes generate obvious differences
between the captured frame and the template. In addition, these differences cannot be
eliminated by face alignment. Therefore, the rotation angle can be estimated according
to the face matching percentage. The calculation of translations is much easier than that
of rotations. The method used here is to track the location of the face and the size of the
face area. Then, the relative location of the face can be determined.
zamfira@unitbv.ro
122 Z. Zhang et al.
In order to evaluate the performance of the VP, 50 intentional cheating attempts per
criterion were tested. The test results corresponding to different criteria are listed in
Table 1. These results show that the proposed VP is reliable with respect to pose trans‐
lations and illumination changes.
For further assessment of the proposed algorithms used in the VP, a pilot test with
three volunteers was conducted. The details can be found elsewhere [43]. The difference
between this pilot test and the prior experiment involving 50 cheating attempts is that
the illumination conditions were changed while the volunteers were asked to bow their
head. Despite the changing conditions, the VP based on the proposed algorithms worked
well with respect to recognizing and tracking the users.
In this paper, a VP based on facial recognition was introduced. In order to overcome the
shortcomings of existing facial recognition methods, a stereo matching method was
proposed to improve the reliability of the facial recognition. In order to evaluate the
reliability of this method, two benchmark analyses were conducted. The first analysis
was designed to determine the impact of the stereo matching method on the reliability
for large pose translations. The second analysis was aiming to test the performance of
the stereo matching method under variable illumination conditions. The results proved
that the proposed method has a higher reliability than the existing facial recognition
methods (template matching and eigenfaces).
In the future, the performance of the VP will be enhanced by adding voice identifi‐
cation and recognition functions, adding screen monitoring functionality, targeting more
complicated suspicious behaviors and optimizing the recognition algorithms.
Although the proposed VP still has certain limitations, it performed well under labo‐
ratory conditions. In addition, it has the potential to replace human proctors in both
distance education and traditional classroom settings.
References
1. http://wcet.wiche.edu/initiatives/research/WCET-Distance-Education-Enrollment-Report-2016.
Accessed Nov 2016
2. Zhang, Z., Zhang, M., Chang, Y., Esche, S.K., Chassapis, C.: A smart method for developing
game-based virtual laboratories. In: Proceedings of the ASME International Mechanical
Engineering Conference and Exposition, IMECE 2015, Houston, Texas, 13–19 November
2015
zamfira@unitbv.ro
A Virtual Proctor with Biometric Authentication 123
3. Zhang, Z., Zhang, M., Chang, Y., Esche, S.K., Chassapis, C.: Real-time 3D reconstruction
for facilitating the development of game-based virtual laboratories. Comput. Educ. J. 7(1),
85–99 (2016)
4. Zhang, Z., Zhang, M., Tumkor, S., Chang, Y., Esche, S.K., Chassapis, C.: Integration of
physical devices into game-based virtual reality. Int. J. Online Eng. 9, 25–38 (2013)
5. Qureshi, F., Terzopoulos, D.: Smart camera networks in virtual reality. In: Proceedings of
First ACM/IEEE International Conference on Distributed Smart Cameras, Vienna, Austria,
25–28 September 2007
6. http://www.proctoru.com/. Accessed Oct 2016
7. http://www.proctortrack.com/. Accessed Oct 2016
8. http://www.softwaresecure.com/remote-proctor-pro-faq/. Accessed Oct 2016
9. http://proctorfree.com/. Accessed Oct 2016
10. http://www.biomids.com/proctoring/. Accessed Oct 2016
11. http://remoteproctor.com/rpinstall/orgselector/orgselector.aspx. Accessed Oct 2016
12. http://www.nytimes.com/2015/04/06/technology/online-test-takers-feel-anti-cheating-
softwares-uneasy-glare.html. Accessed Sept 2016
13. Rasmussen, K.B., Roeschlin, M., Martinovic, I., Tsudik, G.: Authentication using pulse-
response biometrics. In: Proceedings of Network and Distributed System Security
Symposium 2014, San Diego, California, USA, 23–25 February 2014
14. Bača, M., Grd, P., Fotak, T.: Basic principles and trends in hand geometry and hand shape
biometrics. In: New Trends and Developments in Biometrics. INTECH Open Access
Publisher (2012)
15. http://proctorfree.com/blog/the-future-is-now-iris-recognition-technology-makes-id-cards-
redundant. Accessed Oct 2016
16. Proctor, R.W., Lien, M.C., Salvendy, G., Schultz, E.E.: A task analysis of usability in third-
party authentication. Inf. Secur. Bull. 5(3), 49–56 (2000)
17. https://facedetection.com/algorithms/. Accessed Oct 2016
18. Horprasert, T., Harwood, D., Davis, L.S.: A robust background subtraction and shadow
detection. In: Proceedings of 4th Asian Conference on Computer Vision, Taipei, Taiwan, 5–
8 January 2000
19. http://www.cs.rutgers.edu/~decarlo/facetrack.html. Accessed Oct 2016
20. Fröba, B., Külbeck, C.: Real-time face detection using edge-orientation matching. In:
Proceedings of International Conference on Audio- and Video-Based Biometric Person
Authentication, Halmstad, Sweden, 6–8 June 2001
21. Jesorsky, O., Kirchberg, K.J., Frischholz, R.W.: Robust face detection using the Hausdorff
distance. In: Proceedings of International Conference on Audio- and Video-Based Biometric
Person Authentication, Halmstad, Sweden, 6–8 June 2001
22. Vasconcelos, N., Saberian, M.J.: Boosting classifier cascades. In: Proceedings of Advances
in Neural Information Processing Systems 23, Vancouver, British Columbia, Canada, 6–9
December 2010
23. Viola, P., Jones, M.: Fast and robust classification using asymmetric adaboost and a detector
cascade. In: Proceedings of Advances in Neural Information Processing System 14,
Vancouver, British Columbia, Canada, 3–8 December 2001
24. Gokturk, S.B., Bouguet, J.Y., Tomasi, C., Girod, B.: Model-based face tracking for view-
independent facial expression recognition. In: Proceedings of Fifth IEEE International
Conference on Automatic Face and Gesture Recognition, Washington D.C., USA, 20–21 May
2002
25. Viola, P., Jones, M.: Robust real-time object detection. Int. J. Comput. Vis. 57(2), 137–154
(2004)
zamfira@unitbv.ro
124 Z. Zhang et al.
26. Wilson, P.I., Fernandez, J.: Facial feature detection using Haar classifiers. J. Comput. Sci.
Coll. 21(4), 127–133 (2006)
27. http://docs.opencv.org/master/d7/d8b/tutorial_py_face_detection.html. Accessed Nov 2016
28. http://opencv.org/. Accessed Nov 2016
29. Zhang, Z., Zhang, M., Chang, Y., Esche, S.K., Chassapis, C.: A virtual laboratory system
with biometric authentication and remote proctoring based on facial recognition. In:
Proceedings of the 2016 ASEE Annual Conference and Exposition, New Orleans, LA, USA,
26–29 June 2016
30. http://www.goldennumber.net/face/. Accessed Nov 2016
31. http://www.goldennumber.net/facial-beauty-new-golden-ratio/. Accessed Nov 2016
32. Brunelli, R., Poggio, T.: Face recognition: features versus templates. IEEE Trans. Pattern
Anal. Mach. Intell. 15, 1042–1052 (1993)
33. Turk, M., Pentland, A.: Eigenfaces for recognition. J. Cogn. Neurosci. 3, 71–86 (1991)
34. Jafri, R., Arabnia, H.: A survey of face recognition techniques. J. Inf. Proces. Syst. 5(2), 41–
68 (2009)
35. Pentland, A., Moghaddam, B., Starner, T.: View-based and modular eigenspaces for face
recognition. In: Proceedings of IEEE Computer Society Conference on Computer Vision and
Pattern Recognition, Seattle, WA, 21–23 June 1994
36. http://docs.opencv.org/2.4/modules/imgproc/doc/object_detection.html. Accessed Nov 2016
37. Menezes, P., Barreto, J.C., Dias, J.: Face tracking based on Haar-like features and eigenfaces.
In: Proceedings of IFAC/EURON Symposium on Intelligent Autonomous Vehicles, Técnico,
Lisboa, Portugal, 5–7 July 2004
38. https://www.sheffield.ac.uk/eee/research/iel/research/face. Accessed Nov 2016
39. Graham, D.B., Allinson, N.M.: Characterising virtual eigensignatures for general purpose
face recognition. face recognition, pp. 446–456. Springer, Heidelberg (1998). doi:
10.1007/978-3-642-72201-1_25
40. http://vision.ucsd.edu/~leekc/ExtYaleDatabase/ExtYaleB.html. Accessed Nov 2016
41. Georghiades, A.S., Belhumeur, P.N., Kriegman, D.J.: From few to many: illumination cone
models for face recognition under variable lighting and pose. IEEE Trans. Pattern Anal. Mach.
Intell. 23(6), 643–660 (2001)
42. https://msdn.microsoft.com/en-us/library/hh265136.aspx. Accessed Nov 2016
43. Zhang, Z., Zhang, M., Chang, Y., Esche, S.K., Chassapis, C.: A virtual laboratory combined
with biometric authentication and 3D reconstruction. In: Proceedings of the ASME
International Mechanical Engineering Conference and Exposition, IMECE 2016, Phoenix,
Arizona, USA, 11–17 November 2016
zamfira@unitbv.ro
From a Hands-on Chemistry Lab to a Remote Chemistry
Lab: Challenges and Constrains
Abstract. The spread of remote labs in Universities is a current reality. They are
strong e-learning tool which allow students to carry out online experiments over
real equipment and Universities to have e-learning tools for learning methodol‐
ogies such as Blended learning and Distance learning. These remote labs are
developed for many science fields such as electronic, robotic and physic. Never‐
theless it is very difficult to find chemistry remote labs. This paper wants to show
the difficulties of choosing a chemistry lab which can become a remote chemistry
lab, and a first approach of converting a hands-on chemistry lab to remote one.
1 Introduction
zamfira@unitbv.ro
126 S. Elio et al.
These remote labs also cover a great number of science fields (Fig. 2), such as:
• Robotic labs. For instance i-robot lab from The Labshare Institute in Australia. This
remote lab was designed to allow students to explore the concepts of teleoperation
of robots, accuracy of sensors, localization and mapping [4]. Or the Robotic arm
From UNED which allow students to work with a real robotic arm [5].
• Physic remote labs. For instance Archimedes remote labs from Deusto University
where students of secondary school learn the Archimedes’ Principle [6].
• Electronic remote labs. There are a lot of remote labs in this field. But, only two of
them are going to be described:
– The first one, is VISIR, this labs is really interesting for several reason: more than
one user can access it at the same time, several universities has implemented this
labs and are working together in project such as VISIR+ [7] and PILAR [8]. VISIR
allows wiring and measuring of electronic circuits remotely on a virtual work‐
bench that replicates physical circuit breadboards.
– The second one is the Microelectronics Device Characterization remote lab. This
Measures the DC current-voltage characteristics of microelectronics devices such
as diodes and transistors [9].
This section showed several examples of virtual and remote labs in different science
fields. The next sections is going to focus on describe briefly the architecture of remote
labs and the difficulties to create chemistry remote labs due to its nature.
zamfira@unitbv.ro
From a Hands-on Chemistry Lab to a Remote Chemistry Lab 127
The vast majority of remote labs are based on a same architecture. This is composed by:
• Web Server contains services such as control of users, calendar and user interfaces.
It also communicates with user and the lab server.
• Lab server contains the program to act with the real equipment and send the result
to web server.
• Real equipment depends on the remote labs. In the above section, it was shown real
instrumentation such as robotic arm, electronic circuits, motors and pipettes.
• Web cam allows students to see the results of acting with the real equipment.
Depending on web cam students can zoom in or out real instrumentation.
This is the hardware architecture but there are global phases that a remote lab should
fulfill (Fig. 3). These Phases are:
• Initial state. Students must find the instrumentation in a state initial. For instance, in
the Archimedes lab of Fig. 2 the balls must be out of the water or in VISIR lab the
entries of the circuits must be in its initial state.
• Experimentation. This phase can be divides into other, such as action over the labs,
storing student’s actions over the equipment, storing results of these actions, etc.
• Results. Lab should show a report of results of the experiments
• Visualization. All what happens during the experimentation process must be watched
by students through web cams, user interfaces, etc.
zamfira@unitbv.ro
128 S. Elio et al.
Chemistry labs work with liquids, solids and gases. These resources are combined to
create new ones. These experiments need a set of requirements that are really difficult
in by phases of a remote lab.
• State initial:
– Many chemistry laboratories works with fluids. These fluids are mixed and some‐
times evaporated, therefore when students finish their experiments the fluids must
be replaced and the instrumentation must be cleaned.
– Many chemistry labs works with solids and liquids. These can vary his weight or
volume. These can also mix giving as a result other chemical compound. There‐
fore, it is really difficult to give back an initial state without human help.
• Experimentation
– Chemistry labs need handle and weigh solid material. For instance, the experiment
of Reaction of zinc with iodine needs Zinc powder, about 0.5 g, sulfuric acid,
about 20 cm3, etc. Implementing the mechanics to do these measurements in an
automatic way is really complicated.
• Visualization
– Chemistry labs which work with gasses and transparent liquids are difficult to
watch with web cams.
– In some Chemistry labs, odors are also important for students. Remote labs are
not able to provide this sense, although it is possible to use gas sensors.
zamfira@unitbv.ro
From a Hands-on Chemistry Lab to a Remote Chemistry Lab 129
All these reasons show the difficulties of designing and developing a chemistry
remote laboratory.
Once all these constrains were keeping in mind, the department of Electrical and
Computer Engineering Department and Chemistry applied to engineering department
from UNED decide to work in staring the conversion of the hands-on Hydrogen-solar
equipment.
This equipment allows students to carry out hydrogen-solar energy cycle experi‐
ments. To do this, equipment provides a set of elements to convert water to hydrogen
and oxygen, to store these in graduated Cylinders and to consume them in a fuel cell
and produce electrical energy and water. This energy can be used to switch on a bulb or
start a motor.
As it has been told, the equipment is a set of hardware elements which allow
performing this chemical process (Fig. 4). Among them:
• Light source. Sun is replaced by a lamp. This lamp simulates renewable energy.
Students and teacher can move closer and farther the lamp to solar panel. This allows
simulating the variation of light radiation on solar panel.
• Solar panel converts the light luminous energy, which is supplied by the lamp, into
electrical energy. Students and teacher can vary the solar panel orientation and simu‐
late different inclinations.
• Electrolyzer decomposes water into hydrogen and oxygen by using the electrical
energy supplied by the solar panel.
zamfira@unitbv.ro
130 S. Elio et al.
• Fuel cell. It consists of two PEM fuel cells that can be connected in series or in
parallel. They are used to generate electricity from the hydrogen and oxygen
produced by the electrolyzer.
• Load module. It consists of an engine, a lamp and a set of resistors that allow using
the electric energy generated by the fuel cell.
• Measuring devices. It is composed by a voltmeter and an ammeter to visualize the
different voltages and intensities of the electric energy produced and consumed in
each of the processes.
Although this lab requires to be filled with water for the initial state, the rest of the
experimentation can be automated for blended and distance learning.
This hand-on lab can be converted to remote labs. In this first step the department of
electrical and computer of UNED has been focused on load module which can be
replaced by an IoT device, such as Arduino and/or raspberry pi. These devices can
manage a dimmer which can control the intensity of a lamp (Fig. 5).
Arduino and raspberry pi allow remote labs programmers to create a web page where
students can change the intensity of the lamp.
Along with the modification of Load module a web cam connected directly to
Ethernet will allow students to watch the real instrumentation and chemical process.
5 Conclusion
This paper shows the difficulties of creating chemistry labs. To do this, papers describes:
• A state of art of virtual and remote labs and some of the science field where are
applied.
• Global architecture of remote labs and the phases of a remote lab.
• Constrains that have to be considered if someone wants to develop a chemistry remote
lab.
zamfira@unitbv.ro
From a Hands-on Chemistry Lab to a Remote Chemistry Lab 131
• The selection of a chemistry lab that can minimize these constrains and will become
a remote lab.
• And finally, the initial steps of the department of Electrical and Computer Engi‐
neering Department and Chemistry applied to engineering department from UNED
to create a chemical remote lab.
Although a long road lies ahead, the first steps have been done.
Acknowledgement. The authors acknowledge the support of the eMadrid project (Investigación
y desarrollo de tecnologías educativas en la Comunidad de Madrid) - S2013/ICE-2715, VISIR+
project (Educational Modules for Electric and Electronic Circuits Theory and Practice following
an Enquiry-based Teaching and Learning Methodology supported by VISIR) Erasmus+ Capacity
Building in Higher Education 2015 nº 561735-EPP-1-2015-1-PT-EPPKA2-CBHE-JP and PILAR
project (Platform Integration of Laboratories based on the Architecture of visiR), Erasmus+
Strategic Partnership nº 2016-1-ES01-KA203-025327.
References
1. García-Zubia, J., Orduña, P., López-de-Ipiña, D., Alves, G.R.: Addressing software impact in
the design of remote laboratories. IEEE Trans. Industr. Electron. 56(12), 4757–4767 (2009)
2. Gomes, L., Bogosyan, S.: Current trends in remote laboratories. IEEE Trans. Industr. Electron.
56(12), 4744–4756 (2009)
3. Tawfik, M., Sancristobal, E., Martin, S., Diaz, G., Peire, J., Castro, M.: Expanding the
boundaries of the classroom: implementation of remote laboratories for industrial electronics
disciplines. Ind. Electron. Mag. 7(1), 41–49 (2013). IEEE
4. Labshare Labs. http://www.labshare.edu.au/catalogue/rigtypedetail/?id=42&version=1.3.
Accessed 9 Nov 2016
5. Carro, G., Plaza, P., Sancristobal, E., Castro, M.: A wireless robotic educational platform
approach. In: 13th International Conference on Remote Engineering and Virtual
Instrumentation (REV) (2016)
6. Garcia-Zubia, J., et al.: Archimedes remote lab for secondary schools. In: 3rd Experiment@
International Conference, exp.at 2015 (2015)
7. VISIR+ Project: http://www2.isep.ipp.pt/visir/. Accessed 16 Nov 2016
8. PILAR Project. http://ec.europa.eu/programmes/erasmus-plus/projects/eplus-project-details-
page/?nodeRef=workspace://SpacesStore/2d88ecb1-3db1-4a29-93c1-dd2802eec4f6.
Accessed 16 Nov 2016
9. Microelectronics Device Characterization Lab (MIT). http://ceci.mit.edu/projects/iLabs/
Accessed 16 Nov 2016
zamfira@unitbv.ro
Advanced Intrusion Prevention for Geographically
Dispersed Higher Education Cloud Networks
1
Marist College, Poughkeepsie, NY, USA
{casimer.decusatis,Piradon.Liengtiraphan1}@marist.edu
2
BlackRidge Technologies, Reno, NV, USA
tsager@blackridge.us
1 Introduction
Higher education institutions in the U.S. are expected to spend about $10.8 billion on
information technology (IT) in 2016 (up from $6.6 billion last year), primarily driven
by investments in enterprise networks [1]. Globally, the higher education market is
expected to spend over $38.2 billion on IT in 2016 alone [2]. According to EduCause,
a nonprofit organization of IT leaders from higher education [3], the leading issue driving
upgrades for these organizations is information security. Security concerns among
higher education institutions appear to be well justified; the environment in which higher
education institutions operate, and the data which they store, has made them prime
targets for cyberattacks. Recent survey data indicates that 35% of all security breaches
take place in higher education [3]. Among those institutions suffering a breach, over
46% verified advanced persistent threat (APT) activity taking place in their environment
[4]. Higher education institutions collect and retain valuable data such as student,
alumni, and faculty personally identifiable information (PII) including medical records;
research data which may be subject to export control regulations; financial and
accounting data including student tuition, loans, and institution accounting records; and
critical infrastructure or intellectual property information including analytic systems
used for grading and research. This type of information is subject to various local,
national, and international security and privacy compliance regulations, including the
NIST 800 series of security guidelines [5]. In some ways, higher education can be
considered a large enterprise; despite this, higher education is not currently classified as
zamfira@unitbv.ro
Advanced Intrusion Prevention 133
a “mission critical” application by the U.S. federal government [5]. In fact, many large
enterprises employ security policies based on the principle “exclude everything, allow
specific”, while the nature of higher education is just the opposite, and often implements
policies such as “allow everything, exclude specific” in an attempt to promote shared
academic research and education. This can make it particularly challenging to develop
effective security policies for higher education institutions.
A recent example involves the Linux One Community Cloud, a collaboration
between industry and academia to provide free access to an open source Linux devel‐
opment environment [6]. In August 2015, IBM announced a series of enterprise-class
servers which run only the Linux operating system. The Linux One platforms currently
support SUSE, Red Hat, and Ubuntu Linux distributions, along with a variety of
supporting tools such as Apache Spark, MongoDB, and Chef. In order to promote
development and research on this platform, the Linux One Community Cloud makes it
possible for anyone to request a free instance of the Linux One servers and toolsets. This
environment is hosted at the New York State Center for Cloud Computing and Analytics
(CCAC) at Marist College (a private, 6,000 student institution in upstate New York),
and is managed from a IBM development location in Poughkeepsie, NY. However, this
open innovation initiative also means that the cloud hosting Linux One is subject to
continuous cyberattacks from bad actors who attempt to exploit the open access privi‐
leges in this environment. There is a need for an intrusion prevention and authentication
solution which limits access to the cloud development code to only authorized users,
while at the same time preventing malicious reconnaissance attempts to fingerprint the
cloud infrastructure or launch denial of service (DoS) attacks.
In this paper, we present results of a cybersecurity testbed deployed in production
for the Linux One community cloud. Our research addresses the unique cybersecurity
requirements of this environment, including improved authentication as well as identity
and access management within a cloud data center. The key points of novelty for this
work include the use of network-based identities in a hybrid public/private cloud;
specifically, we demonstrate a combination of BlackRidge Technology first packet
authentication and transport layer access control (TAC) technologies. We experimen‐
tally demonstrate user identity management in the Linux One community cloud,
including the novel ability to prevent unauthorized fingerprinting of key network
resources. Further, we have developed original software to parse the logs from these
appliances and related honeypots, performing geolocation and botnet classification. This
work is intended to address the leading concerns expressed in recent surveys of chief
information security officers in academia, and enable replication of our security solution
at other colleges and universities. We deploy BlackRidge Technology TAC virtual
appliances throughout the network which manage user identity based on the first packet
used in transport connection requests. This solution including software developed
specifically for this project which performs geolocation and attribution for all unau‐
thorized access attempts, and enables collection of analytic data on attempted attacks
which can be processed into actionable threat intelligence. Experimental results are
presented, demonstrating that our approach detected and blocked 1,161 unauthorized
access attempts in the first twelve hours of production deployment. Over a period of ten
days, our approach successfully blocked over 18,000 attacks, which we have attributed
zamfira@unitbv.ro
134 C. DeCusatis et al.
to locations in China, Korea, Brazil, Vietnam, and elsewhere. We also demonstrate the
ability to identify insider threats by running our authentication technology inside the
college firewall (an essential enabling feature for a NIST zero trust network [7]). We
present data demonstrating that this approach successfully prevents IP Spoofing and
Denial of Service attacks, and identifies network scanners such as Nessus if they are
operating on the cloud network. This functionality was not possible using conventional
network security approaches.
The paper is organized as follows. Section 1 provides an introduction and motivation
for this work, and an overview of our novel contributions. Section 2 describes TAC and
first packet authentication technologies in more detail. Section 3 provides experimental
results obtained from the Linux One higher education cloud deployment over a 30 day
period. Section 4 includes a summary and conclusions.
zamfira@unitbv.ro
Advanced Intrusion Prevention 135
in the first packet of the TCP connection request. The second gateway enforces the
network access policy by extracting the token, resolving the token to an identity, and
determining the identity’s authorizations. Trusted users (attempting to access the educa‐
tion network) have identity tokens inserted by Gateway A; untrusted users receive no
such authentication tokens. The TAC gateways are configured to protect sensitive
resources, such as the cluster of Linux servers. When the second gateway receives a
connection request, it extracts and authenticates the inserted identity token and then
applies a security policy (such as forward, redirect, or discard) to the connection request
based on the received identity. This gateway acts as a policy enforcement point trans‐
parent to the rest of the system architecture and backwards compatible with existing
network technologies. Trusted users will be authenticated by Gateway B, allowing them
full access to the Linux server cluster. Untrusted users are not recognized by Gateway
B, and their first packet requesting a new session is dropped, along with all responses
at or below the transport layer. In this manner, the untrusted user is unable to determine
that the Linux server cluster exists, and cannot begin to mount an attack. The attempted
access is logged in an external syslog server, which allocates enough memory to avoid
wrapping and over-writing log entries. Existing security information and event manage‐
ment (SIEM) tools can still be used to analyze the logs or generate alerts of suspicious
activity. We note that continuous logging of all access attempts is consistent with the
approach of a zero trust network (i.e. not allowing any access attempts to go unmoni‐
tored). Conventional denial of service (DoS) and port scanner attacks from an untrusted
user are similarly blocked, effectively cloaking the presence of the Linux server cluster
in this example. Note that the conventional IPS platform is no longer required, but may
remain in place since it is transparent to the TAC gateway authentication process. We
may also add features such as honeypots which accept redirect requests from a failed
access attempt at the TAC gateway (for example, SSH honeypots may be configured in
this manner). This enables the collection of attack data which may subsequently be used
to craft actionable threat intelligence, such as attack signatures. Both the identity inser‐
tion gateway and identity authentication gateway appliances can be implemented as
virtual network functions (VNFs) hosted on a virtual server, router, or similar platform.
This approach has several advantages, including separation of security policy from
the network design (i.e. network addresses and topologies) [7]. This approach works for
any network topology or addressing scheme, including IPv4, IPv6, and networks which
use the Network Address Translation (NAT) protocol and is compatible with dynamic
addressing often used with mobile devices. This approach extracts, authenticates, and
applies policy to the connection requests, not only protecting against unauthorized
zamfira@unitbv.ro
136 C. DeCusatis et al.
external reconnaissance of the network devices but also stopping any malware within
the protected devices from calling home (exfiltration). Security policies can be easily
applied at the earliest possible time to conceal network attached devices from unau‐
thorized awareness. By preventing unauthorized scanning and reconnaissance, TAC
disrupts the attacker’s kill chain, blocks both known and unknown attack vectors, and
stops lateral attack spreading within a data center. This approach is low latency and high
bandwidth since packet content is not inspected. Since the network tokens are embedded
in the TCP session request, they do not consume otherwise useful data bandwidth. The
combination of transport access control and a segmented, multi-tenant network imple‐
ments a layered defense against cybersecurity threats, and contributes to non-repudiation
of archival data. These techniques are also well suited to protecting public and hybrid
cloud resources, or valuable, high performance cloud resources such as enterprise-class
mainframe computers and higher education data centers. Further, this approach can be
applied to software defined networks (SDN), protecting the centralized SDN network
controller from unauthorized access, and enabling only authorized SDN controllers to
manage and configure the underlying network. Further, our implementation of TAC uses
an innovative identity token cache to provide high scalability and low, deterministic
latency. The token cache is tolerant of packet loss and enables TAC deployments in low
bandwidth and high packet loss environments.
3 Experimental Results
The Linux One geographically distributed community cloud (Phase One production
environment) created for these experiments is shown in Fig. 3. This cloud interconnects
two physical data centers, namely the Linux One cloud data center hosted at Marist
College near Poughkeepsie, NY; the IBM data center hosted in their Poughkeepsie, NY
facility. The Marist College and IBM Poughkeepsie data centers are located approxi‐
mately 8.5 km apart in upstate New York.
zamfira@unitbv.ro
Advanced Intrusion Prevention 137
Users connect to the Linux One Community Cloud via a secure Internet portal to
an Apache web server at the Marist College data center. Content management
servers in this data center host instances of OpenStack (Liberty and Juno releases),
Maria database server, IBM Java Development Kit (JDK), and IBM BlueMix
DevOps Build Engine. These applications are hosted on virtual machines (VMs)
partitioned in an IBM z Systems 113 enterprise server. It is necessary to securely
authenticate the long distance connection between the Marist College data center and
IBM Poughkeepsie data center, (which houses a processing server and content fulfill‐
ment engine), To authenticate traffic between these two data centers, BlackRidge
appliances implementing TAC and first packet authentication were implemented
between these locations as shown in the figure. A physical appliance was installed
at the edge of the IBM Poughkeepsie data center network, and a virtual appliance
hosted in an IBM z13 enterprise server Zvm virtual partition was installed at the
corresponding edge of the Marist College data center network.
To determine the effectiveness of the TAC appliances at cloaking attached systems,
we performed nmap scans of both the Marist College and IBM Poughkeepsie data center
networks before and after implementing the TAC appliances. Representative scans from the
Marist College and IBM Poughkeepsie data centers before implementing TAC are shown
in Figs. 4 and 5, respectively. From these scans, an attacker can clearly see the open port
22 on the Marist network, running OpenSSH 6.6.1, and a traceroute showing network hops
within the IBM network, among other reconnaissance data that would be useful in plan‐
ning an attack on these systems.
A representative scan after implementing TAC on this network is shown in Fig. 6
(results are equivalent for both the IBM and Marist network segments). Note that we
can no longer detect any open ports, including the exposure previously reported on port
22. All attempts to scan these hosts were successfully blocked by first packet authenti‐
cation, and all responses from the host due to these scans were successfully blocked by
TAC. The scan is now unable to determine the host operating systems, port or IP
addresses, or services running on the host. These results show that we can effectively
block fingerprinting of all devices located behind the TAC gateway.
zamfira@unitbv.ro
138 C. DeCusatis et al.
In order to better understand the attack vectors being used against this higher educa‐
tion cloud, we created a script in Python 2.7 to parse the syslog from a TAC appliance.
This script uses the Python regular expression operator ReGex to retrieve data from the
syslog including source and destination IP address and port numbers. This data is subse‐
quently processed through a geolocation module which we created for a related project
[7] to generate a report of the ISP, ASN, hostname, latitude, longitude, country, state/
province, and city of each attacker in JSON format. The TAC appliance was
programmed to automatically blacklist any IP address which attempted more than 100
accesses to the network within 30 s. The log parser which we have created also classifies
blacklisted IP addresses as potential DoS attacks or port scanners. We also collect data
on the number of attacks generated from unique IP addresses. All of this data is used to
create a profile of the attacker, which can be correlated with known botnets or hacker
groups.
zamfira@unitbv.ro
Advanced Intrusion Prevention 139
For example, during the first 12 h of monitoring the Linux One cloud after installing
the TAC appliance, there were numerous unauthorized attempts to access the system.
At this point the TAC system was placed into enforce mode, and successfully blocked
all subsequent unauthorized access attempts. The TAC appliance remained in enforce
mode for the next 10 days; a list of the top attacking IP addresses, and the top 10 attacking
countries, is shown in Figs. 6 and 7, respectively.
For example, analysis of the TAC appliance logs revealed a DoS attack against port
23 (originating from the Shangdong provide in China). We configured the TAC appli‐
ance to block unauthorized access attempts after 10 s of continuous attempts from a
given site, and to keep these sites blacklisted for one hour. Using this technique, we
successfully blacklisted the DoS attacker while continuing to collect log information on
the attack. In this manner, we have demonstrated that the TAC appliance provides
improved protection by identifying and blocking attacks which were previously unde‐
tected on the education network.
zamfira@unitbv.ro
140 C. DeCusatis et al.
Fig. 6. Number of attacks attempted from the top attacking source IP addresses
Further, we assessed the performance logs of the IBM z Systems enterprise server
in the Marist College data center before and after these attacks. Prior to implementing
the TAC appliance, the server attempted to block unauthorized attacks using network
appliances (such as intrusion prevention systems). This approach was replaced with a
single TAC gateway, protecting all VM’s on the server at the point of entry. We further
demonstrated that the TAC appliance was able to block IP spoofing on the network. By
comparing nmap scans of the network before and after implementing the TAC appliance,
we can show that attempts to perform IP spoofing are effectively blocked by the TAC
appliance. A scan of the network using the Spoofer tool (part of BGP-38 recommended
zamfira@unitbv.ro
Advanced Intrusion Prevention 141
by the National Science Foundation [7]) confirmed that both IPv4 and IPv6 packets
attempting to spoof the network were blocked (including private and routable addresses).
In a related test of egress filtering depth, the BGP-38 tracefilter test found the network
unable to spoof valid, non-adjacent source addresses through even the first IP hop.
Additional statistical data on attacks against this system was obtained using Long‐
Tail, an open source botnet classifier which we developed for a related project at Marist
College [7]. This classifier was used to identify SSH brute force botnet attacks against
the Linux One educational network, and to evaluate the effectiveness of blocking these
attacks using a conventional intrusion prevention system and the TAC appliance. For
this test, we first monitored the total number of attacks against an SSH honeypot
deployed in the Marist College network ingress from the IBM Poughkeepsie site; stat‐
istical analysis of these attacks is shown in Table 1. We then evaluated a commercially
available intrusion prevention system, the Juniper SRX 3600, under the same conditions;
results are shown in Table 2. We can see that the IPS helped reduce the number of attacks,
but did not eliminate them completely. Finally, we deployed the TAC appliance under
the same conditions; results are shown in Table 3. In this case, the combination of first
packet authentication and transport access blocking was able to successfully block all
brute force SSH attacks against the network, and demonstrated a significant improve‐
ment over the commercial IPS system alone.
Table 2. Attacks against the Marist education network mitigated by conventional IPS
Time frame Number Total SSH Average Standard Median Max Min
of days attempts per day deviation
Past day 1 30 N/A N/A N/A N/A N/A
This month 18 897 49.83 39.75 35.5 124 0
Last month 30 369 12.30 12.22 10 43 0
Table 3. Attacks against the Marist education network mitigated by TAC gateways
Time frame Number Total SSH Average Standard Median Max Min
of days attempts per day deviation
Past day 1 0 N/A N/A N/A N/A N/A
This month 18 0 0 0 0 0 0
Last month 30 0 0 0 0 0 0
zamfira@unitbv.ro
142 C. DeCusatis et al.
We have also demonstrated that a TAC gateway placed just inside the Marist College
firewall is useful in nonrepudiation of insider threats. When a bad actor inside the Marist
firewall is detected, efforts to trace the source of the attack traditionally stop at the Marist
NAT gateway. It can be a difficult, time consuming process to trace the IP address which
originated such an attack. However, a TAC gateway placed behind the Marist firewall
(on the Marist side of the NAT) can be used to authenticate the attacker’s source IP
address much more quickly and efficiently. This new functionality should be helpful not
only in discouraging insider threats, but also in helping the college comply with requests
and subpoenas from law enforcement agencies investigating such attacks.
4 Conclusions
References
1. McCarthy, S.: Pivot Table: U.S. Education IT Spending Guide, version 1, 2013–2018. IDC
publication GI255747, April 2015. http://www.idc.com/getdoc.jsp?containerId=GI255747
2. Lowendahl, J., Thayer, T., Morgan, G.: Top ten business trends impacting higher education.
Gartner Group white paper, January 2016. https://www.gartner.com/doc/3186325/top–
business-trends-impacting
3. Grama, J.: Data breaches in higher education. Educause Center for Analysis and Research,
May 2014. https://library.educause.edu/resources/2014/5/just-in-time-research-data-
breaches- in-higher-education
4. Fireye white paper: Cyber threats to the education industry, March 2016. https://
www.fireeye.com/content/dam/fireeye-www/current-threats/pdfs/ib-education.pdf
5. Stoneburner, G., Goguen, A., Feringa, A.: Risk management guide for IT systems. NIST
special publication 800-30, September 2012. http://csrc.nist.gov/publications/
PubsSPs.html#800-30
zamfira@unitbv.ro
Advanced Intrusion Prevention 143
6. Guilen, A., Rutten, P.: Driving Digital Transformation through Infrastructure Built for Open
Source: How IBM LinuxONE Addresses Agile Infrastructure Needs of Next Generation
Applications. IDC white paper, December 2016. https://public.dhe.ibm.com/common/ssi/ecm/
lu/en/lul12345usen/LUL12345USEN.PDF. Last accessed 22 Oct 2016
7. DeCusatis, C., Liengtiraphan, P., Sager, A., Pinelli, M.: Implementing zero trust cloud
networks with transport access control and first packet authentication. In: Proceedings of IEEE
International Conference on Smart Cloud, New York, NY, 18–21 November 2016
8. Amazon Web Services Identity and Access Management, April 2016. https://
aws.amazon.com/iam/. Last Accessed 20 May 2016
9. BlackRidge white paper: Dynamic network segmentation, August 2012. http://www.
blackridge.us/images/site/page-content/BlackRidge_Dynamic_Network_Segmentation.pdf
zamfira@unitbv.ro
Remote Laboratory for Learning Basics
of Pneumatic Control
Abstract. In this paper, a remote laboratory for learning the basic principles of
pneumatic control and realizing pneumatic control schemes is described. Goal is
to develop a remote system for our laboratory through which remote participants
(students, engineers, etc.) would be able to learn some basic principles of pneu‐
matic control. The first stage of developing a unique complex pneumatic scheme
with which several smaller, simpler tasks can be realized, as well as a user inter‐
face for the remote laboratory are shown.
1 Introduction
Following the constant technological progress and the increasing of electronic and
informatic literacy of the new generation of students, growing number of faculties and
universities around the world are introducing distance learning [1]. Distance learning
greatly increases the quality of teaching activities [2] on it, because the students have
the opportunity to organize their timetable and their activities. In the paper [3], is
described one example of a system that enables distance learning from the field of elec‐
trical engineering. Connecting different smaller electronic schemes are carried out using
one complex scheme in that system. The aim of this paper is to develop a remote labo‐
ratory that will enable learning of the basic principles of pneumatic control for the remote
participants.
Pneumatic systems are often finding their application in various branches of industry
due to a large number of advantages of compressed air [4]. For this reason, drawing of
pneumatic schemes and pneumatic control are studied in secondary schools and at the
universities as well. The remote participants, clients, can be students or engineers from
the industry. The great advantage of using remote laboratory for students is that they
can make a practical exercise in case they are absent from regular classes. The great
advantage of using remote laboratory for engineers who are the carrier power of modern
industry is that they can improve their skills throughout their lives (Life Long Learning
- LLL), without being absent from work. In addition to aforementioned, for all the clients
will be available the descriptions of basic principles of pneumatic control as well as
individual components.
zamfira@unitbv.ro
Remote Laboratory for Learning Basics of Pneumatic Control 145
zamfira@unitbv.ro
146 B. Bajči et al.
zamfira@unitbv.ro
Remote Laboratory for Learning Basics of Pneumatic Control 147
For better understanding, one example will be shown in extension. If a client want
to indirectly control a single-acting pneumatic cylinder, it is necessary to go through the
next steps (Fig. 1):
1. By activating the 0V1 valve (2/2) is allowed the supply of compressed air to the
service unit (0Z),
2. By activating 1V2 and 1V3 valves (2/2) is allowed the flow of compressed air from
service unit to the electrically activated 3/2 valve (1S2) and to pneumatically acti‐
vated 3/2 valve (1V6),
3. By activating the 1V5 valve (2/2) is allowed the flow of compressed air from elec‐
trically activated 3/2 valve (1S2) to pneumatically activated 3/2 valve, to its control
connector 12 (1V6),
4. By activating the 1V9 valve (2/2) is allowed the flow of compressed air from pneu‐
matically activated valve (1V6) to the single acting pneumatic cylinder (1A),
5. After the simulation of the physical interconnection of the components, by activating
electrically activated 3/2 valve (1S2) the single-acting pneumatic cylinder (1A) will
extract,
6. By deactivating electrically activated 3/2 valve (1S2) the single-acting pneumatic
cylinder (1A) will retract to its initial position.
Previously is already mentioned that with using the developed pneumatic scheme in this
paper, it is possible to realize eleven different, smaller exercises related to pneumatic
control. As it can be seen from Fig. 1, a large number of electro-pneumatic command
valves are used. Precisely for this reason, for the realization of remote control of this
system, a controller with large number of digital output signals is required. In this paper,
a controller of modular type, CompactRIO is used for this purpose.
A client can access to our laboratory through the CEyeClon platform [5, 6] and needs
to have installed only the CEyeClon Viewer software. It is necessary to request from
administrators an access key for the experiment. In Fig. 2, ways of communications in
our system are shown. When the client logs into the system, he/she connects to a remote
computer through the internet. That computer is physically connected to the controller.
Communication between the PC and the CompactRIO controller is accomplished by
using TCP/IP protocol. The electro-pneumatic command valves are connected to the
digital outputs of the controller. Live monitoring is enabled via web camera. When the
client logs into the system, it is necessary to launch a file called “Remote Laboratory
for Learning Basics of Pneumatic Control” from the desktop.
The homepage of the user interface then opens in internet browser with a list of
exercises. Exercises on this page are divided into two groups. The first group relates to
the direct control of a pneumatic actuator and consists of two exercises, one for single-
acting cylinder and one for double-acting cylinder. The second group relates to the indi‐
rect control of a pneumatic actuator. Within this group are classified nine different exer‐
cises. The user can choose between two languages, English or Serbian. By selection of
one exercise, a new window opens in the browser. In Fig. 3 the user interface for the
zamfira@unitbv.ro
148 B. Bajči et al.
first exercise of indirect control is shown. The first thing that the client can notice on
this page is the title of the exercise. Below the title, the text and the sketch of a physical
realization of the exercise are located.
On the left side of the user interface, below the text of the exercise, is located an area
provided for drawing the pneumatic scheme. Within this area, at the beginning, are
placed only the basic components, necessary for the realization of the selected exercise,
such as cylinders, valves, sensors, etc. A pop-up window appears by pressing the mark
zamfira@unitbv.ro
Remote Laboratory for Learning Basics of Pneumatic Control 149
of one of the components. That window contains the description and the picture of the
selected component, which enables the client to better understand the basic function of
it. In the top left corner a legend is located that explains the meaning of the color of the
pneumatic tubes on the scheme. These tubes are represented with lines. The line is red
when the tube is under pressure and it is black when the tube is not under pressure. Green
line represents a control signal and a blue line represents an exhausted tube. An area
with the commands for connecting certain components is located below the legend. The
mentioned commands appears and disappears, by pressing them, one after another. By
pressing the command, for example, “Connect the push-button with the command
valve” a line between these components is drawn. Also, at the same time, sending a
command to the controller is carried out. Certain 2/2 valve is activated, for this example,
1V6 showed in Fig. 1, and the push-button and the command valve are physically
connected. A turned on light, on the valve, can then be noticed on the camera.
Once the components are connected, it is necessary to press the 1S1 button at the
bottom of the user interface. A simulation is executed on the pneumatic scheme. All the
components will be activated and the cylinder will extract. The color of the pneumatic
tubes will be changed depending on the flow of the compressed air. Also, the extracting
of the cylinder will be seen on the camera. By releasing the push-button, in this exercise,
the cylinder will retract. The user interface works on the same principle for the other
exercises. At the top of the interface are located two buttons, for changing the exercises.
The basis for the realization of the user interface was JavaScript programming language
while LabView was used for programming the controller.
4 Conclusions
In this paper a remote laboratory for learning basics of pneumatic control is shown. A
unique pneumatic control scheme, used for the realization of several smaller exercises
is developed and presented. The use of the developed user interface is explained. The
CEyeClon platform is used for the realization of remote control. In this way is provided
a complete control over client access to the system. This system can be used as an integral
part of distant learning. Development of remote laboratories, like the one described in
this paper, is very important for the improvement of learning activities. In this way, it
is possible to attract the attention of a large number of new students and engineers, on
the study programs where the laboratories are used as well as enable a further education
for anyone interested in pneumatic control.
References
1. Horvat, A., Dobrota, M., Krsmanovic, M., Cudanov, M.: Student perception of Moodle
learning management system: a satisfaction and significance analysis. Interact. Learn. Environ.
23(4), 515–527 (2015)
2. Rodríguez-Sevillano, A.A., Barcala-Montejano, M.A., Tovar-Caro, E., López-Gallego, P.:
Evolution of teaching tools and the learning process: from traditional teaching to edX courses. In:
13th International Conference on Remote Engineering and Virtual Instrumentation (REV), UNED,
Madrid, 24–26 February 2016, pp. 42–49. IEEE (2016). ISBN 978-1-4673-8245-8 (2016)
zamfira@unitbv.ro
150 B. Bajči et al.
3. Bjekić, M., Božić, M., Rosić, M., Antić, S.: Remote experiment: serial and parallel RLC circuit.
In: 3rd International Conference on Electrical, Electronic and Computing Engineering,
IcETRAN 2016, Zlatibor, Serbia, 13–16 June 2016. ISBN 978-86-7466-618-0
4. Šešlija, D., Milenković, I., Dudić, S., Šulc, J.: Improving energy efficiency in compressed air
systems – practical experiences. Thermal Sci. (2016). ISSN 0354-9836
5. Zurcher, T.: Distance education in energy efficient drive technologies by using remote
workplace. In: 11th International Conference on Remote Engineering and Virtual
Instrumentation (REV), 26–28 February 2014. IEEE, ISBN 978-1-4799-2024-2
6. Zurcher, T., Rojko, A., Hercog, D.: Education in industrial automation control by using remote
workplaces. In: 3rd Experiment@ International Conference Online Experimentation (exp.at
2015), University of the Azores, 2–4 June 2015, Ponta Delgada. IEEE, ISBN
978-989-20-5753-8
zamfira@unitbv.ro
The Augmented Functionality of the Physical
Models of Objects of Study for Remote
Laboratories
1 Introduction
The advantages of distance education stimulate the improvement of its components [1],
among which in the last decade, the rapid development of remote laboratories
(RL) took place [2–4]. These laboratories include a server with a set of physical models
of the object of study. For example, in the laboratories of the Grid of Online Laboratory
Devices Ilmenau (GOLDi) there are physical models (PM) Elevators, 3-Axis-Portal,
Production Cell with devices for their control [5]. Users enter the data of technical
experiment from a remote computer. This information is sent over the Internet to the
server of the laboratory, where it is converted into control signals of physical and
(or) virtual model of the object of the experiment. The progress and results of the
experiment are perceived by the user by means of the user perceptual interface [6]. The
outputs of this interface are a user perceptual image and the flow of user commands. An
example of the User Perceptual Image (UPI) is a computer screen with a WEB – the
image of the physical model and the visual part of the virtual model of the object of the
© Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6_15
zamfira@unitbv.ro
152 M. Poliakov et al.
experiment [6]. However, as from the user’s point of view and from the point of view
of the designer of a remote laboratory, the object of the experiment is represented by a
system with physical and virtual elements that interact with each other and with the
environment. For description of such objects in a number of cases the term CPS –
Cyber Physical System is used, which declares the connection of physical objects with
computational algorithms [7]. Despite the importance of solving the issues of managing
remote users of RL and real-time students’ interaction with the model of the object, a
necessary condition for effective use of RL is a sufficient number of experiments with
models of the object of study, as well as the quality and informative value of the User
Perceptual Image.
In Sect. 2 of the article gives an overview of publications on technologies of the
augmented reality that used by the authors to extend the functionality of the PM of RL.
Section 3 describes the models of object of study and interfaces which are involved in
the formation of the image perceived by the user. Conclusions and Acknowledgements
set forth in the Sect. 4 and Acknowledgment section of the article.
2 State of Art
Contemporary researches are aimed primarily at improving the quality and informa-
tional content of the User Perceptual. For this Augmented Reality (AR) technology is
used. AR is a scientific discipline, the essence of which is disclosed in numerous
publications (e.g. [8–12], which holds regular scientific conferences [13], and pub-
lished by the journal [14]).
AR themes are reflected in previous sessions of the International Conference on
Remote Engineering and Virtual Instrumentation. So in works [15, 16] the LEDs’ state
change of physical model of a traffic light is displayed on the remote user’s screen in
the video window. Software RL detects certain changes in the shots of this image,
which are classified as events. Events control the induced overlay video on video
elements of the physical model. In this case, the images of moving vehicles are
induced.
The focus of this work is to use the added functionality of the RL models to
increase the quantity and quality of the experiments without significant changes in the
physical models.
The concept of functional models of objects of study was mentioned in [17] in
connection with the analysis of the structure of the hybrid model. The work presents an
example of the behavior of the physical model of the traffic light in RL RELDES [18].
The experiment is carried out with a hybrid model, which includes a physical model of
a traffic light, complemented behavior of fault diagnosis and simulation of defects of
the traffic light lamps.
In [19] it is proposed to add virtual physical model of the Elevator in the lab
GOLDi [5]. This model was originally used to explore the construction of the FSM
controls. By adding virtual model of flow control commands of users and the virtual
reality of a queue of users waiting for service, experiments with more complex control
algorithms, the experiments on the theories of queues and performance evaluation of
real-time systems become possible. Below are the new results in this direction.
zamfira@unitbv.ro
The Augmented Functionality of the Physical Models of Objects 153
Models and interfaces RL involved in the formation of the User Perceptual Image are
shown in Fig. 1.
Fig. 1. Models and interfaces RL involved in the formation of the user perceptual image
The physical interface includes the flow of information from the physical external
environment and physical model of the object of study to their hybrid model, as well as
the flow of control actions on the object of study by a hybrid model.
The virtual interface includes information from the virtual external environment
and the virtual models of the object of study to their hybrid model, the flow of control
actions on virtual objects of study and the flow of information to synchronize the
physical and virtual models of the object of study.
The role of hybrid models is in the selection, switching and integration of the
information coming through the physical and virtual interfaces depending on the
selected operation mode RL. In addition, the hybrid model generates streams of source
data for the work of the media model of the experiment and receives a stream of user
commands for the control of the experiment.
The network interface performs the standard functions of information exchange
between the server and the remote RL user’s computer.
On the remote user’s computer the media (visual, audio, etc.) model of the
experiment performs and the user command to manage a hybrid/virtual/physical model
of the object of study is processed.
Finally, the user interface implements interaction “sensors user – output device”
and “user input device”.
A physical model of the object of study is the means for the study of the control
system, which also includes the control device and the environment.
The interface of the physical model with the rest of the system (physical interface)
is implemented using sensors and actuators composition of which is shown on Fig. 2.
We will distinguish physical parameters sensors and image sensors. An example of the
first are sensors of electric currents and voltages, the speed of movement of the object
zamfira@unitbv.ro
154 M. Poliakov et al.
zamfira@unitbv.ro
The Augmented Functionality of the Physical Models of Objects 155
remote user’s computer. Each module implements a specific set of virtual functions in
the course of the experiment the object of study.
The object and the form of the generated functionality describe the module of
virtual interface. The object of functionality is the object of study, its physical model,
the external environment, including the flow of user’s commands, as well as technical
and software of RL. The types of the generated functionality are the image of the
object, UPI, parameters, and behavior of the object.
Scale of evaluating the conformity extent of the object’s functionality and the
interface’s module can have the following gradation: absent (f0), reduced (f1), equiv-
alent to model (f2), advanced (f3), equivalent to the object (f4), added (f5), and new (f6).
Feature comparison is shown in Fig. 3.
Fig. 3. A comparison of the functionality of the virtual interface: fCO, fPM - functionality of the
controlled object and its physical model
The bases of the coordinates of the simulated object parameters of the generated
functionality are: observable and controlled variables; unobservable and uncontrollable
variables; full range of observed, unobserved, controlled and uncontrollable variables.
Bases of simulated behavior regarding the objectives of the experiment and modes
of use of the object of study: the behavior in a normal mode; the behavior in emergency
mode; the technical state control of the elements of the object of study; the external
environment control.
The bases of the simulated behaviors with respect to the selected type of controls:
the behavior of discrete control on the basis of the FSM; the behavior of continuous
control based on the structure of the control system, the transfer functions of the
elements of the object of study (or its physical model) and regulators; the behavior of a
hybrid control in which the current state of the system formed by means of discrete
control, and the actions in this state are determined by the model of continuous control.
Time basis of simulated images, parameters, and behaviors includes the current
time; historical trend and forecast.
Virtual image generated by the modules of the virtual interface is dependent on the
modules that define the current virtual behavior, and the parameters of the object of
study and the external environment.
zamfira@unitbv.ro
156 M. Poliakov et al.
The main categories that characterize a virtual image are realism/metaphor; degree
of coverage (whole/part), scale; dimension (2D/3D mono/stereo) and the format in
pixels; type of media (text/photo/video); view direction (on the object/from the object);
illumination angle, the number of survey points; visualization object (object of
study/visual model/external environment/trend/design models (FSM graph, chart, UML
diagram, control program text); consistency with the sensors of the user (“visible”/
“invisible visualization”).
The virtual image must satisfy the requirements of ergonomics and technical
aesthetics.
As mentioned above, virtual interface via the generated functionality is transmitted
in a hybrid model. Structural scheme of the hybrid model is shown in Fig. 4.
The configuration of the hybrid model is controlled by RL. Images, parameters, and
status are connected through the source selector. Using the destination selector setting,
the involved modules receive input information, necessary for initialization and
synchronization.
A physical model (PM) and software image control units, the corresponding FSM,
take up important places in the structure of the hybrid model. The presence of archive
data of experiments on the machine carrier can expand the information basis of the
researcher and allows the use of statistical methods of research.
The results of the hybrid model in the form of a stream of images and values of tags
of the media model are transmitted via the network interface to the remote user’s
computer. A standard network is the Internet. In the context of its use in the RL, it must
meet the requirements in the exchange rate, especially if the UPI contains complex
images. The details of the transformation of tag values of the objects display of the
model in the elements of the UPI image are given in [17].
Examples of additional features that provide augmented functionality are given in
the table for the known physical models of an Elevator and Traffic light.
zamfira@unitbv.ro
The Augmented Functionality of the Physical Models of Objects 157
4 Conclusions
1. The specificity of the RL experiments is that their results are perceived by the user
remotely from the object of study. Therefore, the content and technology of creating
the UPI are key to improve the quality and diversity of experiments. Today the main
zamfira@unitbv.ro
158 M. Poliakov et al.
varieties of UPI are “live” WEB – a picture of the object in the experiment and
animated image controlled by the tags of the virtual model.
2. Key technology of improving the UPI is “augmented” technology. Used varieties of
this technology - augmented reality, augmented visuality, as a rule, do not affect the
behavior of the object of study. The behavior of the object supposes the dependence
of the output response from the internal state. Discrete behavior is specified using
the FSM. The continuous behavior is specified using the structure of backward
linkages and transfer functions of the regulators. Moreover, the new behavior leads
to new functionality of the object of study that allows us to speak about “augmented
functionality”.
3. The implementation of augmented functionality in RL is connected with the
interaction of a number of interfaces (physical, virtual, network, and perceptual user
interface) and models (physical, virtual, hybrid and visual). Added functionality is
synthesized with modules virtual interface managed by the hybrid model.
4. The following concepts are associated with added functionality: the concepts of
object functionality, add gradations, the basis of the coordinates of the simulated
parameters, the basis of simulated behavior regarding the objectives of the study/use
of the object, the basis of the simulated behaviors with respect to the selected type
of control and the basis of the time the simulated images, parameters, and behaviors.
These and other categories of functionality were analyzed.
5. It is proposed to use the term “The Media model of the object of study in the
external environment” instead of the term “visible model”. Categories of images,
which are reflected in the UPI while using added functionality, were analyzed.
The proposed methods enhance the functionality of the models object of study is to
be used for expanding the range of experiments with models of remote laboratories
GOLDi in Ilmenau University of Technology and Zaporizhzhya National Technical
University (Table 1).
Acknowledgment. This work was partially carried out within the European Community Project
“Tempus” ICo-op: Industrial Cooperation and Creative Engineering Education based on Remote
Engineering and Virtual Instrumentation 530278-TEMPUS-1-2012- 1-DE-TEMPUS-JPHES.
The authors are grateful to Ilmenau University of Technology (Germany) and the Zaporizhzhya
National Technical University (Ukraine) for the opportunity to work with remote laboratory
GOLDi.
References
1. Azad, A.K.M., Auer, M.E., Harward, V.J. (eds.): Internet Accessible Remote Laboratories:
Scalable E-Learning Tools for Engineering and Science Disciplines, Engineering Science
Reference, 645 p. (2012)
2. Gravier, C., et al.: State of the art about remote laboratories paradigms - foundations of
ongoing mutations. Int. J. Online Eng. (iJOE) 4(1), 1–9 (2008)
3. Remote and virtual tools in engineering: monograph/general editorship, Karsten Henke,
Dike Pole, Zaporizhzhya, Ukraine, 250 p. (2015). ISBN 978–966–2752–74–8
zamfira@unitbv.ro
The Augmented Functionality of the Physical Models of Objects 159
4. Gomes, L., Bogosyan, S.: Current trends in remote laboratories. IEEE Trans. Industr.
Electron. 56(12), 4744–4756 (2009). doi:10.1109/TIE.2009.2033293
5. GOLDi-labs cloud Website: http://goldi-labs.net
6. Richir, S., Fuchs, P., Lourdeaux, D., Millet, D., Buche, C., Querre, R.: How to design
compelling virtual reality or augmented reality experience? Int. J. Virtual Reality 15(1), 35–
47 (2015)
7. Terkowsky, C., Jahnke, I., Pleul, C., Licari, R., Johannssen, P., Buffa, G., Heiner, M.,
Fratini, L., Valvo, E.L., Nicolescu, M., Wildt, J., Tekkaya, A.E.: Developing tele-operated
laboratories for manufacturing engineering education. Platform for eLearning and Telemetric
Experimentation (PeTEX). Int. J. Online Eng. (iJOE) 6, 60–70 (2010). http://dx.doi.org/10.
3991/ijoe.v6s1.1378. REV2010, Vienna, IAOE, Special Issue 1
8. Wikipedia 2016: http://en.wikipedia.org/wiki/Augmented_reality
9. Cao, M., Li, Y., Pan, Z., Csete, J., Sun, S., Li, J., Liu, Y.: Creative educational use of virtual
reality: working with second life. IEEE Comput. Graph. Appl. 34(5), 83–87 (2014)
10. Hughes, C.E., Stapleton, C.B., Hughes, D.E., Smith, E.M.: Mixed reality in education,
entertainment, and training. IEEE Comput. Graph. Appl. 25(6), 24–30r (2005)
11. Schaf, F.M., Pereira, C.E.: Integrating mixed-reality remote experiments into virtual learning
environments using interchangeable components. IEEE Trans. Industr. Electron. 56, 4776–
4783 (2009)
12. Milgram, P., Colquhoun, H.: A taxonomy of real and virtual world display integration. In:
Ohta, Y., Tamura, H. (eds.) Merging Real and Virtual Worlds, pp. 5–30. Ohmsya Ltd.,
Springer (1999)
13. Vlada, M., Albeanu, G.: The potential of collaborative augmented reality in education. In:
The 5th International Conference on Virtual Learning, ICVL 2010. Targu – Mure, Romania,
29–31 October 2010, pp. 39–43 (2010)
14. Int. J. Virtual Reality. http://www.ijvr.org
15. Maiti, A.A.K., Maxwell, A.: Variable interactivity with dynamic control strategies in remote
laboratory experiments. In: International Conference on Remote Engineering and Virtual
Instrumentation, REV2016, Madrid, Spain, 24–26 February 2016, pp. 399–407 (2016)
16. Smith, M., Maiti, A., Maxwell, A.D., Kist, A.A.: Augmented and mixed reality features and
tools for remote laboratory experiments. In: Int. J. Online Eng. (iJOE) 7, 45–52 (2016).
http://dx.doi.org/10.3991/ijoe.v12i07.5851. Vienna, IAOE
17. Poliakov, M., Larionova, T., Tabunshchyk, G., Parkhomenko, A., Henke, K.: «Hybrid
models of studied objects using remote laboratories for teaching design of control systems».
Int. J. Online Eng. (iJOE) 9, 7–13 (2016). http://dx.doi.org/10.3991/ijoe.v12i09.6128. IAOE,
Vienna
18. Parkhomenko, V., Gladkova, O., Ivanov, E., Sokolyanskii, A., Kurson, S.: Development and
application of remote laboratory for embedded systems design. Int. J. Online Eng. (iJOE) 11
(3), 27–31 (2015). IAOE, Vienna
19. Poliakov, M., Larionova, T., Wuttke, H.-D., Henke, K.: Automated testing of physical
models in remote laboratories by control event streams. In: 2016 International Conference on
Interactive Mobile Communication, Technologies and Learning (IMCL), 17–19 October
2016, San Diego, CA, USA. 94 p., pp. 10–13. IEEE 978-1-5090-1197-1/16/$31.00 ©2016
zamfira@unitbv.ro
More Than “Did You Read the Script?”
Different Approaches for Preparing Students for Meaningful
Experimentation Processes in Remote and Virtual Laboratories
1 Introduction
Project ELLI (Excellent Teaching and Learning in Engineering Science) is a joint project
of the three German universities RWTH Aachen, TU Dortmund University and Ruhr-
University Bochum. Considering teachers’ and learners’ perspectives, the project aims
to improve existing concepts in higher engineering education and to develop new inno‐
vative approaches. In the past years, a pool of remote and virtual labs has been developed
and set up in order to gain flexibility in the usage of experimental equipment in different
pre-set scenarios. Teachers can either use these virtual and remote laboratories in class
for demonstrating engineering practice whereas the labs can support students to indi‐
vidually discover scientific concepts.
The use of labs in general can be distinguished in research, development and training
purposes. In engineering education, labs are often used to introduce students to experi‐
mental work or explain a phenomenon in a realistic way. The project ELLI aims for
several improvements in the field of teaching and learning in engineering science. A
main aspect is to establish remote learning experiences.
zamfira@unitbv.ro
More Than “Did You Read the Script?” 161
Fig. 1. Remote laboratory at TU Dortmund (right) with the graphical user interface for user-
experiment-interaction (left)
zamfira@unitbv.ro
162 D. Kruse et al.
courses. In order to perform experiments in the remote lab the user can use the specially
designed graphical user interface. Using this interface, it is possible to prepare, start,
pause, stop, watch, and even analyze the ongoing experiment (Fig. 1, left part).
Introducing experimentation exercises with the help of remote or virtual equipment into
educational processes is different to the instruction on classical hands on labs. Whereas
in hands on labs normally a scientific assistance guides or supervises the experimentation
process (and the learning process, too), in some of its parts, the essential of a virtual or
remote lab learning is the non-guided and non-supervised process. This process can be
seen in the following stages:
(1) Orientation
(2) Preparation
(3) Performing an experiment
(4) Report experimental results.
Before performing any type of experiment, preparation is needed [4]. A classic hands
on lab preparation is often based on a scriptum or any kind of document that has to be
read by the students before coming to the lab. Such a scriptum contains the theoretical
background and used methodology as well as technical characteristics and at least the
task that should be performed. The students have to get familiar with the content and to
be prepared to be tested on the experimental content. In virtual or remote labs, things
are a bit different. As the whole experience is meant to be highly independent, all aspects
of the process must work in an intuitive and helpful way, without lowering the necessary
effort for the student’s performance. One of the main differences is the feedback on the
student’s preparation. In a classic hands on lab, this is ‘assessed’ by a supervisor during
a short interview, a discussion or the observation of the physical preparation of the
experiment. Whereas these aspects are fitting for the hands on lab, the lack of a supervisor
in a remote or virtual setting leads to new challenges [5]. Here the two main challenges
are the examination of the necessary preparation and the option of giving feedback about
the process of flexibly setting up an experiment. The following approaches are dealing
with these challenges. As the remote laboratories are developed independently at the
two locations, the student’s preparation will be explain separately.
While offering remote learning resources, there is the question if a scriptum is still the
best way of preparing students for a remote experiment. The balance between a chal‐
lenging task and a guided experience is crucial for the whole remote learning process.
Therefore, the preparation phase has to be rethought. A setup for performing experiments
in the field of process technology can contain several apparatus and instruments. For
gaining experience in setting up an experimental plant, it must be possible to change
zamfira@unitbv.ro
More Than “Did You Read the Script?” 163
Fig. 2. A virtual workbench for developing, the repository area at the bottom and the connection
control area on the right side.
zamfira@unitbv.ro
164 D. Kruse et al.
created by choosing the type of connection, choosing its starting point and selecting its
end point. During the process of creating a connection, the symbols placed in the work‐
bench indicate their ability to accept or decline a connection on one of the four sides
with green or red dots, respectively (see Fig. 3).
Fig. 3. Symbols on the virtual workbench during the connection creation process.
In case of a student needing assistance or having finished the setup of a virtual process
scheme, a consistency test runs and checks the flow scheme created. For a virtual process
scheme of a flow testing rig there should be at least one suitable pump, pressure gauges
and a regulation valve connected in one loop. Open connections or missing equipment
is recognized and can be displayed to the student with a hint about how to complete the
setup. A flow scheme containing all necessary equipment with correct connections is
reviewed and reported as complete.
This consistency test can be adjusted in its complexity by adding more information
to each symbol available. Parameters like flow direction, generated pressure, pressure
drop or process fluid parameters can be reviewed in the consistency test. The more
information is respected, the more complex the review process is. However, the accuracy
of the enabled feedback can be enhanced and individualized with this enlarged infor‐
mation [5]. The results of such a consistency test can be used to allow the student’s
access to a real remote lab control or give them advice to review several parts of the
experiment’s documentation [4, 8].
The virtual process scheme is created by using an html 5 framework called phaser
(www.phaser.io). A more common use for this framework is known to develop computer
games for web or mobile applications. Therefore, the functionality of this framework
was highly useful to develop the virtual process scheme. As the code works well on
mobile devices, the virtual process scheme can easily be adapted to mobile use for even
more flexibility.
The explicit example of preparing the remote learning experiment about the meas‐
urement of pressure drop at flow testing rig can be easily adapted to other experiments.
The idea of the virtual process scheme works in each discipline that uses schemes or
drawings to show interaction and connectivity. Examples of use can be electrical circuit
zamfira@unitbv.ro
More Than “Did You Read the Script?” 165
drawings or drawings of mechanical balance of forces. The virtual process puzzle elim‐
inates some of the drawbacks of remote experiments in the field of independent experi‐
ment development and reflection about the state of the student’s preparation. While
eliminating some challenges, it also creates new ones, especially with the consistency
test and its usage to create a meaningful feedback.
zamfira@unitbv.ro
166 D. Kruse et al.
of this lab is the determination of material parameters. This material parameters are
needed to conduct a FEM simulation in the next step. Therefore, a first run of the experi‐
ment using the remote lab is conducted in a lecture or exercise. The lecturer controls the
experiment during lecture in front of the audience. The students can ask questions and
discuss their needs in interaction with the lecturer. In a second step, the students need
to book a time slot to conduct the experiment using the ilab server. In order to help the
students and make a smooth start using the remote lab possible, an online video explains
the most important steps. This video is available without any registration to the ilab
server (see: http://iul.eu/remotelabs/). With this information and help, the students can
conduct their experiments using the remote lab. After conducting the experiments, the
data can be downloaded and the material parameters can be calculated on their own
devices.
Another course context, in which the remote lab plays a crucial role, is a completely
online delivered course for international students, which is taken in advance of their stay
in Germany for the master study program [9]. Part of this course is to conduct online
experiments in internationally mixed students groups using the universal testing
machine in the remote lab for a tensile test. As the students are coming from all over the
world, one of the challenges is that their knowledge and competence in experimentation
theory and practice may differ significantly. Whereas for some of the students inde‐
pendently performed experimentation processes may be normal and largely trained, this
is not the case at all for others. It may be even the case that some students are introduced
to experimentation equipment for the first time in their lives. Nevertheless, for the
experimentation with the remote lab at TU Dortmund, it is important to bring the students
on an adequate level of competence in experimentation and material characterization.
The Authors decided to make use of the differences and build heterogeneous and inter‐
nationally mixed students group for the important preparation phase. Within these phase,
the students did not receive a written script with all the important information, but they
were asked to do their own individual research about material characterization, based
on guiding questions. Figure 4 shows pictures given to the students as a starting point
for their research.
Fig. 4. Pictures used to guide students in their research process for material characterization
Taking the pictures shown in Fig. 4 as a starting point, the students are asked to
answers questions as follows:
zamfira@unitbv.ro
More Than “Did You Read the Script?” 167
1. In the first picture, you see the universal testing machine used at the IUL.
1.1. What are important parts?
1.2. How does such a machine work?
1.3. What is the theoretical background of the tensile test?
1.4. What is it used for?
2. The following pictures show stress strain diagrams.
2.1. What do they show?
2.2. What is the difference between the two diagrams?
2.3. How are they worked out?
2.4. What are important areas?
2.5. Which material properties can be gained through the connected data and how?
Using this approach for experimental preparation, it is possible on the one hand that
students themselves can directly develop knowledge about the respective experimenta‐
tion process. On the other hand, they see and learn where they may have important gaps
in knowledge, especially in comparison to other students. Furthermore, and this may be
the most important aspect, they can learn from each other. As they do have totally
different educational backgrounds, they recognize while answering these questions in
their respective group how far their personal concepts in experimentation differ from
the others’ concepts. With the help of each other, the students can leverage their indi‐
vidual knowledge about tensile testing and are finally on the same and needed level for
successful experimentation. To make sure that all students really are on the same level,
they have to present their research results in the following course meeting, and the most
important aspects are discussed again in the whole group. Observing the students during
the following experimentation process and assessing their results, it becomes clear that
they are well prepared for the experimentation by going through the procedure explained
above. Especially during the discussion of the experiment’s results, they benefit from
their former research in advance of the experimentation. Furthermore, they show good
abilities to connect their results with the explanations given in the literature.
Using remote learning processes in higher engineering education allow a flexible and
individual learning process. Due to the boundary conditions of the physical pre-set setup,
a creative discovery of scientific concepts lying behind the experimentation process is
limited. With different approaches for student activation and preparation, such as virtual
process schemes (VPS), static remote laboratory setups can be used in scenarios that
give a more flexible experience. With this, students are asked to take more personal
responsibility for their research, their personal learning process and gained knowledge.
They can prove it with several creations in VPS, getting feedback about their ideas by
the system. In a next step, such approaches even can be used for organizing the access
to the laboratory environment based on students’ performance during the preparation
process. For example, the access to the remote lab could be allowed only to those students
who received an adequate reflection/feedback to the tasks before the experimentation.
zamfira@unitbv.ro
168 D. Kruse et al.
Even if there is existing research on the usage of preparation activities, there is still
work to be done. Since the ELLI project starts its second runtime of five years in 2016,
the presented approaches are put into practice and evaluated within the next two years.
Hence, research results are expected to be looking at the question how different students
react on different preparation activities and in which intensity different kinds of such
activities are more or less suitable for different types of remote labs.
5 Summary
References
1. Frerich, S., Kruse, D., Petermann, M., Kilzer, A.: Virtual labs and remote labs: practical
experience for everyone. In: Proceedings: IEEE Global Engineering Education Conference
(EDUCON), pp. 312–314 (2014)
2. Terkowsky, C., Jahnke, I., Pleul, C., May, D., Jungmann, T., Tekkaya, A.E.: Pe-TEX@Work:
designing CSCL@Work for online engineering education. In: Goggins, S.P., Jahnke, I., Wulf,
V. (eds.) Computer-Supported Collaborative Learning at the Workplace - CSCL@Work,
Computer-Supported Collaborative Learning Series, vol. 14, pp. 269–292. Springer, New York
(2013). ISBN 978-1-4614-1739-2
3. Tekkaya, A.E.: Metal forming. In: Grote, K.-H., Antonsson, E.K. (eds.) Handbook of
Mechanical Engineering, Chap. 7.2, pp. 554–606. Springer, Heidelberg (2009)
4. Bochicchio, M.A., Longo, A.: The importance of being curricular: an experience in integrating
online laboratories in National Curricula for High Schools. In: Proceedings of 11th
International Conference on Remote Engineering and Virtual Instrumentation (REV), pp. 450–
456 (2014)
5. Graven, O.H., Samuelsen, D.A.H.: Remote laboratories with automated support for learning.
In: Proceedings of 10th International Conference on Remote Engineering and Virtual
Instrumentation (REV), pp. 1–5 (2013)
6. Kruse, D., Frerich, S., Petermann, M., Ortelt, T.R., Tekkaya, A.E.: Remote labs in ELLI: lab
experience for every student with two different approaches. In: Proceedings of IEEE Global
Engineering Education Conference (EDUCON), pp. 469–475 (2016)
7. Wuttke, H.D., Hamann, M., Henke, K.: Integration of remote and virtual laboratories in the
educational process. In: Proceedings of 12th International Conference on Remote Engineering
and Virtual Instrumentation (REV), pp. 157–162 (2015)
zamfira@unitbv.ro
More Than “Did You Read the Script?” 169
8. Dias, F., Matutino, P.M., Barata, M.: Virtual laboratory for educational environments. In:
Proceedings of 11th International Conference on Remote Engineering and Virtual
Instrumentation (REV), pp. 191–194 (2014)
9. May, D., Tekkaya, A.E.: Using transnational online learning experiences for building
international student working groups and developing intercultural competences. In:
Proceedings of American Society for Engineering Education’s 123rd Annual Conference and
Exposition “Jazzed about Engineering Education”, 26th–29th June 2016, New Orleans,
Louisiana, USA (2016). doi:10.18260/p.27171
zamfira@unitbv.ro
Collecting Experience Data from Remotely
Hosted Learning Applications
1 Introduction
Nowadays, organizations, companies and universities are collaborating in the
deployment and integration of learning applications. These applications range
c Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6 17
zamfira@unitbv.ro
Collecting Experience Data from Remotely Hosted Learning Applications 171
from simple tools like interactive assessment applications, to others for domain-
specific learning environments like remote laboratories. Thus, students may com-
monly access learning tools which are hosted in other universities or organiza-
tions, and rarely use applications that are actually deployed in their university’s
servers. The ability to integrate multiple learning applications from different
organizations allows sharing resources and reducing costs in the deployment
of learning systems. In this sense, Learning Tools Interoperability (LTI) [7] is
the main current leading technology for integrating learning applications with
platforms like Learning Management Systems (LMS), portals, learning object
repositories, and other educational environments, including Massive Open Online
Course (MOOC) platforms.
On the other hand, the integration of learning applications also requires data
collection as well as tools interoperability. Data collection allows learning sys-
tems to implement Learning Analytics (LA) processes. LA are useful to measure,
analyze and report about learners in order to optimize their learning. For exam-
ple, interactions and steps followed by learners in a remote lab could be used
to analyze the learning experience. Regarding the data collection, Tin Can API
(sometimes known as Experience API or xAPI) [10] is a specification for learn-
ing technology that makes this possible. This API captures data in a consistent
format about learners activities and enables dynamic tracking of activities from
any learning system. In addition, Tin Can API uses a Learning Record Store
(LRS), a data store system that serves as a repository for learning records.
Both learning technologies, LTI and Tin Can API, are supported by nowa-
days LMS, either natively or through plugins. Several works have previously
used Tin Can API in order to apply LA, for example, SmartKlass, [8] a multi-
platform solution that enables data tracking through a dashboard and that can
be embedded in Moodle or any other LMS. Other works have used LTI to get
interoperability between LMS and remote tools, for example, [11] shows how
to develop an external tool for e-Assessment. Also we can find specific solu-
tions that provides ad-hoc integration of both technologies, among them, [1]
describes an e-learning architecture with analytic capabilities aimed at training
Unmanned Autonomous Vehicles (UAV) operators. However, there is no seam-
less integration between these technologies in order to provide learning systems
with experience data from remotely hosted learning applications.
In this sense, our proposal defines a learning system architecture ready to
apply advanced LA techniques on experience data collected from remotely hosted
learning applications through a seamless integration between LTI and Tin Can
API. The key outcomes of our proposal are:
zamfira@unitbv.ro
172 F.J. Garcı́a Clemente et al.
Following this Sect. 1 on the objective and structure of this paper, the Sect. 2
presents the motivation example, which is used throughout all the paper to intro-
duce the concepts related to our proposal. In the Sect. 3, we describe our pro-
posal, and how to get a seamless integration of the learning technologies. Based
on this, the Sect. 4 shows an implementation of our proposal. Subsequently, the
fifth section discusses a specific deployment, which shows the integration using
resources located in three European universities. Finally, conclusions and future
work are drawn in the Sect. 6.
Fig. 1. Use case where the LTI and Tin Can API technologies are used.
zamfira@unitbv.ro
Collecting Experience Data from Remotely Hosted Learning Applications 173
and the other acts as the Tool Consumer, as shown in Fig. 1. The Tool Provider
shares learning applications from simple Javascript-based physics simulations to
complex remote laboratories. The Tool Consumer uses LTI services to provide
local learners with access to remote tools. Learners use their learning space
through web browser and their interactions are stored into an LRS via Tin
Can API.
The learner’s LMS can store learning experience data like the time at which
the learner logged in and out, the time she spent connected, or a session count
into the LRS. LMS collected data could be analyzed by LA software to report
useful information in order to assess and evaluate the learning experience. How-
ever, the data collected by the LMS is of no use if we want to analyze the
learning experience in depth. Other learning data is absolutely necessary, such
as the learner’s interactions with the learning applications, mouse and keyboard
events, button clicks or changes in input elements. Advanced LA software based
in data mining and data analysis can process this kind of interactions for classi-
fication and/or clustering. For example, LA software could automatically group
learners and identify who find more difficulties to interact or solve a task defined
in the learning application.
Tin Can API provides a mechanism to collect learner’s interactions in learn-
ing applications, but it is not currently supported by LMS in this way. LMS
usually use Tin Can API to store data related to learner’s experience extracted
from their own database. Therefore, LMS lack a mechanism to collect experience
data from learning applications. Moreover, the collection of learning experience
data becomes more complex when learning applications are remotely hosted.
Especially, proper authentication and right management mechanisms between
the LMS, the tools, and the LRS are missing.
zamfira@unitbv.ro
174 F.J. Garcı́a Clemente et al.
Fig. 2. Proposal for seamless integration between LTI and Tin Can API.
The LTI link requires a manual configuration process that consists in the
exchange of OAuth credentials. Teachers or course managers are in charge of
the LTI configuration and so, they can decide when a tool is included or shared
in the LMS. When a tool is included in a course, a learner can launch it. The
LMS internal process consists in a Basic LTI Launch Request where the Tool
Consumer provides the learner’s browser with all LTI parameters (OAuth para-
meters, context information, user identification and other learning information)
and then the browser uses them to get access to the tool delivered by the Tool
Provider. When the Tool is loaded in the learner’s browser, it can connect to a
remote lab if it is necessary. In this case, the tool must include the access cre-
dentials, required to get camera images and interact with actuators and sensors.
The LTI parameters below are taken from a Basic LTI sample launch data.
user id = 288816824
resource link id = 18551-bb669-e1e416
resource link title = System Activity
context id = 456434513
launch presentation document target = iframe
launch presentation return url = http://unilabs.dia.uned.es/lab32/return.php
lis person name full = “Felix J. Garcia”
lis person contact email primary = fgarcia@um.es
lti message type = basic-lti-launch-request
lti version = LTI-1p0
tool consumer instance guid = unilabs.dia.uned.es
tool consumer instance name = UNILABS
tool consumer instance description = UNILABS (LMS Moodle)
tool consumer instance url = http://unilabs.dia.uned.es
zamfira@unitbv.ro
Collecting Experience Data from Remotely Hosted Learning Applications 175
The parameter user id uniquely identifies the user, while the lis parameters
contain information about the user account that is performing the LTI launch
request. The specific meaning of the content in these fields is defined by Learning
Information Services (LIS) [6]. The parameters launch presentation describe
the kind of browser window/frame where the Tool Consumer has launched the
Tool. The fields tool consumer instance give details of the Tool Consumer and
the ouath are produced by the signing process. The oauth consumer key para-
meter identifies which Tool Consumer is sending the message allowing the Tool
Provider to look up the appropriate secret for validation.
In addition to the standard LTI parameters, the creator of a LTI link can
add custom key/value parameters to a launch, which are to be included with
the launch of the LTI link. When there are custom parameters, each custom
parameter is included into POST data when a Basic LTI launch is performed.
Creators of LTI links should limit their parameter names to lower case and to
use no punctuation other than underscores.
Our solution proposes to include a set of key/value pairs into the optional
custom section in the LMS that originally authored the link (i.e. Tool Consumer).
These custom parameters define the Tin Can connection between the Tool and
LRS Proxy. For example, the following LTI custom parameters complete the
previous Basic TLI sample launch data.
The field custom T inCan base endpoint contains the LRS proxy endpoint
service, while custom T inCan activity id identifies the Tin Can object and
custom T inCan verbs defines the verbs that must be used in the Tin Can
statements. These fields are required. Note that the Tin Can actor is identi-
fied unequivocally by the parameter lis person contact email primary.
These parameters are sent back to the external tool when the tool is launched.
If the LTI link is imported and then exported, the custom parameters should
be maintained across the import/export process unless the intent is to redefine
the link.
zamfira@unitbv.ro
176 F.J. Garcı́a Clemente et al.
local LMS (typically a username and password) and then get access to the learn-
ing space where he/she could find Tools directly without additional authentica-
tion process in other remote LMS. This LTI process, based on OAuth protocol
and called single sign-on (SSO), allows users to enter their credentials to gain
access to multiple systems just once.
In the same way, our solution proposes that learner gains access to LRS proxy
using a SSO mechanism and so avoiding a new authentication process. Specif-
ically, when the Tool is launched, it is presented into a type of browser win-
dow/frame (identified by the LTI field launch presentation document target)
embed in the learner’s learning space. That allows the Tool to get access to LRS
Proxy service located into the learner’s LMS using the current session.
{
”actor”: {
”mbox”: ”mailto:fgarcia@um.es”
},
”verb”: {
”id”: ”http://unilabs.dia.uned.es/xapi/verbs/changed”,
”display”:{”en-US”: ”changed”}
},
”object”:{
”id”: ”http://unilabs.dia.uned.es/xapi/ActivitySystem”
},
”result”: {
”extensions”: {
”http://unilabs.dia.uned.es/xapi/extensions/name”: ”processes”,
”http://unilabs.dia.uned.es/xapi/extensions/value”: ”10”
}
}
}
The actor object could have two properties, “name” and “mbox”, but only
“mbox” uniquely identifies the user. Verbs in Tin Can are URIs, and should be
paired with a short display string. Typically, the object will be a tool and the
result will include the extensions fields in order to provide a complete description
about the learner interaction.
Additionally, our solution proposes that the allowed verbs are set by the LTI
custom field custom T inCan verb. Therefore, the external application should
zamfira@unitbv.ro
Collecting Experience Data from Remotely Hosted Learning Applications 177
only send statements with these verbs. In this sense, the application must be
aware about possible verbs that can be requested. Considering the learning
application is running in a web browser, we propose that valid verbs must be
associated to HTML events. For example, the onchange event is related to the
changed verb. In addition, the event is triggered by actions inside a HTML ele-
ment and even it might include relevant action values. These event elements
are included into the statements using the extensions fields, as showed in the
previous example.
4 Implementation
Plugins enable the addition of new features and functionality to Moodle, such
as new activities, new quiz question types, new reports, integration with other
systems and many more. Specifically, LRS proxy is a web service into a local
plugin.
The following description declares the service including the name that iden-
tifies the plugin, web service functions and internal properties.
$services = array(
’LRS Proxy’ =>array(
’shortname’ =>’lrsproxy’,
’functions’ =>array (’lrsproxy echo text’, ’lrsproxy store statement’,
’lrsproxy store statements’, ’lrsproxy retrieve statement’,
’lrsproxy fetch statements’, ’lrsproxy store activity state’,
’lrsproxy retrieve activity state’, ’lrsproxy fetch activity states’,
’lrsproxy delete activity state’, ’lrsproxy clear activity states’),
’restrictedusers’ =>1,
’enabled’ =>0
)
);
The function lrsproxy echo text is only for testing purposes. The rest of func-
tions are divided into two groups. One is related to the functions for storing,
retrieving and fetching statements. These functions are used by Tools and might
also be used by other applications that can manage Tin Can statements. The other
group is for storing, retrieving and fetching states. These functions might be used
by Tools that wants to save arbitrary documents in the context of a particular
learner and particular Tool, for example, a snapshot of the learner’s experience.
zamfira@unitbv.ro
178 F.J. Garcı́a Clemente et al.
In relation to the internal implementation, the plugin was deployed using Tin
CanPHP [10], which provides a PHP library for implementing the Tin Can API.
Moreover, this library includes examples to show how to use Tin Can endpoints
services available to a SCORM Cloud account.
model.addLRSListeners = function(verbs) {
...
if(verbs.indexOf(’moved’) >-1){
document.addEventListener(’mousemove’, model.sendMovedInteraction);
document.addEventListener(’touchmove’, model.sendMovedInteraction);
}
...
};
zamfira@unitbv.ro
Collecting Experience Data from Remotely Hosted Learning Applications 179
Figure 4 presents the user interface for the application that was deployed by
EjsS in order to show a real-time graphics with the system activity and two
form inputs with the number of CPU cores online and the number of processes
running, as well as buttons to increase or decrease both input values.
In relation to this remote laboratory, note that the load average is a mea-
sure of system activity, calculated by the operating system and expressed as a
fractional number. In order to ensure adequate performance, the load average
should ideally be less than the number of CPU cores in the system. However,
learners can change the number of processes or CPU cores online to visualize
how the load average evolves.
zamfira@unitbv.ro
180 F.J. Garcı́a Clemente et al.
The Moodle configuration for the LTI link is shown in the Sect. 3.1 and an
example of Tin Can statement generated by this application is shown in the
Sect. 3.3. SCORM Cloud provides a simple interface for LRS endpoint service
configuration and statement viewer. However, it could be replaced by other LRS,
for example, Learning Locker [5].
Other existing remote laboratories can be included into the architecture
if the LRS proxy client is integrated in the application, i.e. if the application
processes the custom LTI fields and sends the Tin Can API statements to the
LRS proxy. Although this functionality is implemented into EjsS library, it could
be extracted and used independently.
Moreover, LRS could be shared by several Learning Management Systems
and so, they could even share LA tools in the future. In this way, organizations
can increase their goals of sharing resources and reducing costs in the deployment
of learning systems.
Finally, while our implementation uses Moodle, the same elements could be
deployed with other LMS, for example openEdX [2] or Graasp [3]. In fact, since
our proposal is based on standard technologies, the integration between different
LMS is supported.
References
1. Dodero, J.M., González-Conejero, E.J., Gutiérrez-Herrera, G., Peinado, S., Tocino,
J.T., Ruiz-Rube, I.: Trade-off between interoperability and data collection perfor-
mance when designing an architecture for learning analytics. Future Gener. Com-
put. Syst. 68, 31–37 (2017)
2. edX. Open edX: Open Courseware Development Platform. https://open.edx.org/.
Accessed 31 Oct 2016
3. EPFL React Group: Grassp project. http://graasp.eu/. Accessed 31 Oct 2016
4. Clemente, F.J.G., Esquembre, F.: EjsS: A JavaScript library and authoring tool
which makes computational-physics education simpler. In: Poster Presented at the
XXVI IUPAP Conference on Computational Physics (CCP), Boston, USA (2014)
5. HT2 Labs: Learning locker. https://learninglocker.net/. Accessed 31 Oct 2016
zamfira@unitbv.ro
Collecting Experience Data from Remotely Hosted Learning Applications 181
6. IMS Global Learning Consortium: IMS global learning information services best
practice and implementation guide. http://www.imsglobal.org/lis/. Accessed 31
Oct 2016
7. IMS Global Learning Consortium: Learning tools interoperability. https://www.
imsglobal.org/activity/learning-tools-interoperability. Accessed 31 Oct 2016
8. Learning Analytics Technologies for Education: KlassData. http://klassdata.com/.
Accessed 31 Oct 2016
9. Rustici Software: SCORM cloud. https://cloud.scorm.com/. Accessed 31 Oct 2016
10. Rustici Software: Tin Can API. https://tincanapi.com/. Accessed 31 Oct 2016
11. Sierra, A.J., Martı́n-Rodrı́guez, A., Ariza, T., Muñoz-Calle, J., Fernández-Jiménez,
J.J.: LTI for interoperating e-Assessment tools with LMS. In: Methodologies and
Intelligent Systems for Technology Enhanced Learning, 6th International Confer-
ence, pp. 173–181. Springer, Switzerland (2016)
12. UNED Labs: Moodle LRS proxy. https://github.com/UNEDLabs/moodle-local
lrsproxy. Accessed 31 Oct 2016
zamfira@unitbv.ro
“Remote Wave Laboratory” with Embedded
Simulation – Real Environment
for Waves Mastering
1 Introduction
Waves and their phase sensitive interference and superposition are important phe-
nomena constituting a major problem in students’ teaching of waves and optics, due to
the necessary students’ imagination. Then, the phenomena of interference of waves and
their superposition are difficult to understand. The proposed “Remote Wave Labora-
tory” is aimed at real measurements of the phase and the most frequent phenomena of
phase-sensitive wave superposition on real physical instrumentation with multiple use
and applications. As a teaching tool for better understanding of real measurements
serves the embedded real multiparameter simulation of the observed phenomena
introduced for the first time in our remote experiments.
2 Purpose or Goal
The whole system of the remote experiment (RE) “Remote Wave Laboratory” is
conceived to enable demonstrating the basic concepts of wave phenomena, as:
– The concept of the basic parameters of harmonic waves - the amplitude, the fre-
quency, the period, the initial phase, the phase velocity and the wavelength,
– The concept of the instantaneous phase (corroborated with our electronic phase
laboratory) as the function of the elapsed time and the path covered by the wave (in
relation to two periodicities of waves - in time and in space),
– The concept of the phase sensitive interference and the superposition of
parallel/perpendicular waves.
where the amplitude A and the initial phase D/ of the resulting signal is
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
A¼a 2ð1 þ cosðDuÞÞ; ð2Þ
sinðDuÞ
tgðD/Þ ¼ : ð3Þ
1 þ cosðDuÞ
which reduces in particular cases Du = 0 and p/2 into the straight line and circle,
respectively, in other cases of Du gives an ellipse.
• Phase measurements - it is then possible to determine the phase shift of both
signals (waves) Du from the position of the ellipse as shown Fig. 1.
Y
Du ¼ arcsin ð5Þ
H
zamfira@unitbv.ro
184 F. Schauer et al.
Fig. 1. Scheme for the phase shift Du of two waves determination using Eqs. (4) and (5)
Fig. 2. Examples of student’s work on hands - on experiment (a) Phase shift Du dependence on
the difference of the wave path Dx, (b)–(d) Superposition of two waves for phase shift
Du = 0 rad (b), Du = p/2 rad (c), Du = p rad (d), all for the wavelength k = 35.3 cm; the upper
panel shows signal of both waves, the middle panel phase-sensitive superposition-Lissajous
figures (for perpendicular waves) and bottom panel phase-sensitive interference signal (for
parallel waves)
zamfira@unitbv.ro
“Remote Wave Laboratory” with Embedded Simulation 185
Loudspeaker posiƟoning
ISES microphones
AC generator
Fig. 3. Schematic arrangement (upper panel) and the real RE “Remote Wave Laboratory”
(lower panel) with the loudspeaker as the acoustic wave source, two acoustic detectors 1 and 2
and the driving motor for moving detector 2, producing the phase shift Du of both signals,
corresponding to the detectors´ distance Dx
sound waves, the middle panel the phase-sensitive superposition - Lissajous figures (for
perpendicular waves) and the bottom panel shows the phase-sensitive interference
signal (for parallel waves).
zamfira@unitbv.ro
186 F. Schauer et al.
Fig. 4. Example of the view of the RE web page “Remote Wave Laboratory”, measured data,
(left) and simulation of the observed phenomenon (right); from upper graph: both
signals-perpendicular superposition- interpherence; the position of the movable detector is
visible in the live stream
zamfira@unitbv.ro
“Remote Wave Laboratory” with Embedded Simulation 187
Fig. 5. The general mathematical unit for ISES remote experiments enabling both arithmetical
operations and differential equations solutions
zamfira@unitbv.ro
188 F. Schauer et al.
When the client starts RE, the Measureserver begins the communication with
ISES HW. Then the RE is ready to perform all the required measurements according to
the web page instructions given by the client. Then, the Measureserver obtains
experimental data from the ISES modules (meters, sensors and probes) and transports
them again on the client’s web page for the analyses [3].
The embedded ES works in a similar way, replacing the ISES module by the
mathematical solver and providing the data for graphical comparison with the mea-
sured data. The difficult problem was the synchronization of both measured and sim-
ulated data into one time dependent graphical representation to study the role of model
parameters on the resulting signals.
4 Conclusions
The remote environment “Remote Wave Laboratory” provides the following knowl-
edge from waves:
– Phase of the wave as a function of the covered distance (with respect to the ref-
erence signal and its linearity),
– To examine parameters of the wave - the phase velocity and the wavelength in a
medium, the amplitude, the frequency and the period of the wave,
– To examine the concept of the coherence of two acoustic waves,
– To show the phase sensitive interference of two parallel waves and find the con-
ditions for extremes,
– To show the phase sensitive superposition of two perpendicular waves and to find
the phase shift and amplitude to the reference wave,
– To find the integer quotient of frequencies of an unknown waves.
Acknowledgement. The support of the project of the Swiss National Science Foundation
(SNSF) - “SCOPES”, No. IZ74Z0_160454 is highly appreciated. The support of the Internal
Agency Grant of the Tomas Bata University in Zlin No. IGA/FAI/2016 for PhD students is
acknowledged.
zamfira@unitbv.ro
“Remote Wave Laboratory” with Embedded Simulation 189
References
1. Ozvoldova, M., Schauer, F.: Remote laboratories in research-based education of real world.
In: Frankfurt, F.S. (ed.), p. 157. Peter Lang International Academic Publisher (2015) ISBN
978-80-224-1435-7
2. Gerza, M., Schauer, F., Dostal, P.: Embedded simulations in real remote experiments for ISES
e-Laboratory. In: EUROSIM 2016, Oulu, Finland, pp. 653–658. ISBN 978-1-5090-4119-0
3. Gerza, M., Schauer, F.: Intelligent processing of experimental data in ises remote laboratory.
Int. J. Online Eng., 58–63 (2016). ISSN 1861-2121. Austria
4. Inspiration of Prof. F. Esquembre in Solver Compiling is Appreciated
zamfira@unitbv.ro
Remote Laboratories: For Real Time Access to Experiment
Setups with Online Session Booking, Utilizing a Database
and Online Interface with Live Streaming
1 Introduction
Laboratory experiments are the integral part of Engineering Education. The main focus
is to gain access to these lab experiments over the internet using various integration
tools. Remote laboratory (also known as online laboratory, remote workbench) is the
use of telecommunications to remotely conduct real (as opposed to virtual) experiments,
at the physical location of the operating technology, Enabling the students to utilize these
technology from a separate geographical location. Supported by resources based on new
information and communication technologies, it is now possible to remotely control a
wide variety of real laboratories.
zamfira@unitbv.ro
Remote Laboratories: For Real Time Access to Experiment Setups 191
2 Architecture of Guacamole
In cloud computing environment, there are various important issues, including standard,
virtualization, resource management, information security, and so on. Among these
issues, desktop computing in virtualized environment has emerged as one of the most
important ones in the past few years. Currently, users no longer use a powerful, more-
than-required hardware but share a remote powerful machine using light weight thin-
client. A thin-client is a stateless desktop terminal that has no hard drive. All features
typically found on the desktop PC, including applications, sensitive data, memory, etc.,
are stored back in the server when using a thin client. These thin clients may not neces‐
sarily be a totally different hardware but can also be in the form of PCs. Thin clients,
software services, and backend hardware make up thin client computing, a remote
desktop computing model [1]. Guacamole is not a self-contained web application and
is made up of many parts. The web application is actually intended to be simple and
minimal, with the majority of the grunt work performed by lower-level components.
Users connect to a Guacamole server with their web application. The Guacamole client,
written in JavaScript, is served to users by a web server within the Guacamole server.
Once loaded, this client connects back to the server over HTTP using the Guacamole
protocol. The web application deployed to the Guacamole server reads the Guacamole
protocol and forwards it to guacd, the native Guacamole proxy. This proxy actually
interprets the contents of the Guacamole protocol, connecting to any number of remote
desktop servers on behalf of the user [2] (Fig. 1).
zamfira@unitbv.ro
192 B. Kalyan Ram et al.
stack. It actually only understands the Guacamole protocol, which is a protocol for
remote display rendering and event transport. While a protocol with those properties
would naturally have the same abilities as a remote desktop protocol, the design prin‐
ciples behind a remote desktop protocol and the Guacamole protocol are different: the
Guacamole protocol is not intended to implement the features of a specific desktop
environment. As a remote display and interaction protocol, Guacamole implements a
superset of existing remote desktop protocols [1].
Adding support for a particular remote desktop protocol (like RDP) to Guacamole
thus involves writing a middle layer which “translates” between the remote desktop
protocol and the Guacamole protocol. Implementing such a translation is no different
than implementing any native client, except that this particular implementation renders
to a remote display rather than a local one.
2.2 GUACD
• guacd is the heart of Guacamole which dynamically loads support for remote desktop
protocols (called “client plug-ins”) and connects them to remote desktops based on
instructions received from the web application.
• guacd is a daemon process which is installed along with Guacamole and runs in the
background, listening for TCP connections from the web application. guacd also does
not understand any specific remote desktop protocol, but rather implements just
enough of the Guacamole protocol to determine which protocol support needs to be
loaded and what arguments must be passed to it. Once a client plug-in is loaded, it
runs independently of guacd and has full control of the communication between itself
and the web application until the client plug-in terminates (Fig. 2).
zamfira@unitbv.ro
Remote Laboratories: For Real Time Access to Experiment Setups 193
This specific remote laboratory setup is made up of Motor - Generator setups, PLC
trainer setup and Process control trainer setup.
zamfira@unitbv.ro
194 B. Kalyan Ram et al.
4 Cassandra Database
zamfira@unitbv.ro
Remote Laboratories: For Real Time Access to Experiment Setups 195
value, Cassandra performs a read repair in the background to update the stale values.
The following figure shows a schematic view of how Cassandra uses data replication
among the nodes in a cluster to ensure no single point of failure (Fig. 4).
The first time that a user seeks access to an application, the Login Server:
– Authenticates the user by means of user name and password
– Passes the client’s identity to the various applications
– Marks the client being authenticated with an encrypted login cookie
In subsequent user logins, this login cookie provides the Login Server with the user’s
identity, and indicates that authentication has already been performed. If there is no login
cookie, then the Login Server presents the user with a login challenge. To guard against
sniffing, the Login Server can send the login cookie to the client brow er over an
encrypted SSL channel. The login cookie expires with the session, either at the end of
a time interval specified by the administrator, or when the user exits the browser. It is
never written to disk. A partner application can expire its session through its own explicit
logout.
1. Single Sign-On Application Programming Interface (API)
(a) The Single Sign-On API enables:
(i) Applications to communicate with the Login Server and to accept a user’s
identity as validated by the Login Server
(ii) Administrators to manage the application’s association to the Login Server
(b) There are two kinds of applications to which Single Sign-On provides access:
(i) Partner Applications
(ii) External Applications
zamfira@unitbv.ro
196 B. Kalyan Ram et al.
2. Partner Applications
Partner applications are integrated with the Login Server. They contain a Single Sign-
On API that enables them to accept a user’s identity as validated by the Login Server.
3. External Applications
External applications are web-based applications that retain their authentication
logic. They do not delegate authentication to the Login Server and, as such, require a
user name and password to provide access. Currently, these applications are limited to
those which employ an HTML form for accepting the user name and password. The user
name may be different from the SSO user name, and the Login Server provides the
necessary mapping (Fig. 5).
6 Port Forwarding
zamfira@unitbv.ro
Remote Laboratories: For Real Time Access to Experiment Setups 197
7 A Record
An A record maps a domain name to the IP address (IPv4) of the computer hosting the
domain. Simply put, an A record is used to find the IP address of a computer connected
to the internet from a name. The A in A record stands for Address. Whenever you visit
a web site, send an email, connect to Twitter or Facebook or do almost anything on the
Internet, the address you enter is a series of words connected with dots. For example, to
access any website you enter a URL for instance www.google.com. At the name server
there is an A record that points to the IP address 8.8.8.8. This means that a request from
your browser to www.google.com is directed to the server with IP address 8.8.8.8. A
Records are the simplest type of DNS records, yet one of the primary records used in
DNS servers [7]. You can actually do quite a bit more with A records, including using
multiple A records for the same domain in order to provide redundancy. Additionally,
multiple names could point to the same address, in which case each would have its own
A record pointing to the that same IP address.
The HTTP-based video interface provides the functionality for requesting single and
multipart images and for getting and setting internal parameter values. The image and
CGI requests are handled by the built-in web server. The mjpg/video.cgi is used to
request a Motion JPEG video stream with specified arguments. The arguments can be
zamfira@unitbv.ro
198 B. Kalyan Ram et al.
specified explicitly, or a predefined stream profile can be used. Image settings saved in
a stream profile can be overridden by specifying new settings after the stream profile
argument [8].
Deployment is the term used for the process of installing a web application (either a 3rd
party WAR or your own custom web application) into the Tomcat server. Web appli‐
cation deployment may be accomplished in a number of ways within the Tomcat server.
– Statically, the web application is setup before Tomcat is started
– Dynamically; by directly manipulating already deployed web applications (relying
on auto-deployment feature) or remotely by using the Tomcat Manager web appli‐
cation
The Tomcat Manager is a web application that can be used interactively (via HTML
GUI) or programmatically (via URL-based API) to deploy and manage web applica‐
tions. There are a number of ways to perform deployment that rely on the Manager web
application. Apache Tomcat provides tasks for Apache Ant build tool. Apache Tomcat
Maven Plug-in project provides integration with Apache Maven. The desired environ‐
ment should define a JAVA_HOME value pointing to your Java installation. Addition‐
ally, you should ensure the Java javac compiler command run from the command shell
that your operating system provides.
10 Network Architecture
10.1 Firewall
10.2 Load-Balancer
A load balancer is a device that acts as a reverse proxy and distributes network or appli‐
cation traffic across a number of servers. Load balancers are used to increase capacity
(concurrent users) and reliability of applications. They improve the overall performance
zamfira@unitbv.ro
Remote Laboratories: For Real Time Access to Experiment Setups 199
zamfira@unitbv.ro
200 B. Kalyan Ram et al.
The server machine runs on windows server 2012 and makes use of the remote desktop
service to configure and host software developed to control the hardware systems from
server machine.
Thin client is a lightweight computer that is purpose-built for remote access to a server
(typically cloud or desktop virtualization environments). It depends heavily on another
computer (its server) to fulfill its computational roles. The specific roles assumed by the
server may vary, from hosting a shared set of virtualized applications, a shared desktop
stack or virtual desktop, to data processing and file storage on the client’s or user’s behalf.
This is different from the desktop pc (fat client), which is a computer designed to take
on these roles by itself.
Thin clients occur as components of a broader computing infrastructure, where many
clients share their computations with a server or server farm. The server-side infrastruc‐
ture makes use of cloud computing software such as application virtualization, hosted
shared desktop (hsd) or desktop virtualization (vdi). This combination forms what is
known today as a cloud based system where desktop resources are centralized into one
or more data centers. The benefits of centralization are hardware resource optimization,
reduced software maintenance, and improved security.
zamfira@unitbv.ro
Remote Laboratories: For Real Time Access to Experiment Setups 201
Statistics is the study of numerical information, which is called data. People use statistics
as tools to understand information. Learning to understand statistics helps a person react
intelligently to statistical claims. Statistics are used in the fields of business, math,
economics, accounting, banking, government, astronomy, and the natural and social
sciences. Over all session statics is put in the admin portal. Where the admin will have
the privilege to check the overall user sessions, how many session are booked and
cancelled. The scheduler help to book a slot at the required time as per the user needs.
And the lab can be accessed at the particular time slot booked by the users (Fig. 7).
Fig. 7. Scheduler
zamfira@unitbv.ro
202 B. Kalyan Ram et al.
zamfira@unitbv.ro
Remote Laboratories: For Real Time Access to Experiment Setups 203
13 Conclusion
Remote labs are the natural choice for accessing physical laboratories online to enhance
the accessibility of both Software and Hardware infrastructure in Engineering colleges
[12]. In the context of India, the data shows that the utilization of Laboratory resources
is very low and the accessibility of laboratory resources to the students is sparse [13].
The topics presented in this paper addresses the technological architecture and the tools
needed for implementation of an effective Remote Lab Infrastructure from the perspec‐
tive of OS independent, Browser independent and Application independent solution.
zamfira@unitbv.ro
204 B. Kalyan Ram et al.
References
1. Wang, S.-T., Chang, H.-Y.: Development of web-based remote desktop to provide adaptive
user interfaces in cloud platform. World Acad. Sci. Eng. Technol. Int. J. Comput. Electr.
Autom. Control Inf. Eng. 8(8), 1572–1577 (2014)
2. http://guacamole.incubator.apache.org
3. Tsai, C.-Y., Huang, W.-L.: Design and performance modeling of an efficient remote
collaboration system. Int. J. Grid Distrib. Comput. 8(4) (2015)
4. Cassandra. https://www.tutorialspoint.com/cassandra/cassandra_introduction.htm
5. SSO. https://docs.oracle.com/cd/A97337_01/ias102_otn/portal.12/a86782/concepts.htm
6. Port Forwarding. https://en.wikipedia.org/wiki/Port_forwarding
7. Introduction to A-record. https://support.dnsimple.com/articles/a-record/
8. VideoAPI. http://www.axis.com/files/manuals/vapix_video_streaming5237_en_1307.pdf
9. Apache Tomcat. http://tomcat.apache.org/
10. Gallardo, A., Richter, T., Debicki, P., et al.: A rig booking system for on-line laboratories.
In: IEEE EDUCON Education Engineering– Learning Environments and Ecosystems in
Engineering Education Session T1A, p. 6 (2011)
11. Scheduler. https://www.techopedia.com/definition/25078/scheduler
12. Kalyan Ram, B., Arun Kumar, S., Mallikarjuna Sarma, B., Bhaskar, M., Chetan Kulkarni,
S.: Remote software laboratories: facilitating access to engineering softwares online. In: 13th
International Conference on Remote Engineering and Virtual Instrumentation (REV), p. 394
(2016)
13. Kalyan Ram, B., Hegde, S.R., Pruthvi, P., Hiremath, P.S., Jackson, D., Arun Kumar, S.: A
distinctive approach to enhance the utility of laboratories in Indian academia. In: 12th
International Conference on Remote Engineering and Virtual Instrumentation (REV), p. 235
(2015)
zamfira@unitbv.ro
Web Experimentation on Virtual and Remote
Laboratories
1 Introduction
Students need to understand the theoretical and practical fundamental concepts
in order to achieve a quality education in any field, hence, experimentation in
c Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6 20
zamfira@unitbv.ro
206 D. Galan et al.
traditional laboratories is essential [5]. The high costs associated with equipment,
space, and maintenance staff, impose certain constraints on resources. Virtual
and Remote laboratories (VRLs) try to overcome these limitations [4,14]. Differ-
ent empirical studies [3,19] have shown that both, VRLs and traditional labora-
tories, can obtain similar learning outcomes. Furthermore, VRLs provide inter-
esting additional advantages: they support experimentation about unobservable
phenomena and avoid health risks, such as radioactivity, chemical reactions, or
electricity [6,9].
A laboratory is meant to offer experimentation possibilities. Experimentation
can be defined as the process of extracting data from a system by exerting it, not
only through its inputs, but also through the model parameters. Traditionally,
users of VRLs were expected to perform experiments by scripting algorithms in
a certain simulation language or by interacting with the controls and buttons of
the applications graphical user interface (GUI).
Most of the modern modeling or simulation tools already include script-
ing facilities that allow users to script certain types of experiments [12,15,16].
Among them it can be found ACSL [2], EcosimPro [1], and Dymola [8]. Advanced
Continuous System Language (ACSL) was one of the first commercially available
modeling and simulation tools designed for simulating continuous systems. ACSL
includes a programming language that supports creating experiments. Dymola
also support a script facility that makes it possible to load model libraries, set
parameters, set start values, simulate and plot variables by executing scripts.
However, the major drawback is that if these tools are used to create a labora-
tory with educational purposes, final users (students mainly) will have to know
how the laboratory was implemented and handle fluently a specific programming
language just to perform any experiment.
Due to this disadvantages, most of the VRLs for educational purposes are
geared towards performing experiments by interacting with the GUI. For these
labs, visualization and interactivity are features of special importance, [11,13].
It is highly recommended the use of images or animations in order to help users
to understand more easily the system under study. Current developments in
interactivity allows users to visualize the response of the system to any external
or internal change, [10,18]. These features, rich visual contents and the possibility
of an instantaneous visualization of the system response make VRLs a human-
friendly tool to learn, helping users to achieve practical experience.
Despite all these improvements, there are certain limitations that need to be
solved. Consider, for example, a VL with a PI control of the level of water in
a tank. A typical process for an experiment in which several PIs are compared
could be:
zamfira@unitbv.ro
Web Experimentation on Virtual and Remote Laboratories 207
This set of actions cannot be executed with the accuracy needed or in reason-
able time by just interacting with the GUI. For example, pausing the simulation
in an exact moment is practically impossible. Other repetitive tasks such as tak-
ing tens of measurements to perform an analysis of the results are tedious and
provide no educational value so it is preferable not to ask for them.
Alternatively, it would be preferable to code the experiment using a flexible,
intuitive and user-friendly experimentation language, then run it automatically
and finally visualize the results or the plots. In other words, the solution will be
joining the two conceptions of how to perform experiments in a lab (interaction
and script programming).
The main goal of this work is to enrich existing VRLs with an application
that enables the creation and execution of automated experiments. To achieve
this objective, a new Application Programming Interface (API) and a set of func-
tions which VRLs should conform to provide the desired experimentation capa-
bilities, has been designed. On the basis of the general specifications obtained
from the most commonly used simulation languages, authors have added some
new requirements to achieve an universal, full-fledged specification that provides
more general and flexible features.
In order to test the viability of the proposed experimentation application,
authors’ implementation uses JavaScript labs developed with the modeling tool
Easy java(script) Simulations (EjsS), [11]. EjsS is a software tool that helps the
user with the creation of interactive simulations in Java, or JavaScript. EjsS has
been designed to be used by scientists without special programming skills, and
has proven to simplify the creation of simulations for scientific and engineering
purposes. An excellent proof of the EjsS potential is the ComPADRE repository
[7], which hosts free online resource collections, supporting students and teachers.
Among these resources, users can find more than 500 applications created with
EjsS. These labs are enriched with the capability to execute experiments, which,
in the presented approach, are scripts coded with Blockly, [17]. Blockly is an
easy and intuitive graphical programming language.
Despite the huge advantages and great utilities offered by EjsS there was
not a way to use them for allowing users to create simulation experiments. This
limitation is not restricted to EjsS; PhET simulations, [20], also available to
download for free, present the same problem too.
The paper is structured as follows. Section 2 presents the Experiment Appli-
cation and its benefits. Section 3 discusses the implementation of the language
and the blocks needed to represent experiments. Section 4 shows an example
that uses the experimentation language in practice. Finally, Sect. 5 discusses the
results and describes further work.
zamfira@unitbv.ro
208 D. Galan et al.
The first two elements (Blockly and Google Charts) that comprise the appli-
cation GUI (see Fig. 1) are explained in this section. The other two, the API
and the experimentation language are explained in Sect. 3. Notice that the lab
is not part of the application. If the lab, whether a virtual lab, a remote lab,
an hybrid lab or a simulation, implements the API proposed by the authors the
ExApp can be used.
In the first place, Blockly is the selected tool for the design of the experiments.
It is a free and open source library that adds a visual code editor to web and
Android apps. The Blockly editor uses graphical puzzle-like blocks to represent
concepts like variables, logic expressions, loops, and any element of a traditional
programming language. It allows users to apply programming principles with-
out having to worry about syntax or the laboratory structure. Blockly is used in
lots of learning applications as: Blockly Games (a set of educational games that
Fig. 1. ExApp GUI (Blockly Code and Google Charts) with a virtual lab modeling a
bouncing ball
zamfira@unitbv.ro
Web Experimentation on Virtual and Remote Laboratories 209
teach programming concepts), MIT’s App Inventor (to create applications for
Android), Code.org (to teach introductory programming to millions of students
in their Hour of Code program), Wonder Workshop (to control their Dot and
Dash educational robots), the Open Roberta project (to program Lego Mind-
storms EV3 robots), or ScratchyCAD (a web based parametric 3D modeling tool
which allows users to create 3D objects). Using Blockly to create experiments
for VRLs rather than other programming languages is a valuable asset from the
experience of the authors. As VRLs can be used by any person, with or without
programming skills, Blockly is the easiest way to start creating algorithms to
conduct experiments. Furthermore, this code editor offers interesting features
that favor the web use, maintaining the power of traditional languages (imple-
mented with JavaScript, minimal type checking supported, easy to extend with
custom blocks, localized into 50+ languages, ...).
The data analysis is provided by Google Charts, [21]. This free and open
source library is used to visualize data on a website. Google Charts provide a
large number of ready-to-use chart types. It is able to represent from simple
line charts to complex hierarchical tree maps. It is highly customizable and
supports dynamic data and controls to create interactive dashboards. It also
offers functions to import and export data to other formats. As Blockly, it is a
JavaScript library so its incorporation to an online tool is simple and clean.
1. Lab designers. They are in charge of creating the model, the view, deciding
which variables are going to be visualized in charts and adding some interac-
tive elements to control the execution of the lab by changing some variables
or internal functions. If the designer uses a tool that implements the API
proposed in this paper (EjsS, for example), he/she will not need to change a
single thing in the lab implementation in order to use ExApp. Furthermore,
the designer could focus only in the model definition and the view, charts or
any interactive element are not longer needed to control the lab. Since ExApp
has access to every variable, the final user can decide the way to work with
the lab and the data to show in the charts. This means that the time needed
to create a lab is reduced and the experiences proposed to the students are
not limited by the design.
2. Teachers. They have to define the lab experiences for the students. If the
lab is open and not restricted by the designer pretensions, the teacher will
have plenty of possibilities to propose different kind of experiments to the
students. From simple algorithms to discover the important variables of a
system, to create from the scratch a controller for the level control of a water
tank. Deploying ExApp and the lab on a web page is as simple as preparing
an HTML with the two elements. Authors’ next step is to include ExApp
as a Moodle plugin. In this way, the lab, the ExApp, the experiment files
zamfira@unitbv.ro
210 D. Galan et al.
and the results would be managed by Moodle. The correction of these type
of interactive experiments using Blockly is as easy as running the student’s
file and evaluating the results obtained. Regarding the evaluation, teachers
may give value to whether the correct result is obtained as well as how the
student reached to that solution. Teachers have the possibility to analyze the
experiments structure, to study the algorithms used and to perform the stu-
dents experiments as many times as needed just with one click. An additional
advantage is that the time needed by teachers to explain how to use the tool is
extremely low comparing with other simulation tools that allow the creation
of experiment scripts. Even more, Blockly and other similar tools as Scratch
are currently being used in elementary schools. This means near future users
will not need any extra explanation about how to used it, because students
will be familiar with these tools.
3. Students. They are the final users of the lab and ExApp. Currently, Blockly
is the first step to start learning programming skills, so even students with no
programming knowledge will find ExApp an easy tool to code their scripts.
Blockly offers visualization features as highlighting blocks which are executed
at a certain time so it is very easy to follow the execution flow of the exper-
iment and correct possible mistakes. For the same lab, students can face
different assignments depending on their skills which promote, among oth-
ers, imagination to solve the assignments, learning interest, critical thinking,
being challenged and inquiry-based learning. Also, by scripting the experi-
ment, students avoid tedious or repetitive tasks that lack of any educational
value. They are able to exchange, compare y confront experiments with teach-
ers or other students. Visualizing, collecting and analyzing results is easier
thanks to Google Charts.
3 Implementation
To achieve the objective of controlling every aspect of a VRL an interface between
ExApp and the VRL is needed. Such API should then contain the following
elements:
These elements, how they conform the experimentation language and the
way they are implemented in ExApp, are described in the following subsections.
zamfira@unitbv.ro
Web Experimentation on Virtual and Remote Laboratories 211
variables and lab functionality (in EjsS labs, the model variable). Once ExApp
and the lab are linked, all variables from the lab system are classified by type
and prepare for their use in the code. Optionally, an XML file can be configured
to show more or less Blockly blocks in order to create from the most simple
to the most complex algorithm. By default, the XML is configured with all the
possible blocks.
zamfira@unitbv.ro
212 D. Galan et al.
The API should implement different ways to control the lab execution. If it is
a lab which evolution depends on time, instructions to start, pause or stop the
lab are necessary. The In every step do block can be used to executed a code
in every step of the simulation. Also, more complex functions are needed, like
events (do something when a given condition is met). For example, “run the sim-
ulation until the level of the tank is greater than 10” or “run the simulation and
increase the set point by 50% when t = 10”. The lab should implement the func-
tion model.addEvent(conditionCode, actionCode) and model.addFixedRel(Code)
in order to allow these type of statements.
Figure 4 shows how the experimentation language implements the functions
to add code to every step and to add events to the lab. First of all, the lab is
reset and then the code, print variable z, is added to the step. After that an
event is added. The condition is 3 minus the variable t from the lab. And the
action consist in pausing the execution of the VRL. After this, the lab is started.
When the variable t equals 3, the lab will pause.
zamfira@unitbv.ro
Web Experimentation on Virtual and Remote Laboratories 213
4 Example of Use
This section presents several examples to show the usefulness of ExApp and the
advantages of it use described throughout the paper. Each of these examples
contains a brief description of the experiment, the code of the experiment, their
results and the advantages of working with the experiment editor. Experiments
have been developed for the water tank system VL. Their general features and
functionality are detailed below.
zamfira@unitbv.ro
214 D. Galan et al.
dh (Qin − Qout)
= (1)
dt A
where h represents the current tank water level and A the cross section of the
tank. The input flow, Qin, and the output flow, Qout, are given by Eqs. 2 and 3
respectively.
Qin = K1 ∗ a1 (2)
Qout = K2 ∗ a2 ∗ 2 ∗ g ∗ h (3)
where K1 represents the first valve input flow, a1 the first valve perturbation.
K2 is the second valve output flow, a2 the second valve perturbation, the gravity
is represented by g and h the current tank water level.
The virtual laboratory was developed using Easy java(script) Simulations (EjsS).
EjsS simulations are created by specifying a model to be run by the EjsS engine
and by building a view to visualize a graphical representation of the system
modeled and to interact with it. Department of Computer Science and Automatic
of UNED commonly uses EjsS simulations as virtual or remote laboratories.
The main intention of these examples is to show the ExApp power and useful-
ness, for this reason the virtual lab is as simple as possible. The GUI only shows
the tank, the in and out pipes and the level of water in the tank. Notice that
zamfira@unitbv.ro
Web Experimentation on Virtual and Remote Laboratories 215
there are no interactive controls, plots or variable indications. The lab imple-
ments the Eqs. 1, 2 and 3, consequently, the parameters and variables of this
equations are the only ones implemented in the lab.
zamfira@unitbv.ro
216 D. Galan et al.
zamfira@unitbv.ro
Web Experimentation on Virtual and Remote Laboratories 217
by the controller tuning parameters and the controller error, e(t). PI controllers
have two tuning parameters to adjust, K and Ti. PI controllers provide a bal-
ance of complexity and capability that makes them by far the most widely used
algorithm in process control applications. The PI controller implemented for this
experiment has the form shown in Eq. 6.
K
U = Ke(t) + e(t)dx (6)
Ti
Figure 8 shows the experiment script and the charts obtained at the end of the
experiment. The experiment is divided in five parts for better readability. The
first one is used to create and define the charts. The initialization part set the
initial values to prepare the lab for the experiment. The controller implementa-
tion part is the main script of the experiment. By using the “In every step do”
block the input flow is calculated using Eq. 6. Because of the limitations of the
valve some conditions are added for not allowing negative input flows and setting
a top maximum value. Events are used to change the set-point value at 50 s and
to finish the experiment at 100 s. Once everything is defined, the experiment is
executed. The chart at the left part shows the level and the set-point over time
and the one at the right part shows the input flow and the output flow over time.
5 Conclusion
To expand the experimental activities that can be carried out in virtual and
remote laboratories, a new web application and its corresponding implementa-
tion has been presented in this paper. Current alternative approaches to auto-
mate experiments require to interact and consequently to use the same code
that implements the laboratory, which implies the use of the same language in
which the lab is written. In contrast, in the authors approach, it is possible to:
access and modify all the laboratory variables, create algorithms and functions,
and control the execution of the experiment. Additionally, users can execute the
experiment step by step or run the whole script with a modifiable interval of time
between code sentences. The developed web application will be implemented as
a Moodle plugin in the near future. From the authors point of view, these type of
plugins can change the way of performing experiments, creating new experiences
in Learning Management Systems (LMS).
To illustrate the potential and ease of use of the language, an example has
been described in detail: the water tank system.
Authors are in the process of testing the initial design creating different types
of experiments of practical use in teaching Automatic Control and Physics. Ini-
tial results show that this implementation is both simple and flexible, supplying
users with a great deal of control over the running simulation. The combination of
JavaScript and Blockly has been crucial in making the proposed implementation
very natural. The way EjsS implements the VRLs allows external applications
to easily access all its variables without any required modification in the lab
applications already developed. In a more general context, authors believe the
zamfira@unitbv.ro
218 D. Galan et al.
API proposed in this work can effortlessly be adapted to different lab implemen-
tations or to any future standard protocol.
Acknowledgments. This work has been funded by the Spanish Ministry of Econ-
omy and Competitiveness under the projects EUIN2015-62577, DPI-2013-44776-R and
DPI2016-77677-P.
References
1. EA internacional ecosimpros website. http://www.ecosimpro.com/
2. MGA software ACSL reference manual, version 11 (1995)
3. Brinson, J.R.: Learning outcome achievement in non-traditional (virtual and
remote) versus traditional (hands-on) laboratories: a review of the empirical
research. Comput. Educ. 87, 218–237 (2015)
4. Brodersen, A.J., Bourne, J.R.: Virtual engineering laboratories. J. Eng. Educ. 83,
279–285 (1994)
5. Cellier, F.E., Greifeneder, J.: Continuous System Modeling. Springer, New York
(2013)
6. Chiu, J.L., DeJaegher, C.J., Chao, J.: The effects of augmented virtual science
laboratories on middle school students’ understanding of gas properties. Comput.
Educ. 85, 59–73 (2015)
7. Christian, W., Esquembre, F., Barbato, L.: Open source physics. Science
334(6059), 1077–1078 (2011)
8. Mattsson, S.E., Brck, D., Elmqvist, H., Olsson, H.: Dymola for multi engineer-
ing modeling and simulation. In: Proceedings of the 2nd International Modelica
Conference (2002)
9. de Jong, T., Linn, M.C., Zacharia, Z.C.: Physical and virtual laboratories in science
and engineering education. Science 340, 305–308 (2013)
10. Dormido, S., Dormido-Canto, S., Dormido, R., Sánchez, J., Duro, N.: The role of
interactivity in control learning. Int. J. Eng. Educ. 21(6), 1122 (2005)
11. Esquembre, F.: Adding interactivity to existing Simulink models using Easy Java
simulations. Comput. Phys. Commun. 156, 199–204 (2004)
12. Feisel, L., Peterson, G.D.: A colloquy on learning objectives for engineering edu-
cational laboratories. In: ASEE Annual Conference and Exposition, Montreal,
Ontario, Canada (2002)
13. Heck, B.S.: Future directions in control education [guest editorial]. IEEE Control
Syst. 19(5), 36–37 (1999)
14. Heradio, R., de la Torre, L., Galan, D., Cabrerizo, F.J., Herrera-Viedma, E.,
Dormido, S.: Virtual and remote labs in education: a bibliometric analysis. Com-
put. Educ. 98, 14–38 (2016)
15. Law, A.M., Kelton, W.D.: Simulation Modeling and Analysis, 2nd edn.
McGrawHill, New York (1991)
16. Law, A.M., Kelton, W.D.: Simulation Modeling and Analysis. McGrawHill,
New York (2001)
17. Marron, A., Weiss, G., Wiener, G.: A decentralized approach for programming
interactive applications with javascript and blockly. In: Proceedings of the 2nd
Edition on Programming Systems, Languages and Applications Based on Actors,
Agents, and Decentralized Control Abstractions, pp. 59–70. ACM (2012)
zamfira@unitbv.ro
Web Experimentation on Virtual and Remote Laboratories 219
18. Sánchez, J., Morilla, F., Dormido, S., Aranda, J., Ruipérez, P.: Virtual and remote
control labs using Java: a qualitative approach. IEEE Control Syst. 22(2), 8–20
(2002)
19. Sun, K.T., Lin, Y.C., Yu, C.J.: A study on learning effect among different learning
styles in a web-based lab of science for elementary school students. Comput. Educ.
50(4), 1411–1422 (2008)
20. Wieman, C.E., Adams, W.K., Perkins, K.K.: PhET: simulations that enhance
learning. Science 322(5902), 682–683 (2008)
21. Zhu, Y.: Introducing Google chart tools and google maps API in data visualization
courses. IEEE Comput. Graph. Appl. 32(6), 6 (2012)
zamfira@unitbv.ro
How to Leverage Reflection in Case of Inquiry
Learning? The Study of Awareness Tools
in the Context of Virtual and Remote
Laboratory
1 Introduction
In the context of inquiry learning that leads to knowledge building and deep
learning [11], Virtual and Remote Laboratories (VRL) gain more and more inter-
est from the research community, as the Go-Lab European project that involved
more than fifteen partners demonstrates it. However, research in this area mainly
focus on the technical and technological issues instead of emphasizing the peda-
gogical expectations to enhance learning. Yet, some research conducted around
remotely controlled track-based robots [17] showed that, among other benefits,
reflection and metacognition could emerge [21].
On the other hand, during the last decade, a significant number of researchers
studied how awareness tools could be used to promote reflection. A wide variety
c Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6 21
zamfira@unitbv.ro
How to Leverage Reflection in Case of Inquiry Learning? 221
1
http://www.openstack.org.
zamfira@unitbv.ro
222 Venant et al.
zamfira@unitbv.ro
How to Leverage Reflection in Case of Inquiry Learning? 223
zamfira@unitbv.ro
224 Venant et al.
visualization tools dealing with such data have to require very few attention to
be understood and beneficial for learners [27]. We adopted a simple color code
(i.e., green if the indicator is set to 1, red if it is set to 0) to represent, as progress
bars, learners’ performance. The tool distinguishes the learners’ level of practice
during the session within the system (i.e., since they logged in the system - see
progress bar My current session in Fig. 2), and their level of practice taking into
account the whole set of actions they carried out since they started working
on the given practical activity (i.e., not only the current session, but also all
previous sessions related to the given activity - see progress bar My practical
activity in Fig. 2). This tool also comprises a progress bar to reflect the level of
practice of the whole group of learners enrolled in the practical activity (i.e., all
the sessions of all users - see progress bar All participants in Fig. 2). Each time a
command is executed by a learner, the progress bars are automatically updated
with a coloured item (see next section). Finally, the social presence tool (see
Sect. 2.1) exposing the users currently working on the same practical activity
has been enhanced: the session level of practice of each user is displayed using a
smaller progress bar (see bottom right corner of Fig. 1).
Through the current and general progress bars, learners can get aware of the
progression of their level of practice regarding a given activity; they are also
able to compare their current level with their average level. In conjonction with
the group progress bar, learners can position themselves in relation to peers and
become more engaged in learning tasks [18]. In addition, the progress bars of
the social presence tool allow learners to identify peers that perform better, and
thus to get support from them using other awareness tools (see further).
Let us note that the indicator on which the social comparison tool stands
on, i.e., the technical rightness, is not specific to computer science. In most of
STEM disciplines, such an indicator may be captured: a given instruction is
executed (respectively not executed) by an equipment if it is (respectively is
not) technically/semantically well-formulated.
zamfira@unitbv.ro
How to Leverage Reflection in Case of Inquiry Learning? 225
zamfira@unitbv.ro
226 Venant et al.
Design and Learning Scenario. The tool features visualization and analysis
of detailed information about interactions between users and remote resources.
Users are able to consult the commands they carried out during a particular
session of work, or since the beginning of a given practical activity. The tool has
been designed to let users easily drill down into deeper and fine-grained analysis
of their work, but also to let them discover how peers have solved a given issue.
Figure 3 shows the graphical user interface of this tool: the top of the interface
exposes a form to allow users to refine the information they want to visualize,
whereas the main panel exposes the selected data. To facilitate the projection
of the information, the filtering features include the possibility to select a given
user, a particular session of work and, if applicable, one or several resources
used during the selected session. The actions matching with the selected criteria
are then exposed to users as timelines. Each node of a timeline represents a
command, and is coloured according to its technical rightness. In addition, the
details of a command can be visualized by putting the mouse over the matching
node; in that case, the date the command has been carried out, the action and
the output are displayed into the area appearing on Fig. 3.
This reflection-on-action tool allows users to browse the history of the actions
they carried out, and thus brings learners into a reflective learning situation
where they can analyze their practical work sessions in details. In addition,
zamfira@unitbv.ro
How to Leverage Reflection in Case of Inquiry Learning? 227
learners can easily focus, thanks to the coloured-coded artifact, on the difficulties
they experienced. Also, combined with the social presence tool, learners are able
to easily seek immediate help from peers by analyzing the commands executed
by users currently performing well into the system.
zamfira@unitbv.ro
228 Venant et al.
zamfira@unitbv.ro
How to Leverage Reflection in Case of Inquiry Learning? 229
The three tools presented in this section have been designed, coded and
integrated into the existing Lab4CE environment. An experimentation based on
the enhanced system has then been set up; the design, results and analysis of
this study are exposed below.
4 Experimentation
zamfira@unitbv.ro
230 Venant et al.
learning unit) with their actual experience (i.e., what they did receive from the
learning unit). In our experimentation, learners actual experience of both groups
has been compared: the control group evaluated the Linux computers, whereas
the Lab4CE group had to evaluate our system. In addition, the System Usabil-
ity Scale (SUS), recognized as a quick and reliable tool to measure how users
perceive the usability of a system [6], has been delivered to students.
The first scale (i.e., relevance) expresses the learners’ interest in the learning
unit regarding future professional practices. The Lab4CE group evaluated the
system with a slightly higher mean score and a higher concentration of scores
distribution. Since this category deals more with the topic of the learning unit
itself than the supporting environment, high differences were not expected.
The second scale relates to reflection and critical thinking. Even if the tradi-
tional environment assessed by the control group does not provide any awareness
zamfira@unitbv.ro
How to Leverage Reflection in Case of Inquiry Learning? 231
and/or reflection tools, the plots do not show a significant difference between
both groups, but slightly higher mean score and median for the Lab4CE group
only. We make here the hypothesis that learners did not realize they were engaged
in the reflection process while consulting the Lab4CE awareness tools. Indeed,
according to the system usage statistics, a mean of almost 42% of the students of
the Lab4CE group have used the reflection-on-action tool to review each of their
own sessions. On the other hand, we think that students of the control group
have considered the reflection processes occurring within the classroom instead
of considering the processes generated through the computer system only.
Feedback from both groups are quite equivalent regarding the interaction
scale which measures the extent of learners’ educative dialogue and exchange of
ideas. Here, results from the Lab4CE assessment were expected to be higher than
those returned by the control group as Lab4CE provides a chat where students
can exchange instant text messages, and a total of 166 messages have been posted
during the 3 sessions. In addition, almost 30% of the Lab4CE students have
worked at least once with a peer using the collaborative feature (see Sect. 2.1).
Again, we think that students are not aware of being involved in an interaction
task when exchanging ideas with peers.
Results about the peer support are also quite the same for both groups, even
slightly lower in the Lab4CE group. Beside our previous hypothesis that can
explain such unexpected results (here again, 47% of the Lab4CE students have
used the reflection-on-action tool), this scale reveals a potential improvement of
our platform. Learners have significantly used the reflection tools to analyze the
work done by peers, but the system does not currently provide learners with such
awareness information. The peer support scale is about the feeling of learners on
how peers encourage their participation, or praise or value their contributions.
We believe that providing students with awareness information about analysis
performed by peers on their work would increase that perception.
The last scale evaluates how messages exchanged between students, and
between students and tutors, make sense. Scores from the Lab4CE group are
characterized by a higher concentration of distribution and a little higher class
mean. These results tend to confirm that providing students with reflection tools
helps them to get a better comprehension of their interactions with each other.
In addition to the statistics commented in the previous paragraphs, interest-
ing data are the number of peers sessions analysis the day the first report had
to be submitted: almost 43% of the Lab4CE students analyzed at least one ses-
sion of a peer using the reflection-on-action tool. We assume that these learners
didn’t know how to achieve the objectives of the practical work, and thus sought
for help from peers sessions: the mean level of performance of users whose the
session has been analyzed is 90 (for a highest score of 100).
Finally, the social comparison tool which, by default, is hidden within the
user interface (see Fig. 1), has been displayed by most of users at each session
even if this rate slightly decreases when the level of performance increases. This
finding is in line with research about social comparison tools. Their impact on
cognitive and academic performance has been thoroughly examined, and main
zamfira@unitbv.ro
232 Venant et al.
results showed that informing learners of their own performance relative to others
encourages learning efforts and increases task performance [18].
System Usability Scale. The score of the SUS has been computed according
to [7]. The SUS score was 62.4 for the control group, while a SUS score of a
73.6 was attributed to the Lab4CE system. According to [2], the Linux-based
computers have been evaluated as below than acceptable systems in terms of
usability, while Lab4CE has been qualified as good regarding this criteria.
References
1. Arnold, K.E., Pistilli, M.D.: Course signals at Purdue: using learning analytics to
increase student success. In: Proceedings of the 2nd International Conference on
Learning Analytics and Knowledge, pp. 267–270. ACM (2012)
2. Bangor, A., Kortum, P., Miller, J.: Determining what individual SUS scores mean:
adding an adjective rating scale. J. Usability Stud. 4(3), 114–123 (2009)
3. Boud, D.: Situating academic development in professional work: using peer learn-
ing. Int. J. Acad. Dev. 4(1), 3–10 (1999)
4. Boud, D., Keogh, R., Walker, D.: Reflection: Turning Experience into Learning.
Routledge, New York (2013)
5. Broisin, J., Venant, R., Vidal, P.: Lab4CE: a remote laboratory for computer edu-
cation. Int. J. Artif. Intell. Educ. 25(4), 1–27 (2015)
6. Brooke, J.: SUS: a retrospective. J. Usability Stud. 8(2), 29–40 (2013)
zamfira@unitbv.ro
How to Leverage Reflection in Case of Inquiry Learning? 233
7. Brooke, J., et al.: SUS-a quick and dirty usability scale. Usability Eval. Ind.
189(194), 4–7 (1996)
8. Advanced Distributed Learning (ADL) Co-Laboratories: Experience API. https://
github.com/adlnet/xAPI-Spec/blob/master/xAPI-About.md. Accessed 21 Nov
2016
9. Collins, A., Brown, J.S.: The computer as a tool for learning through reflection.
In: Learning Issues for Intelligent Tutoring Systems, pp. 1–18. Springer, New York
(1988)
10. Davis, D., Trevisan, M., Leiffer, P., McCormack, J., Beyerlein, S., Khan, M.J.,
Brackin, P.: Reflection and metacognition in engineering practice. In: Using Reflec-
tion and Metacognition to Improve Student Learning, pp. 78–103 (2013)
11. De Jong, T., Linn, M.C., Zacharia, Z.C.: Physical and virtual laboratories in science
and engineering education. Science 340(6130), 305–308 (2013)
12. Durall, E., Leinonen, T.: Feeler: supporting awareness and reflection about learn-
ing through EEG data. In: The 5th Workshop on Awareness and Reflection in
Technology Enhanced Learning, pp. 67–73 (2015)
13. Edwards, S.H.: Using software testing to move students from trial-and-error to
reflection-in-action. ACM SIGCSE Bull. 36(1), 26–30 (2004)
14. Govaerts, S., Verbert, K., Klerkx, J., Duval, E.: Visualizing activities for self-
reflection and awareness. In: International Conference on Web-Based Learning,
pp. 91–100. Springer, Heidelberg (2010)
15. Howlin, C., Lynch, D.: Learning and academic analytics in the realizeit system. In:
E-Learn: World Conference on E-Learning in Corporate, Government, Healthcare,
and Higher Education, pp. 862–872 (2014)
16. Jonassen, D.H.: Instructional design theories and models: a new paradigm of
instructional theory. Des. Constr. Learn. Environ. 2, 215–239 (1999)
17. Kist, A.A., Maxwell, A., Gibbings, P., Fogarty, R., Midgley, W., Noble, K.: Engi-
neering for primary school children: learning with robots in a remote access lab-
oratory. In: The 39th SEFI Annual Conference: Global Engineering Recognition,
Sustainability and Mobility (2011)
18. Kollöffel, B., de Jong, T.: Can performance feedback during instruction boost
knowledge acquisition? Contrasting criterion-based and social comparison feed-
back. Interact. Learn. Environ. 24(7), 1–11 (2015)
19. Lowe, D., Murray, S., Lindsay, E., Liu, D.: Evolving remote laboratory architectures
to leverage emerging internet technologies. IEEE Trans. Learn. Technol. 2(4), 289–
294 (2009)
20. Matthew, C.T., Sternberg, R.J.: Developing experience-based (tacit) knowledge
through reflection. Learn. Individ. Differ. 19(4), 530–540 (2009)
21. Maxwell, A., Fogarty, R., Gibbings, P., Noble, K., Kist, A.A., Midgley, W.: Robot
RAL-ly international-promoting stem in elementary school across international
boundaries using remote access technology. In: The 10th International Conference
on Remote Engineering and Virtual Instrumentation, pp. 1–5. IEEE (2013)
22. Michinov, N., Primois, C.: Improving productivity and creativity in online groups
through social comparison process: new evidence for asynchronous electronic brain-
storming. Comput. Hum. Behav. 21(1), 11–28 (2005)
23. Miller, T.: Formative computer-based assessment in higher education: the effective-
ness of feedback in supporting student learning. Assess. Eval. High. Educ. 34(2),
181–192 (2009)
24. Prensky, M.: Khan academy. Educ. Technol. 51(5), 64 (2011)
25. Schön, D.A.: The Reflective Practitioner: How Professionals Think in Action. Basic
Books, New York (1983)
zamfira@unitbv.ro
234 Venant et al.
26. Seibert, K.W.: Reflection-in-action: tools for cultivating on-the-job learning condi-
tions. Org. Dyn. 27(3), 54–65 (2000)
27. Sweller, J.: Cognitive load theory, learning difficulty, and instructional design.
Learn. Instr. 4(4), 295–312 (1994)
28. Taylor, P., Maor, D.: Assessing the efficacy of online teaching with the construc-
tivist online learning environment survey. In: The 9th Annual Teaching Learning
Forum, p. 7 (2000)
29. Venant, R., Vidal, P., Broisin, J.: Evaluation of learner performance during prac-
tical activities: an experimentation in computer education. In: The 14th Interna-
tional Conference on Advanced Learning Technologies, ICALT, pp. 237–241. IEEE
(2016)
30. Wilson, J., Jan, L.W.: Smart Thinking: Developing Reflection and Metacognition.
Curriculum Press, Carlton (2008)
zamfira@unitbv.ro
Role of Wi-Fi Data Loggers in Remote Labs
Ecosystem
Abstract. All data are important and useful but what is more important is the
way this data is used. Wi-Fi Data-logger is a major step towards making use of
data for effective management of a remote lab. The purpose is to build a
real-time data-logger with Wi-Fi capabilities to remotely monitor the equipment
status and environmental conditions inside a remote lab containing high-end
electrical and electronic machinery. This device should be adaptive, flexible,
easy to use and should give deterministic results to take action.
The structure of Wi-Fi data logger consists of two zones: (a) de-
vice-level-hardware zone and (b) server-level-software zone.
(a) A micro-controller is connected to various sensors such as Temperature,
Humidity, Gas, motion sensors and to fault testing lines of the equipment and
peripherals. The data is continuously obtained in real time is pumped through
Wi-Fi over TCP/IP or UDP protocols to a server computer.
(b) It consists of a simple program running on the server computer to receive
the data from micro-controller through Wi-Fi and organize it. This program
also has a script running which throws up possible a warning in case of
malfunctioning and possible solution with step-wise instructions is
displayed.
Key Outcomes include: (a) Seamless integration of the device with the
existing machinery requiring minimal effort (b) Protection to components
(c) Over 40% reduction in the time required to detect and fix an issue achieved
by impeccable synchronous effort of device and software.
Thus, these Wi-Fi data-loggers enhance the way remote labs operate by
taking care of safety issues and increasing the stability of the whole remote labs
architecture. This technology can pave way for more complex architecture of
remote labs and the evolution of Wi-Fi data-logger technology will result in
evolution of remote labs.
1 Introduction
Remote labs are undoubtedly the future of laboratory education as they provide
opportunities for effective and holistic learning for students and researchers with
limited access to laboratories by providing them with that extra flexibility and time
required to complete that experiment or make that breakthrough or pull out an amazing
research paper like this one [1]. Remote labs have been increasing in numbers with
advanced technology day by day with new labs set up in the places spanning all
domains [2]. Remote labs serve as a bridge between virtual and real labs as well as
serving as they can be used not only in the field of education, but also for doing any
measurement-task with real laboratory instruments [3].
The general architecture of a remote lab consists of an experiment set-up inside a
room whose structure includes a computer with hardware connected to it, webcam,
microphone, feedback of experimental results back to the computer [4]. More impor-
tantly everything inside connected to the outside world through the internet or the
intranet accessed through a webserver as shown in Fig. 1.
Data loggers are small devices with capability to accumulate mass data through
acquisition cards and store them in memory before dumping to mass storage device.
Data loggers can be made more purposeful with the advancement of technology and
their adaptive implementation can lead to collection of critical data which can be of
immense importance [5]. However, data loggers are generally overlooked by most
researchers due to their simplicity and deemed by industrialists to be an extra feature
rather a required feature. We wish to change that by showing the important role that
data loggers play in remote labs ecosystem and their huge influence and boost to the
system.
zamfira@unitbv.ro
Role of Wi-Fi Data Loggers in Remote Labs Ecosystem 237
2 Approach
In this section, the build of the data logger whose description has been told towards the
end of introduction is discussed in depth. Figure 2 shows the architecture of the data
logger in general with two branches of its build – (a) Hardware level and (b) Software
Side. Each of the two branches will be dealt in detail in this section and the whole
approach to build a data logger and implementing it in the remote lab ecosystem will be
discussed.
Choosing the Laboratory. The key to choosing a laboratory is by analyzing its down
time and checking for the feasibility of installing the data logger with minimal cost and
infrastructure changes. Our primary criteria to choose a laboratory are that it should
have a sufficiently high down time and low installation cost.
For this, different laboratories were identified and their downtimes (Down Time Per-
centage = Total Downtime/Total time) as well as approximate cost factor (Cost of
installation of data logger/Cost of remote lab set up). Based on this data, the best of the
lot was chosen by Factor of Decision which is weighted sum of the DTP and Cost
Factor together according to their proportionality as shown in Fig. 3.
Factor of decision is 25% when downtime is 25% and cost factor is 25%. Lesser the
cost and more the downtime, Factor of decision goes up. So, for a cost-effective
installation should be greater than 25% and for a more effective purpose, above 35%
was chosen. As shown in the Fig. 3. Five out of the eight have a Factor above 35% and
these laboratories were first choice for data logger installation. This understanding
gives us a clear idea of laboratories and a head start by helping us the best ones on
which the data logger can have its most impact on.
zamfira@unitbv.ro
238 V.V. Gowripeddi et al.
Identifying the Failures. Failures can be due to varied reasons ranging from simple
overheating to equipment malfunction. In this section, some of the failures across the
whole remote labs ecosystem are listed out.
1. Sudden Variation in power can cause the system to fail.
2. Faulty machine lines can lead to malfunctioning.
3. Failure of temperature maintenance system can cause severe damage to compo-
nents due to overheating or overcooling.
4. Increased humidity and water deposition might brick the system.
5. Poor maintenance of infrastructure is an important cause of damage.
6. Use of components or hardware products which are not rated sufficiently high for
parameters like current, voltage, temperature can cause burning out of the com-
ponents resulting in a major failure.
7. Other failures can be attributed to rapid, violent, and unexpected changes that can
occur.
8. Loose connections can also be an issue.
9. Mechanical stress between components can lead to damage of critical moving
parts.
10. Error in software code that can put the system in infinite loop.
11. If proper security measures are not in place it can lead to misuse.
12. Overflow of memory can hang the system.
zamfira@unitbv.ro
Role of Wi-Fi Data Loggers in Remote Labs Ecosystem 239
Identifying the Suitable Hardware. It is important to choose the hardware such that
it transcends across most types of laboratories and laboratory equipment and the only
change that would be required will be change in the implementation of code and
connections (Table 1).
As you can see in the given table, Type B is adaptable to more labs than Type A, so
Type B is preferred over Type A. Similarly, Type C is preferred over Type D. Even if
cost of Type B is slightly greater than that of Type, it is worthy of using as it saves up
on cost of spares [6]. Figure 4 shows an example of a Wi-Fi based microcontroller.1
1
The mentioned microcontroller was used for microcontroller lab and the product contains a Cypress
WICEDTM chip. It is sold by Adafruit Industries based in New York.
zamfira@unitbv.ro
240 V.V. Gowripeddi et al.
Adding Components to Board. This is a 3-stage process where in the components are
put in a circuit on a breadboard to test their working. This is illustrated by the Fig. 5.
Then the components are soldered manually onto the PCB board and made as shown.
Fig. 5. Adding components to board: (a) Testing the hardware design on breadboard
(b) Soldering the components onto a circuit (c) Design of PCB and production
Encapsulating and Enclosing the Hardware Platform. This step involves packing
the whole hardware side in a high-grade case by making required openings for Inputs
and outputs through the device. Figure 6(a) and (b) clearly describe the packaging and
its features.
zamfira@unitbv.ro
Role of Wi-Fi Data Loggers in Remote Labs Ecosystem 241
Fig. 6. (a) Hardware packaged in a IP60 box (b) Openings through the case are well sealed.
Choosing the Right Software for Hardware Side as Well as Client Side. This step
involves choosing the software that is most suited to embedded programming [7] and
client side software application [8].
Arduino is an open-source electronics platform based on easy-to-use hardware and
software.
LabVIEW is an integrated development environment designed specifically for
engineers and scientists building measurement and control systems.
Arduino is chosen for:
• Simplicity
• Strong Hardware – Software interaction
• Code at an Embedded C level
• Open Source and a huge Community for support
• Large database of libraries and binaries
LabVIEW is chosen for:
• Excellent Design in form of front panel and block diagram
• Built in Libraries and tools
• Precision measurement reading
• Highly Customizable
Programming the Hardware. Arduino IDE was used to program the microcontroller
by embedding C code onto the device. Figure 7 illustrate how the same device can be
adapted to read different parameters which makes the device universal and adaptive.
Programming the Client Side. Client side programming is done through LabVIEW,
various loops, conditions are designed. All the conditions, restrictions and boundary
conditions are set in the LabVIEW block diagram with indicators and user output data
on Front Panel. Figure 8 illustrates LabVIEW programming logic done on block
diagram window of LabVIEW software.
zamfira@unitbv.ro
242 V.V. Gowripeddi et al.
Fig. 7. (a), (b) and (c) show how with a few lines of modification in the code different
parameters can be read.
zamfira@unitbv.ro
Role of Wi-Fi Data Loggers in Remote Labs Ecosystem 243
3 Working
3.1 Different Cases of Operation
Following set of Figures illustrate how monitoring front panel looks like in different
cases of operation.
Figure 9 shows a typical remote lab monitoring screen where everything looks
okay. Temperature and Humidity are under control. Fault lines are off and a message is
displayed that indicating the same.
Figure 10 shows a warning state of operation where the temperature is higher than
usual but temperature is not high enough to cause a damage to components, message is
displayed for operator indicating the same and instructions are provided for operator to
sort this error in simple terms.
zamfira@unitbv.ro
244 V.V. Gowripeddi et al.
Figure 11 illustrate the error state of operation where the lines are faulty and red
light indicates the machine has stopped running. Message is displayed indicating the
same and a solution is provided for the operator. Since, this condition may require
expertise, notification using auto alerting is sent to concerned personnel.
zamfira@unitbv.ro
Role of Wi-Fi Data Loggers in Remote Labs Ecosystem 245
4 Outcomes
Fig. 13. Shows how cost of datalogger and its installation fares against total remote lab cost for
different kinds of labs.
zamfira@unitbv.ro
246 V.V. Gowripeddi et al.
Fig. 14. Compares the time required to detect a failure and correct it with datalogger (in red) vs.
Without datalogger (in blue)
zamfira@unitbv.ro
Role of Wi-Fi Data Loggers in Remote Labs Ecosystem 247
Fig. 15. Downtime Comparison with and without dataloggers for different labs
Fig. 16. Compares cost of data logger, yearly savings due to datalogger and savings made over
five years
zamfira@unitbv.ro
248 V.V. Gowripeddi et al.
5 Conclusion
The discussion of this paper started with importance of Remote Labs in current context
and need for Wi-Fi Datalogger for efficient functioning of remote labs was well
established. Approach to build the data logger was discussed from scratch. The
foundation of Datalogger from choosing the lab to choosing components and hardware
was discussed. The building of datalogger in hardware, software and server aspect is
well illustrate in Sect. 2. Working of the datalogger, with live screens from different
labs and different states of operation is shown in working section. Data logger was
judged on the parameters of integration costs and effort, time to detect a failure and
rectify it, efficiency and finally cost perspective. The results clearly prove the effec-
tiveness of data logger. The importance of data loggers in remote labs ecosystem is
well established through this paper.
Acknowledgment. The authors wish to extend thanks to various universities and industries
across India and across the world for providing with opportunities to test the datalogger archi-
tecture and make findings.
References
1. Auer, M.E.: Virtual lab versus remote lab. In: 20th World Conference on Open Learning and
Distance Education (2001)
2. Ram, B.K., Kumar, S.A., Sarma, B.M., Mahesh, B., Kulkarni, C.S.: Remote software
laboratories: facilitating access to engineering softwares online. In: 2016 13th International
Conference on Remote Engineering and Virtual Instrumentation (REV), pp. 409–413. IEEE,
February 2016
3. Pruthvi, P., Jackson, D., Hegde, S.R., Hiremath, P.S., Kumar, S.A.: A distinctive approach to
enhance the utility of laboratories in Indian academia. In: 2015 12th International
Conference on Remote Engineering and Virtual Instrumentation (REV), pp. 238–241. IEEE,
February 2015
4. Esche, S.K., Chassapis, C., Nazalewicz, J.W., Hromin, D.J.: An architecture for multi-user
remote laboratories, dynamics (with a typical class size of 20 students), 5, 6 (2003)
5. Outram, J.D., Outram, R.G.: Adaptive data logger. U.S. Patent No. 4,910,692, 20 March
1990
6. Yunlong, F., Fang, A., Li, N.: Cortex-M0 processor: an initial survey. Microcontrollers
Embed. Syst. 6, 33 (2010)
7. D’Ausilio, A.: Arduino: a low-cost multipurpose lab equipment. Behav. Res. Methods 44(2),
305–313 (2012)
8. Gontean, A., Szabó, R., Lie, I.: LabVIEW powered remote lab. In: 2009 15th International
Symposium for Design and Technology of Electronics Packages (SIITME). IEEE (2009)
9. Auer, M., Pester, A., Ursutiu, D., Samoila, C.: Distributed virtual and remote labs in
engineering. In: 2003 IEEE International Conference on Industrial Technology, vol. 2,
pp. 1208–1213. IEEE, December 2003
10. Aloni, E., Arev, A.: System and method for notification of an event. U.S. Patent
No. 6,965,917, 15 November 2005
zamfira@unitbv.ro
Role of Wi-Fi Data Loggers in Remote Labs Ecosystem 249
11. Shnayder, V., Hempstead, M., Chen, B.R., Allen, G.W., Welsh, M.: Simulating the power
consumption of large-scale sensor network applications. In: Proceedings of the 2nd
International Conference on Embedded Networked Sensor Systems, pp. 188–200. ACM,
November 2004
12. Tinga, T.: Application of physical failure models to enable usage and load based
maintenance. Reliab. Eng. Syst. Saf. 95(10), 1061–1075 (2010)
13. Vuletid, M., Pozzi, L., Ienne, P.: Seamless hardware-software integration in reconfigurable
computing systems. IEEE Des. Test Comput. 22(2), 102–113 (2005)
14. Robinson, R.: Cost-effectiveness analysis. BMJ 307(6907), 793–795 (1993)
15. Tanner, M., Eckel, R., Senevirathne, I.: Enhanced low current, voltage, and power
dissipation measurements via Arduino Uno microcontroller with modified commercially
available sensors. APS March Meeting Abstracts (2016)
zamfira@unitbv.ro
Flipping the Remote Lab with Low Cost Rapid
Prototyping Technologies
Abstract. This work proposes the idea of flipping the remote lab.
A flipped remote lab would consist on requesting students to build a
remotely accessible experiment, so that teachers would test the lab in
order to evaluate it, instead of creating it themselves. Building a remote
lab is a multidisciplinary activity that involves using different skills and
which promotes long-life learning and creativity. Also, by assigning this
task to work in groups, students would also build up abilities such as
teamwork, communication and leadership. Because creating a remote
lab is a complex task, the idea is to use the experience acquired dur-
ing many years of development and use of virtual and remote labs for
teaching engineering and physics, to simplify the process and make it
manageable for students. Given the current state of the technology, pro-
viding students with some guidelines and reference designs should be
enough to make feasible for them to develop a remote experiment.
1 Introduction
zamfira@unitbv.ro
Flipping the Remote Lab with Low Cost Rapid Prototyping Technologies 251
engineering and physics, to simplify the process and make it manageable for
students. Given the current state of the technology, providing students with
some guidelines and reference designs should be enough to make feasible for
them to develop a remote experiment.
Recently, low cost single board computers such as Raspberry Pi or Beagle-
bone Black, and 3D printing technologies, which allow for rapid prototyping of
mechanical systems, have become pervasive. These tools provide an interesting
framework that can assist the creation of remote labs. The hardware framework
is complemented with reusable software components, a web-based architecture,
and standard communication protocols to reduce the development costs and
efforts. Based on this paradigm, an easily replicable remote lab architecture is
proposed, using 3D printed parts designs that have been open source licensed to
allow for free use or modification, as well as software components that implement
the different subsystems of the lab, and elements that either can be gathered from
old electronics devices or are cheap and commonly used components.
Some examples can be found in the literature about flipped classroom
adapted to blended learning [5,8], and engineering subjects [7], as well as of
laboratories based on single-board low-cost platform either hands-on [3] and
remote [1,4,6].
Currently, a virtual and remote lab of an air flow levitation system has been
built using the proposed methodology. It consists of a small object that has to
be lifted using the air flow generated by a fan inside a cylinder. The position
of the levitating object is measured with an infrared distance sensor and it is
used to control the rotation speed of the fan. The prototype will be incorporated
into a master degree course in control engineering. Some of the benefits expected
from the experience are to provide students with a global insight of engineering
processes, increase their motivation to research about different sensing technolo-
gies, and promote their creativity.
2 Approach
Since the idea is to let students build their own systems, the remote lab design
has to be well thought and structured. There are a few requirements the lab
should verify to be realizable:
If these requisites are met, not only students could be able to build the lab
as part of an assigned work, but maybe even those who would want to have
their own experimentation platform can afford to construct one on their own.
The lab should use open-source technologies, mainly for reducing costs in order
zamfira@unitbv.ro
252 J. Chacón et al.
to meet the first requirement, but also because this approach encourages to
acquire knowledge by tinkering with the system design, propose modifications
or enhancements, and so on. There are some other aspects that have been con-
sidered in order to keep the costs of building the remote lab low. The first one is
to use materials that are easy to obtain and have a reasonable cost. For exam-
ple, the IR sensor and single-board computer are cheap and can be bought in
virtually any electronic component shop. Also, components should be reused
whenever it is possible. The fans can be easily gathered from an old PC or other
electronic devices. Taking benefit from the boom rapid prototyping technologies
is also important: 3D printing allows to reduce greatly the cost of mechanic pro-
totyping, and it is relatively easy to have access to a 3D printer, either at the
university, specialized shop or online services which print your designs.
The design of the laboratory can be decomposed into several tasks, some of
which have to be done by the educator, and others that have to be prepared to
be assigned to students. The tasks that corresponds to educators are:
The design and construction of the plant is a thorough engineering process from
which students can benefits, acquiring a learn-by-doing understanding of how to
convert an idea or a concept to a practical solution.
In the next paragraphs, a (not exhaustive) review is provided of the hard-
ware platforms that are available to develop electronic systems. It is followed
by a review of some open source cad tools to model structural components and
electronic circuits, and finally it will be discussed the architecture followed by
our previous designs, which can be used as a reference (but not the only and
definitive solution) for future labs.
zamfira@unitbv.ro
Flipping the Remote Lab with Low Cost Rapid Prototyping Technologies 253
Hardware. Since the release of the first Raspberry Pi model, a bunch of single
board computers have appeared intending to fit developers needs, which range
from small DIY projects such as home media centers or domotic appliances,
to high performance research computing. Most of these boards are specifically
focused on the maker community, students and educators, so they are fully open-
source hardware (Fig. 1).
Fig. 1. Screenshots of two popular open-source CAD software tools, (a) FreeCAD and
(b) OpenSCAD
Software. CAD tools assist the designer to model the physical components
which will be part of the system, in our case the structural parts and the elec-
tronics circuits. It is out of the scope of this work to discuss the pros and cons of
the so many options available. However, it is worth to mention at least some of
the most popular open-source alternatives that cover the lab needs: FreeCAD,
OpenSCAD, KiCAD.
FreeCAD is an open-source 3D CAD software tool very popular among the
3D printing community. It has many features, parametric design, multiplatform
(works on Linux, Windows and Mac), a fully customizable GUI, and native
support for python scripting and extensions.
OpenSCAD is another popular tool, mostly used to design 3D printed parts.
Unlike FreeCAD, it uses a non-graphical with a different modelling approach.
It is based on a specific description language, so the creation process is more
similar to traditional programming. One of the advantages of this approach is
the flexibility to parameterize designs.
zamfira@unitbv.ro
254 J. Chacón et al.
The electronic circuits and the PCBs has been created with the software
KiCaD, a multiplatform and open-source tool that have the support of the
CERN, which started the KiCad project and have made important contribu-
tions to it as part of the Open Hardware Initiative (OHI)1 .
As in the case of 3D printing, there are many PCB manufacturers where you
can send your circuit design and have your PCB with professional quality and a
moderate cost or, following the maker paradigm, you can build your own circuit
with a CNC PCB milling machine or a chemical etching process.
At the end of this stage, the incomes are the structural parts and electronic
circuits needed to construct the plant.
Datalogging. Once the values have been acquired, it is needed to store them
in order to be accessed whenever be required. For that purpose there are many
options, but again it is recommended to use a standard solution. There are time
series database systems (TSDB) that are specialized on time series management,
such as InfluxDB, graphite, OpenTSDB or RRDtool.
Control. The remote labs have a local controller implemented, which can be as
simple or as sophisticated as needed. In the case of a control engineering lab, it
must be a central part of the design, but even in other cases it is always needed
to take some safety measures, to assure that the system cannot be harmed by
accident or by a malicious user.
1
https://home.cern/about/updates/2015/02/kicad-software-gets-cern-treatment.
zamfira@unitbv.ro
Flipping the Remote Lab with Low Cost Rapid Prototyping Technologies 255
Printed Parts. Most structural elements have been printed in a Prusa Mendel
i3 3D printer, a very popular and affordable RepRap printer, available at the
authors’ department. The 3D parts has been modeled with FreeCAD.
zamfira@unitbv.ro
256 J. Chacón et al.
The GUI. The interface design is clean and simple, sharing the same layout
with the virtual lab: there is a view of the system on the left, which is obtained
from the laboratory webcam (the equivalent to the 3D visualization in the vir-
tual lab), some plots on the right showing the time evolution of the interesting
variables (the height of the lifting object measured by the IR sensor, the set-
point and the control signal sent to the fan). Finally, at the bottom there is a
control panel which allows to modify some system parameters, as the controller
gains or the setpoint, and the connection buttons which are analogous to the
simulation execution control ones in the virtual lab. Figure 2 shows the remote
lab web interface, designed with EjsS and the RIP Model Element (an add-on)
which enhances EjsS with RIP interconnection capabilities.
zamfira@unitbv.ro
Flipping the Remote Lab with Low Cost Rapid Prototyping Technologies 257
4 Conclusions
In recent times, it is not unusual that students, even of first courses of engi-
neering, have at least basic knowledge of the mentioned development platforms,
and a good predisposition to use them. In spite of that, the popularity of 3D
printing technologies and the do-it-yourself (DIY) and the maker community
can be an attractive way of drawing the students’ attention. As an example of a
similar approach, some universities already proposed robotic competitions where
students are asked to solve some problems using basic construction kits. These
activities, which have demonstrated to benefit students’ development, are not
very different in nature compared to the one proposed in this work. Therefore,
it is expected to obtain great profit from the flipped remote lab.
References
1. Balula, S., Henriques, R., Fortunato, J., Pereira, T., Borges, H., Amarante-Segundo,
G., Fernandes, H.: Distributed e-lab setup based on the Raspberry Pi: the hydro-
static experiment case study. In: 2015 3rd Experiment@ International Conference
(exp.at 2015), pp. 282–285 (2015)
2. Chacón, J., Farias, G., Vargas, H., Visioli, A., Dormido, S.: Remote interoperability
protocol: a bridge between interactive interfaces and engineering systems. IFAC-
PapersOnLine 48(29), 247–252 (2015)
3. Krauss, R.: Combining Raspberry Pi and Arduino to form a low-cost, real-time
autonomous vehicle platform. In: 2016 American Control Conference (ACC), pp.
6628–6633, July 2016
4. Michels, L.B., Gruber, V., Schaeffer, L., Marcelino, R., da Silva, J.B., de Resende
Guerra, S.: Using remote experimentation for study on engineering concepts through
a didactic press. In: 2013 2nd Experiment@ International Conference (exp.at 2013),
pp. 209–211, September 2013
5. Shi, J., Yuan, S., Zou, Q.: From practice to experiment: Development and enlight-
enment of flipped classroom in China. In: 2016 International Symposium on Edu-
cational Technology (ISET), pp. 94–98, July 2016
6. Simão, J.P.S., Lima, J.P.C., Heck, C., Coelho, K., Carlos, L.M., Bilessimo, S.M.S.,
Silva, J.B.: A remote lab for teaching mechanics. In: 2016 13th International Con-
ference on Remote Engineering and Virtual Instrumentation (REV), pp. 176–182,
February 2016
7. Toner, N.L., King, G.B.: Restructuring an undergraduate mechatronic systems cur-
riculum around the flipped classroom, projects, labview, and the myrio. In: 2016
American Control Conference (ACC), pp. 7308–7314, July 2016
8. Zhang, H., Meng, L., Han, X., Yuan, L., Wang, J.: Exploration and practice of
blended learning in HVAC course based on flipped classroom. In: 2016 International
Symposium on Educational Technology (ISET), pp. 84–88, July 2016
zamfira@unitbv.ro
Remote Experimentation with Massively
Scalable Online Laboratories
1 Introduction
Massively Open Online Courses (MOOC) have the potential to reach vast audi-
ences both in geographic and socioeconomic scope. Currently, many universities,
including Stanford and MIT, use online coursework to augment educational pro-
grams for their students, provide professional programs for a fee, and offer video
lectures as MOOCs to the general public. These universities use online course-
work as an augmentation of physical classrooms in a flipped classroom approach,
where students study online education materials to enable increased interaction
between students and teachers. This enhances educational efficiency and depth of
learning. Moreover, some professional certificate programs, such as Udacity, focus
on online coursework to reach their students. All of these concepts rely on the abil-
ity of online coursework to be a scalable and effective means of education [1].
Although current techniques of video streaming allow users to easily view
lectures online, and, in some cases talk to advisers or teaching assistants via video
call, there currently is not any means of including experiments into an online
coursework environment. However, experiments, which normally take place in a
laboratory environment, with severe time and cost restrictions, are a crucial part
c Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6 24
zamfira@unitbv.ro
Remote Experimentation with Massively Scalable Online Laboratories 259
zamfira@unitbv.ro
260 L.T. Neustock et al.
it into a data-set, and subsequently into a MSOL, the platform provides access
to an otherwise economically-restricted, advanced laboratory experience. More-
over, since the laboratory is provided over a web-page, the barrier of entrance is
low. It can be accessed from all over the world and has very low acquisition costs
for the students and educational institutes. Each additional student requires only
enough resources to respond to their web requests. Therefore, it can be used in
resource poor areas when laboratory equipment is unavailable, or it can aug-
ment existing remote or in-class education. Additionally, users will be able to
access experiments and instruments that would otherwise require extensive prior
training, are dangerous, very expensive or not available.
Moreover, the entrance barrier to create a laboratory by turning an existing
experiment into a data set is very low as well. It only needs to be done once and
can easily be archived by a simple computer program. Today, most experiments
are already run by a computer, which means that only one layer of automatic
sweeping through possible permutations needs to be added. The MSOL platform
provides the required automation tools and storage facilities.
To encourage team building and peer to peer learning, interactive social fea-
tures, as described in Sect. 3, are added to the website. This design, alongside the
low entrance barriers, allows the creation of a large online community composed
of small sub-communities that encourage “Purposeful Conversation”.
3 Implementation of MSOL
3.1 Turning an Experiment into a Data-Set
The first step in turning an existing experiment into a MSOL is to record it in
all possible stages and save the corresponding information, such as values from
sensors or images of the experiment. Most modern experiments are already con-
figured to be computer controlled. The computer controls allow for repeatability
and accuracy in a research setting. This computer control also allows a program
to iterate through all possible states of all controls automatically with only very
little extra effort. If the sensor data and associated images are recorded with
each state in an automatic sweep, then this data is all that is required to cre-
ate a virtual experiment compatible with the MSOL interface. The majority of
relevant experiments can be turned into data in this fashion.
With decreasing storage costs and increased internet bandwidth, it is rea-
sonable to store more than 105 images per experiment, which provide the view
of the experiment at each permutation as if the observer was in the room. The
images can then be stored on a local hard drive, uploaded on a server, and simul-
taneously accessed by thousands of different users. Reviewing data just as a list
would be a very tedious task; therefore, the platform provides an interactive
interface that only shows relevant pictures and data, given the control states
that the user is interested in. This reduces the required bandwidth and increases
the scalability.
In general, before running the automation, the number of planned permuta-
tions should be considered and evaluated concerning its feasibility. Yet, in most
zamfira@unitbv.ro
Remote Experimentation with Massively Scalable Online Laboratories 261
Sample Case: For the purposes of this paper, we chose to demonstrate the
functionality of the MSOL platform with a diffraction experiment. This can
be found on the current online version of the MSOL platform at http://www.
ilabs.education/. Diffraction at a grating is a fundamental concept in optics,
by which the wave nature of light can be explored. The diffraction experi-
ment used here includes two different lasers and three different grating spacings.
A photo-detector that can be moved along the diffraction pattern is utilized as
an indicator, displaying the varying optical intensity due to the diffracted laser
light. The uploaded experiment also contains a light switch. A picture of the
setup can be seen in Fig. 1. The recording was done with the help of a simple
python script iterating over all permutations, recording 24,000 different data
Fig. 1. Experimental setup of the diffraction experiment: (a) Sketch of the setup
(b) Photo of the lab
zamfira@unitbv.ro
262 L.T. Neustock et al.
points with pictures. This data is uploaded through an upload interface. This
diffraction experiment will function as an example through the rest of this paper.
Fig. 2. List of experiment that a user can conduct with the ability to search for a
particular topic, title, or author.
zamfira@unitbv.ro
Remote Experimentation with Massively Scalable Online Laboratories 263
Fig. 3. Display of the laboratory experiment in the MSOL platform. The experiment
is visible along with the controls to provide input. The indicator overlay is visible.
Fig. 4. (a) Overlay at the beginning of the lab, which gives a first overview and basic
information (b) Part of the tutorial guiding through the functionalities of the platform
zamfira@unitbv.ro
264 L.T. Neustock et al.
Fig. 5. The MSOL platform with visible functions overlay while creating a meeting to
collaborate on the same experiment with several users.
either via a URL or other buttons for posting the experiment on Facebook or
Twitter. Additionally, the interface allows the user to comment on the lab using
his/her Facebook account; sharing their excitement, giving suggestions, or asking
questions to a broad audience. In our example, they could ask about the physics
behind diffraction, share their findings or simply express their excitement.
Thirdly, while conducting the experiment, the user is able to record the data
of the experimental stages he or she is going through in a personalized lab book,
if the record option is selected. While recording, the indicator data points are
automatically added to a text field in the lab book interface, accessible via the
functions overlay. This data can subsequently be downloaded as a .csv-file and
be used for creating plots. This is similar to how an actual experiment would be
used as well. The lab book interface is displayed in Fig. 6. Thus, for our example,
Fig. 6. Overlay with recorded data, showing indicator values for several control set-
tings.
zamfira@unitbv.ro
Remote Experimentation with Massively Scalable Online Laboratories 265
people will be able to record and compare the sensor data of optical intensity
for different diffraction gratings, or laser wavelengths.
In summary, the MSOL platform as described in these sections, is able to
accurately recreate the experience of the laboratory by providing an interactive
input and response system. In addition, social features enhance its usability in
an online learning environment.
References
1. Dalgarno, B., et al.: Effectiveness of a virtual laboratory as a preparatory resource
for distance education chemistry students. Comput. Educ. 53, 853–865 (2009)
2. Hesselink, L., et al.: Stanford cyber lab: internet assisted laboratories. Int. J. Dis-
tance Educ. Technol. 1(1), 22–39 (2003). Chang, S.-K., Shih, T.K. (eds.), Idea Group
Inc.
zamfira@unitbv.ro
Object Detection Resource Usage Within a Remote
Real-Time Video Stream
1 Introduction
Remote Access Laboratories (RALs) provide a service whereby experimental rigs, key
hardware or software can be accessed and operated over a network remotely (Benetazzo
et al. 2000). Remote access provides the ability to deliver training and practical expe‐
rience to a larger cohort of students due to the increased availability of equipment. Most
RAL systems supply a live video stream of the equipment under control and include a
user interface to initiate tests and receive the results.
Augmented Reality (AR) shows a real-world environment with additional, computer
generated information. This allows a user of a service to experience a live action event,
but have the event enhanced through computer generated interactive sensory feedback
(Milgram and Kishino 1994). Users are generally provided sensory information of the
event which might not otherwise be directly viewable and thus extend the range of
information presented.
To date, Augmented Reality and Remote Access Laboratories have not been well
integrated. Incorporating AR into RALs has the potential to improve the practical expe‐
rience by supplying a rich sense of interactive control and immersion in the environment.
Combining these however introduces additional complexity and concern to current RAL
environments.
zamfira@unitbv.ro
Object Detection Resource Usage 267
Remote Access Laboratories are generally affected by network induced delays (Kist
et al. 2014), where the primary source of bandwidth consumption is typically via the
live video stream of the remote experimental rig. Integration of multifaceted systems
such as AR into RAL environments can therefore exacerbate these issues resulting in
potential rapid consumption of available ICT resources.
RALs provide an important resource to schools and universities (Fisher and Jensen
1980). Schools can utilize RALs at the fraction of a cost to purchasing and up keeping
didactic resources. Universities cover a diverse range of students, from different time-
zones, and demographics, requiring access to resources at any time (Gustavsson 2003).
Applying AR processes to the RAL environment provides an extra level of interaction
with the equipment, allowing for an improved user involvement and immersion into the
test environment (Azuma 1997).
An important aspect of implementing AR features is Computer Vision (CV). This
usual requires extensive data procession and is therefore resource intensive. Improving
the reality and immersion of RALs through AR enhancements comes at cost of these
resources.
CV models, used in security and surveillance fields to detect and track objects, tend
to need training or use off-line processing of recordings because of the limitations of the
technology (Fisher and Jensen 1980). Applying object detection and tracking methods
to the equipment within the live video stream of the experimental rig, increases
complexity. It may consume resources to a point that any expected advantage AR is to
provide, ends up nullified. This leads to the key question of what are the additional ICT
resources that are needed for the inclusion of AR services into a RAL environment. This
paper investigates this question.
This work outlines typical resources utilized by common CV models. It describes
an implementation of those models to measure and ascertain the impact each model has
on ICT resources. Consumption of the host computers memory and processors are
reported, along with the time taken to process each video frame. These results can then
be used to ascertain the minimum set of resources required for combined AR and RAL
configurations. Additionally, these figures can also be applied to other models as a
baseline of the known resource consumption underlying other processes.
This paper is structured as follows. Section 2 provides a brief overview of the current
RAL and AR works, focusing on any overlapping areas. Issues pertaining to AR
resources are addressed in Sect. 3, where special computer vision aspects are measured
and explained. Section 4 highlights methods to improve user immersion within the AR
RAL environment when bandwidth limitations exist. Section 5 concludes this paper.
2 Current Research
The field combining AR within the RAL environment is relatively new. Augmented
Reality is a component of the Virtual Continuum. This is a sliding scale representing
full reality at one extreme, and a completely virtual environment on the other extreme.
Virtual test rigs and experiments have been used in engineering education and the engi‐
neering industry for more than twenty years. With the recent advances in computing
zamfira@unitbv.ro
268 M. Smith et al.
availability and network capacity, these original activities have little resemblance to the
current RAL systems (Overstreet and Tzes 1999). Virtual rigs provide graphical repre‐
sentations of instruments which were controlled over a proprietary data bus (Fisher and
Jensen 1980). Overstreet and Tzes (1999) have produced a client/server configuration
which has quickly promoted a rapid uptake of web-based RAL configurations. Virtual‐
ized equipment has dominated the field. More recently infrastructure costs and capabil‐
ities have caught up with expectations.
Expanding bandwidth provides the infrastructure that is required for live video feeds
for remote systems, now readily available (Stauffer and Grimson 2000). The inclusion
of video streams into RAL systems has also provided the impetus to the field to expand
into other non-science and engineering fields. Diverse schools and faculties are utilizing
RAL to enhance their pedagogical outcomes. Disciplines as diverse as Nursing (Maiti
et al. 2016) and surveying have benefited from practical remote control of technical
equipment.
RAL systems currently depend on live video streams for the user to observe the
operation of the equipment; however, interaction with the equipment is limited. Famil‐
iarization with technical equipment is somewhat restricted without the senses being
engaged with the functionality (Ester et al. 1996). Early mixed reality systems have
utilized fully virtualized instrumentation (Maiti et al. 2013). These mixed reality systems
where developed completely in-house, utilizing local resources. Support equipment
consisted of computer hardware and applications to simulate the environment, and
provide users with virtualized objects. Virtual Reality systems have not reached the
hype, mostly due to the lack of ICT capability and capacity. Some users of full virtual
environments also consider the experience unsettling (Fig. 1).
In recent years, AR has undertaken extensive growth in all aspects of computing.
This includes a wide variety of mobile devices. Mobile AR (Azuma 1997; Maiti et al.
2013; Fazli et al. 2009; Ester et al. 1996) systems have helped to promote the technology
through a series of convenient applications such as the addition to Google’s StreetView
called StreetLearn (Wagner and Schmalstieg 2003). Applications such as StreetLearn,
demonstrate the technology, helping to further promote research and development.
The majority of AR operations are performed for our visual sense. As the field, has
progressed, additional senses, such as tactile feedback systems were incorporated. As
such, immersion into the augmented environment has become easier to implement.
Azuma (1997) reported works of some of the first see-through head-mounted devices,
capable of viewing the current environment, overlaid with computer generated objects.
Augmented Reality has also expanded into education, helping users to visualize 3D
objects in real-time (Maiti et al. 2016). Using desktop or handheld devices, a magic-lens
effect is achieved where coded images within books, or on cards are detected, interpreted,
and rendered into complete 3D representations of topical items.
Remote Access Laboratories and Virtual Reality originally overlapped in engi‐
neering education and engineering industry fields. Remote instrument virtualization
provided a means to operate electronic test equipment over local networks. By the
1990’s, simple AR started to appear (Milgram and Kishino 1994) as a result of improved
computing resources. This form of AR, in experimentation, only supported basic sensory
zamfira@unitbv.ro
Object Detection Resource Usage 269
data. Sensor data was displayed on virtual instrumentation, while watching videos of
the experiment.
Current AR systems in RALs have limited abilities. Very few works combine the
two technologies. Combined systems focus on visual enhancements, with some works
on the other senses. Works cover some practical implementations such as taxonomies
between hands-on and remote experimentation (Maiti et al. 2013), and more computer/
electronic test-bed (Fazli et al. 2009) systems. Many works have demonstrated the tech‐
nology through elaborate configurations. Engaging students with the technology has
produced systems such as an AR Racing Car games (Grimson et al. 1998), and 3D
modelling systems, all promoting the technologies capabilities.
Hence visual methods are typically used with AR, which rely on Computer Vision.
These CV models commonly used in industry to capture objects within video scenes,
are resource intensive to the extent that it can force significant portions of the processing
to occur off-line. Understanding the resource requirements for both AR and RAL struc‐
tures is necessary to develop effective sensory feedback systems suitable for implemen‐
tation. Basic estimates about RAL system resource limitations exist (Kist et al. 2014),
but there is little work done that has investigated the impact of the two technologies in
combination.
Fig. 1. Virtual continuum. Full reality on the left and a full virtual environment on the right.
3 AR Resources
The majority of current AR works focuses on the visual sense, while CV techniques for
object detection and tracking are employed to understand the scene in the live video
stream. This section will explain the various CV models which can be used with AR
systems to detect and track objects in the video stream. Resource monitoring and meas‐
urements are presented to demonstrate the additional ICT burden imposed by AR
processes.
3.1 Background
Any system implementing AR processes, has to ensure that the users of those systems
are able to engage and interact in a timely manner. The sense of immersion within AR
applications soon fails if registration, tracking and timing errors interfere with the system
processes. Remote Access Laboratories already require a variety of hardware, computer,
software and networking resources. Without including the additional re-source load
imposed by the AR processes, the RAL system could become degraded. Consequently,
AR resource usage must be determined and minimized to maintain effective synchro‐
nization and immersion.
zamfira@unitbv.ro
270 M. Smith et al.
Augmented Reality interprets video scenes using two data modes: remote data sets
and local data sets. The use of local data sets is demonstrated in AR systems using
fiducial markers, which render 3D models (Grimson et al. 1998) when the marker is
detected. Remote data sets are impacted by the network resources available. Desktop
and mobile AR systems have had to delegate the object detection and graphic processes
to separate systems (Wagner and Schmalstieg 2003) so as to cope with the computing
resource demands. This delegation reduces the local resources needed to render the
virtual objects that interact with the current environment.
Developing visual AR systems hinges heavily on CV models. Previous CV works
on video streams provide comprehensive object identification and tracking. Unfortu‐
nately, CV models rely on off-line or post processing of the video stream. Very few
systems provide live or real-time interpretation of the video stream because of the heavy
load on ICT resources.
Computer Vision techniques are expected to materialize physical objects from
multidimensional datasets (e.g. video frame), with the same level of competency as the
human eye and brain. Within video streams, CV systems must attempt to compensate
for shadows, lighting variations, a moving background (such as trees moving in the
wind) and periodic object movements.
To help understand the variations in each video scene, the CV systems require
extensive training. Statistical analysis (Fazli et al. 2009), clustering (Ester et al. 1996)
and frame subtraction systems (Stauffer and Grimson 2000) require considerable
processing to handle data sets consisting of 20–30 frames per second, with a minimal
resolution of 76,800 pixels per frame (typical 240 × 320 frame size). This equates to
307.2 kB of data for a 32 bit RGB encoded frame. A total of 1,536,000 pixels, or 6,144 kB
per second must be processed, which is beyond the capabilities of all but dedicated
hardware. Compounding the problem is the quality of the network services. As the
connection quality deteriorates, the number of frames available for processing also
diminishes. Good network connections provide smooth transitions between frames, but
increases the resource consumption to process those frames. The two strategic resources
counter-balance effective immersion of the AR experience.
Statistical Models
Statistical analysis of a pixel relies heavily on the historical data for the pixel. The
standard
zamfira@unitbv.ro
Object Detection Resource Usage 271
( ) ∑K ( )
p xN = 𝜂 xN ;𝜃j (1)
j=1
calculates the probability of the pixel being a foreground or background object through
its distribution of preceding frames. Cataloguing a single pixel via its distribution adds
to the overall processing requirements, and accumulates to significant levels. Testing
involved storing pixel arrays N deep (20 pixels). The previous 20 pixels for a coordinate,
are used as a Gaussian model to derive the status of the pixel. The current parameters
of the pixel are compared to the Gaussian parameters to ascertain if the pixel status has
changed. Time and processing costs involve statistical calculations for every pixel on
every frame. Below, in Fig. 3, are the processing times for each frame, using normal
distribution. The formula below was applied to each pixel, with no weighting of the
distribution, so as to keep the processing requirements to a minimum.
Figure 2 shows a reasonably consistent period of approximately 140 ms for each
frame and is much larger than the required 50–33 ms frame rate for standard video feeds.
This demonstrates that processing live video in the current configuration, will allow only
every third or fourth frame to be processed.
200
GMM
Frame Subtraction
180
160
140
process time (ms)
120
100
80
60
40
20
0
1 9 17 25 33 41 49 57 65 73 81 89 97 105 113 121 129 137 145 153 161 169 177 185 193 201 209
Frame number
Fig. 2. Statistical (GMM) frame analysis and frame subtraction frame analysis: frame processing
time
zamfira@unitbv.ro
272 M. Smith et al.
frame, the color values may fluctuate for many reasons such as; slight ambient lighting
changes, shadows, reflective surfaces, and the camera’s internal CCD variations.
For this test, simple raster like processing measures the difference between pixels of
the same coordinate (x, y), from the current frame and the previous frame. A threshold
of 10% was used. Each pixel has a maximum value of 255 per color channel, so if the
difference between pixels was less than 25, it was set as white, otherwise it was set to
black.
Frame subtraction testing on the video file, produced better results than the statistical
and DBSCAN models, as shown in Figs. 2 and 3. A consistent 110 ms processing time
occurred for each frame. These results are still far from ideal, with every second or third
frame needing to be ignored if implemented in an AR environment.
12000
10000
8000
6000
4000
2000
0
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106 113 120 127 134 141 148 155 162 169 176 183 190 197 204 211
-2000
Clustering Models
Clustering methods do not take a pixel in isolation, but must analyze all unclassified
pixels within its neighborhood. All pixels within a radius (depending on the clustering
model) are required to be verified as to their suitability to be a member of current group.
Additionally, a pixel must be directly density-reachable (Ester et al. 1996) to the core
pixels before it can be considered a member of the cluster. While the task is not tech‐
nically challenging, the iterative nature of a O(n log(n)) time complexity system,
consumes precious resource time.
A pixel, under DBSCAN, can be a member of only one cluster. For each frame, the
test involves checking each pixel’s (Px) neighborhood, scanning a radius of pixels out.
Unclassified pixels within the region are tested, and if suitable, marked as part of P(x)
cluster. Processing costa increase as the number of clusters, the radius of the neighbor‐
hood, and the number of pixels’ reachable increases.
Performing a DBSCAN pass on each of the frames which have undergone processing
(such as frame subtraction of statistical analysis), adds significant delays. Figure 4 shows
a relatively consistent frame processing period until frame 149. At this point, the video
scene has an increase in object motion, and DBSCAN processing increases as a result.
This delay is at totally unacceptable levels. The summary of results shown in (Stauffer and
zamfira@unitbv.ro
Object Detection Resource Usage 273
Grimson 2000). tally the resource usage for the framework models tested. (Stauffer and
Grimson 2000) results only consist of the CV model attributes, and do not include the user
interaction functions, sensor data or other RAL resource needs.
100
80
60
40
20
0
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106 113 120 127 134 141 148 155 162 169 176 183 190 197 204 211
Offline video processing systems are tuned for graphical tasks, and are beyond the
capabilities of generic desktop systems of remote laboratory users. The summary of
results shown in Table 1 tally the resource usage for the framework models tested.
Table 1 results only consist of the CV model attributes, and do not include the user
interaction functions, sensor data or other RAL resource needs.
The vision systems within a RAL environment can consist of single or multiple
cameras. An AR implementation must be expected to process the live video stream(s),
interpret the scene, accept sensor data, render video overlays, and retransmit data back
to the remote laboratory. User’s interaction and satisfaction with the AR RAL config‐
uration will wholly depend on the consumption of the ICT resources during this process,
and how the management of the resources can ensure user immersion in the experiment
or practical session.
zamfira@unitbv.ro
274 M. Smith et al.
4 AR Improvements
The network topographies of RALs have been investigated (Maiti et al. 2013) to an
extent that they are well understood, and provide the offset to all timing techniques and
calculations discussed in this section. Network delays must be constantly and carefully
monitored and controlled so that the accumulation of all delays remains within accept‐
able levels for AR RAL environments.
Augmented Reality resource consumption revolves around interpreting the scene
between the frames of the live video stream. Any technique employed to improve AR
responsiveness must assume a minimum network latency.
Previously, old military visual systems, using cathode ray tubes, would hijack the
interlacing scheme to interweave tactical information into the image. In today’s envi‐
ronment, image overlays are the primary method to incorporate post-generation images
into the stream. Taking a leaf from the old frame interlacing techniques military systems,
image overlays can be given the opportunity to skip the current frame, providing addi‐
tional time for any intensive processing. Network latency times of 25 ms to 50 ms would
require that every second frame to be skipped, as a minimum resource necessity.
Changes within the scene between frames vary at different rates, depending on the type
of experiment/exercise being performed. Fast changing scenes may not find frame skip‐
ping an acceptable solution, while reasonable static scenes could be updated at much
greater intervals.
Rendering virtual objects is also dependent on the timing of sensor data received
from the experimental rig. Reception of the live video stream and the sensor data (also
streamed through the same link) complicates the synchronization of rendered virtual
objects. Ensuring the live action within the video stream matches sensor data informa‐
tion, adds to the processing overheads. With less active video scenes, it is possible to
pause or limit the need for continual analysis of each frame within the stream (Maiti
et al. 2016).
Real-time statistical analysis of a remote laboratory’s full video stream is time-
consuming through the number of pixels that must be processed. Reducing this bottle‐
neck can be achieved through the following techniques.
4.1 Windowing
Within every video scene, there are regions that are of no interest to the experiment, or
have no function. Within the gear experiment shown below in (Grimson et al. 1998),
separate regions are of interest at specific times. For example, monitoring only the top
half of an experiment, or the center region of the view is probably sufficient for some
demonstrations. This limits the processing needs of augmented systems to a much
smaller subset of data.
4.2 Training
CV techniques, based on statistical modelling, all benefit from training (Grimson et al.
1998) where each pixel’s color distribution is calculated from preloading, or training
zamfira@unitbv.ro
Object Detection Resource Usage 275
from existing video data. The Gaussian mixture model defines the probability distribu‐
tion of a pixel. The number of distributions is a factor of the available memory (Maiti
et al. 2016), and processor power. As additional frames are received into the CV
processes, the distributions are updated. If the Gaussian distributions are performed on
the stream before the experiment begins, then this training will provide a baseline for
comparison during actual rig operations. For every new frame, the pixel color values
are checked against the distributions. It is common to use a standard deviation of 2.5
(Ester et al. 1996) to determines the threshold for a pixel, marking it as either a back‐
ground or foreground object. Foreground objects are the detected objects, which are
tracked. Training will reduce runtime processing costs.
4.3 Client/Server
Workstations at remote locations will vary in capabilities, which limits the base level
resources acceptable for effective AR. Placing hardware capable of performing the
intensive graphical processing at the host, ensures that all client access can receive the
full AR immersion.
Clients receive a video stream that already has the image overlays included. Data
from sensors is also processed at the host. Clients receive the full and complete video
stream, including all feedback data, plus the interface to operate the various controls and
devices. Synchronization of screen transactions and sensor readings are simpler, with
only the user interaction requiring alignment with the video scene.
User interaction within a remote laboratory experimental rig is mostly through
controlling the various equipment. User actions trigger small data packets to the server.
Smaller data requirements to the server, have smaller network loading needs. Conse‐
quently, network delays from user interactions should be minimized and impact
modestly on the users’ immersion level. Processing responsive user input at the server
allows the server to supply the complete rendered scene back to the user. With any
reasonable network access, user input should undergo minimal delays to the resultant
feedback images.
5 Conclusion
Augmented Reality systems capable of integrating with RALs have many hurdles to
overcome to provide services across the wide range of practical and experimental envi‐
ronments. Efficient utilization of ICT resources is paramount for comprehensive and
effective immersion within the remote environment. Augmented Reality for RALs relies
on a responsive network conduit and synchronization between the visual information
and user interactions. All network delays must be accounted for when determining AR
configurations. With reduced capabilities in any of these pathways, then the immersive
effect of AR becomes a liability rather than a benefit.
Incorporating AR functionality into RALs involves consuming additional ICT
resources. A computer system undertaking the AR processes for a remote laboratory
will require additional memory, processor, and network resources. Executing CV
zamfira@unitbv.ro
276 M. Smith et al.
algorithms used within the AR processes provides a metric of the baseline resource
usage. Testing the three major computer vision methods - clustering, frame subtraction
and statistical - with actual remote experiment video streams demonstrated the imme‐
diate current shortcomings of the technology.
Object detection and tracking are both essential functions for AR and this paper has
identified the direction for future work to mitigate these limitations. Limiting the region
of interest within the video image can provide significant gains in terms of overall
responsiveness and reduction of resource usage. Providing some training to the vision
systems also benefits the responsiveness but requires additional management of the
experimental rig. Ensuring the users of a remotely controlled experiment have sufficient
resources can be mitigated by having the major computer vision processing done at the
host location by hardware better suited to the task.
These resource management plans will be further tested to ascertain their capabilities
and effectiveness for AR systems with a RAL environment.
References
Benetazzo, L., Bertocco, M., Ferraris, F., Ferrero, A., Offelli, C., Parvis, M., Piuri, V.: A web-
based distributed virtual educational laboratory. IEEE Trans. Instrum. Meas. 49(2), 349–356
(2000)
Milgram, P., Kishino, F.: A taxonomy of mixed reality visual displays. IEICE Trans. Inf. Syst.
77(12), 1321–1329 (1994)
Kist, A.A., Maiti, A., Maxwell, A.D., Orwin, L., Midgley, W., Noble, K., Ting, W.: Overlay
network architectures for peer-to-peer remote access laboratories. In: 2014 11th International
Conference on Remote Engineering and Virtual Instrumentation (REV), pp. 274–280. IEEE,
February 2014
Overstreet, J.W., Tzes, A.: Internet-based client/server virtual instrument designs for real-time
remote-access control engineering laboratory. In: Proceedings of the 1999 American Control
Conference, vol. 2, pp. 1472–1476. IEEE, June 1999
Fisher, E., Jensen, C.W.: PET and the IEEE 488 Bus (GPIB). OSBORNE/McGraw-Hill, Berkeley
(1980)
Gustavsson, I.: A remote access laboratory for electrical circuit experiments. Int. J. Eng. 19, 409–
419 (2003)
Azuma, R.T.: A survey of augmented reality. Presence Teleoper. Virtual Environ. 6(4), 355–385
(1997)
Wagner, D., Schmalstieg, D.: First steps towards handheld augmented reality. In: ISWC, vol. 3,
p. 127, October 2003
Maiti, A., Kist, A.A., Maxwell, A.D.: Estimation of round trip time in distributed real time system
architectures. In: Telecommunication Networks and Applications Conference (ATNAC), 2013
Australasian, pp. 57–62. IEEE, November 2013
Fazli, S., Pour, H.M., Bouzari, H.: A novel GMM-based motion segmentation method for complex
background. In: 2009 5th IEEE GCC Conference & Exhibition, pp. 1–5. IEEE, March 2009
Ester, M., Kriegel, H.P., Sander, J., Xu, X.: A density-based algorithm for discovering clusters in
large spatial databases with noise. In: KDD, vol. 96, no. 34, pp. 226–231, August 1996
Stauffer, C., Grimson, W.E.L.: Learning patterns of activity using real-time tracking. IEEE Trans.
Pattern Anal. Mach. Intell. 22(8), 747–757 (2000)
zamfira@unitbv.ro
Object Detection Resource Usage 277
Maiti, A., Kist, A., Smith, M.: Key aspects of integrating augmented reality tools into peer-to-
peer remote laboratory user interfaces. In: 2016 13th International Conference on Remote
Engineering and Virtual Instrumentation (REV), pp. 16–23. IEEE, February 2016
Grimson, W.E.L., Stauffer, C., Romano, R., Lee, L.: Using adaptive tracking to classify and
monitor activities in a site. In: Proceedings of 1998 IEEE Computer Society Conference on
Computer Vision and Pattern Recognition, 1998, pp. 22–29. IEEE, June 1998
zamfira@unitbv.ro
Integrating a Wireless Power Transfer System
into Online Laboratory: Example
with NCSLab
1 Introduction
Wireless Power Transfer (WPT) technology has drawn growing attention in recent
years. Limitations on electric wires are no longer problems for the transformation of the
electric field to the magnetic field, and then to the electric field. In [1], a far-field
technique using propagating electromagnetic waves that transfer energy the same way
as radios transmit signals is presented. In contrast to the far-field technique, M. Soljacic
[2] introduced a near-field (inductive coupling) technique operating at distances less
than a wavelength of the signal being transmitted. As the near-field technique requires
relatively low frequency compared with far-field technique, it attracts much research
attention since it was proposed [3–6].
WPT systems designed in Wuhan University [7–9] use inductive coupling technique.
WPT systems in [8, 9] which could be potentially used for high voltage power cable
monitoring were first introduced. All of the systems above adopt a simple 2-coil structure
easy for implementation rather than 3-coil [10, 11] or even multi-coil structure [12, 13].
Regarding conventional design, the tuning of parameters has been a problem.
Traditionally, finding the best working point of the WPT system is difficult as there are
a great number of coupled parameters to tune. What’s more, the experimenter of the
WPT system has to be on site to carry out the experiment with limitations on time and
location as well as safety issue.
State-of-the-art technology is able to keep people informed of the latest trends and
hotspot in the related field. However, conventional, it is not easy to share students with
the latest technology either for cumbersome equipment or device needs careful atten-
tion. For a WPT system with complicated structure and even high voltage generated in
the Tx and Rx side while energized, classroom demonstration is difficult.
The complicated implementation of physical system makes it impossible for every
university and institution to build a set of WPT system. Thus, it is urgent to address the
sharing issue to provide open access for experimentation and research, especially for
education of state-of-the-art technology.
The tuning issue, education issue along with sharing issue has brought out the idea
of Remote Control WPT (RCWPT) system based on networked control [14, 15] which
is a hotspot. There are already a great many of online laboratories which can provide
remote control of physical equipment. For example, in [16], the remote control of
electric and electronic instruments is introduced in NetLab, GOLDi-labs in [17] allows
users to remote control a 3-axis portal and in [18] a remote inclined plane laboratory for
displacement measurements versus time is presented.
NCSLab (Networked Control System Laboratory) is a hybrid online laboratory
which provides both physical and virtual test rigs for remote experimentation. Previ-
ously, only physical and virtual test rigs in control engineering are setup in NCSLab.
For example, fan speed control system [19], dual tank [20], and DC motor [21].
In all, there are 20 virtual rigs and six physical test rigs in NCSLab. However, as
one of the advantages of NCSLab, test rigs in geographically diverse locations can be
integrated into NCSLab [22]. Theoretically, all test rigs that match the interface of
NCSLab can be successfully deployed.
However, it remains to be found whether it is possible to utilize NCSLab to explore
the WPT system. For example, whether it is able to remotely find out the efficiency,
best working point and optimum frequency of the WPT system. Given that NCSLab is
powerful platform that new types of test rigs can easily be deployed into its framework
through the pre-customized interface, a WPT system could be potentially integrated
into NCSLab as well.
The WPT system is a physical test rig containing multiple electric and electronic
parts that all need careful attention. As various widgets such as textboxes, charts and
zamfira@unitbv.ro
280 Z. Lei et al.
gauges are integrated into NCSLab, it provides convenience to remotely monitor and
tune parameters in a visual mode. The WPT system in this paper uses a simple 2-coil
structure as other ones in Wuhan University.
The rest of the paper is organized as follows. In Sect. 2, the NCSLab architecture is
presented. Two of the specific features of NCSLab are also introduced in this
part. Section 3 describes the principle of a two-coil WPT system adopted in this paper.
In Sect. 4, the integration of WPT system into NCSLab is explored, in which the
controller, USB data acquisition card and control algorithm are discussed in details.
Section 5 gives an example of a well-configured monitoring interface of the WTP
system in NCSLab. The paper is concluded in Sect. 6.
2 NCSLab Architecture
Evolved through over 10 years with the latest upgradation, NCSLab provides full
access at www.powersim.whu.edu.cn/ncslab with the advantage of 24/7 with HTML5
technology fitted in. Apart from common features of remote laboratories [23, 24],
NCSLab has its specific features, two of which are introduced as follow.
1. Free from plug-ins
Web-based online laboratory offers convenience without any software installation.
However, potential web crash and updating issues caused by plug-ins remains to be
addressed. The finalization of HTML5 provides alternative to other 3D engines
which needs plug-ins for rendering. As previous Flash 3D engine is replaced by
HTML5 technology in NCSLab [25, 26] with more and more web browsers sup-
porting HTML5, experimenter can conduct various experiments in NCSLab in the
web browser free from plug-ins.
2. 3D Virtual roaming
Apart from the tree structure (laboratory - sub-laboratory - test rig) of the NCSLab,
virtual roaming which can be parallelly accessed is also provided for the experi-
menter. Same as in the physical laboratory, the experimenter can go to the virtual
laboratory building with keyboard and mouse. Several sub-laboratory room will
appear when walking into the main building. If the experimenter chooses one of the
sub-laboratory and walks in, a series of virtual experimental equipment will lie on
the virtual desks in front of the experimenter. Each virtual test rig is ready for
experimentation if it is “picked up” by the experimenter.
Figure 1 shows the current architecture of NCSLab in Wuhan University.
Researchers from all over the world can access the system to carry out experiments
with registered username and password as all the test rigs are open for experimentation.
Test rigs in control engineering as well as WPT system in electric and electronic
engineering are integrated into NCSLab.
zamfira@unitbv.ro
Integrating a Wireless Power Transfer System 281
To provide a RCWPT system, the key issue is to find an appropriate parameter for
control using inductive coupling. Another problem to be addressed is to offer
observable result for monitoring. Therefore, a simple two-coil structure WPT system is
the best option.
The circuit model of two-coil WPT system using magnetically coupled resonator is
shown in Fig. 2, in which the Tx coil and the Rx coil share the same resonant fre-
quency. As can be seen in Fig. 2, an AC voltage source drives a RLC branch on the Tx
side, which is able to create a high frequency magnetic field on the Tx side. Once the
Tx coil is energized at the resonant frequency, the Rx coil can recover the energy from
zamfira@unitbv.ro
282 Z. Lei et al.
RP1 Rp2
AC iron core
L1 L2
bulb
k
C1 C2
the field converted from electric power transmitted through the magnetic field between
the two coils. Finally the Rx coil can drive a load bulb for observation.
Using Kirchhoff’s voltage law (KVL), the two-coil model depicted in Fig. 2 can be
analyzed as
1
I1 ðR1 þ jxL1 þ Þ þ jxI2 M ¼ Vs ð1Þ
jxC1
1
I2 ðR2 þ jxL2 þ Þ þ jxI1 M ¼ 0 ð2Þ
jxC2
where R1 ¼ Rp1 , R2 ¼ Rp2 and the M is the mutual inductance between the Tx and Rx
coil. The relationship between coupling coefficient k and mutual inductance M are
pffiffiffiffiffiffiffiffiffiffi
M ¼ k L1 L2
To simplify the two circuit Eqs. (1) and (2), Z1 and Z2 are defined as the impen-
dence of the both circuit loops as
1 1
Z1 ¼ R1 þ jxL1 þ ; Z2 ¼ R2 þ jxL2 þ
jxC1 jxC2
Z2 Vs jxMVs
I1 ¼ ; I2 ¼ ð3Þ
Z1 Z2 þ x2 M 2 Z1 Z2 þ x2 M 2
zamfira@unitbv.ro
Integrating a Wireless Power Transfer System 283
A A
+
S1 D1 S3 D3 RP1 RP2
V
iron core
Vd V
L1 L2 bulb
S2 D2 S4 D4 k
C1 C2
- To S2 and S3
DDS
Frequency
Generation To S1 and S4
Controller
USB DAQ Current and
Card Voltage
NCSLab
Camera
Server
Figure 4 shows the RCWPT system in the physical laboratory, it can be seen that
there is no electric connection between the Tx and Rx coils. The physical system can
definitely be used for hands-on WPT experiment on site with forementioned limita-
tions. After integration, the RCWPT system called Wireless Power Transfer in
NCSLab can be accessed at http://www.powersim.whu.edu.cn/ncslab in the Compli-
cated System sub-laboratory for remote experimentation.
Due to the relocation of the laboratory, there is no enough space for the WPT
system. Thus, the current WPT system is setup at the corner of the laboratory. For the
sake of legibility, the system in Fig. 4 used a picture taken in May, 2016, which is
exactly the same system as the current one except for the distance between the two
coils. The location of the system demonstrates the advantage of the RCWPT for saving
space.
Apart from basic electronic components, in order to integrate the WPT system into
NCSLab to build a RCWPT, a controller, a USB DAQ (Data Acquisition) card and an
algorithm are three key factors.
zamfira@unitbv.ro
284 Z. Lei et al.
Frequency
Rx coil Tx coil
generator
H-bridge high
Rectifier frequency
converter
Bulb
Controller
DAQ card
Fig. 4. Remote controlled WPT system (taken in May, 2016 in the old laboratory)
zamfira@unitbv.ro
Integrating a Wireless Power Transfer System 285
current, and the voltmeter is to measure the output voltage. The experimenter is able to
watch the monitoring result in the web page, in which the brightness of the bulb shows
the output power of the system.
Traditionally for other WPT systems in Wuhan University, a direct digital syn-
thesizer (DDS) module controlled by a MCU (microcontroller unit) is adopted to
generate the accurate square wave exciting signal [9]. Using the keyboard on the MCU
controller, the output frequency can be tuned from 0.1–1 MHz with the step size of
10 Hz. To achieve remote control of the WPT system, the controller is connected to the
frequency generator. Parameters such as inciting frequency, sweep frequency and
sweep amplitude can be remotely reset as long as it can be found from the control
algorithm.
zamfira@unitbv.ro
286 Z. Lei et al.
The parameters in the algorithm such as frequency “Hz” block and sweep fre-
quency and amplitude in “Sweep Setting” block can be found and tuned in the tree
structure of the monitoring and control interface of NCSLab, and signals such as input
current, voltage and output current and voltage could be monitored using various
widgets offered by NCSLab.
A WPT system can be integrated into NCSLab using hardware and algorithm men-
tioned in Sect. 4. The remote control platform NCSLab adopts Web structure, which
means experimenters don’t have to install any client applications. With the latest
technologies brought in, the platform is completely plug-in free, so the experimenter
just has to register and login to conduct the experiment on RCWPT system.
As the WPT system is for remote control rather than power delivery, the power
transfer efficiency and transferred power is not the priority in this paper, thus, the
RCWPT system is built without precise calculation. With the use of various widgets
provided by NCSLab, the system is able to monitor signals and parameters such as
current, voltage, power and frequency. Parameters such as frequency and
sweep-frequency amplitude can be easily controlled in the user-defined interface.
Signals can be collected easily using widgets like charts and gauges. More importantly,
it helps to remotely explore the optimum transfer frequency by tuning the exciting
frequency, sweeping frequency and sweeping amplitude.
In order to analyse the power transfer efficiency and optimum frequency, data such
as the input and output power, and working frequency should be collected. In partic-
ular, to obtain the optimum frequency, the WPT system should be set in
sweep-frequency mode, which is shown in Fig. 7(a). The resonant frequency is
180.75 kHz at the distance of 13 cm with sweep frequency 0.4 and sweep amplitude
1000 Hz, the transfer efficiency can be obtained. Figure 7(b) shows the RCWPT
system working at resonant frequency, in which the output current and voltage are
1.172 A and 5.782 V, respectively. It can be calculated that the output power is
6.777 W. From Fig. 7(b), it can be seen clearly that the bulb is brighter than the
moment in Fig. 7(a), in which the output current and voltage are 0.8707 A and
2.782 V, respectively, and the output power is 2.422 W.
Once the state-of-the-art WPT technology is integrated into NCSLab, it is able to
provide remote access for the teachers and students. On one hand, the teacher can
clearly explain the RCWPT system through classroom demonstration. On the other
hand, the students can carry out the WPT experiment individually anytime anywhere
with customized control and monitoring interface. The integration brings technology
close to students with less cost and more convenience.
zamfira@unitbv.ro
Integrating a Wireless Power Transfer System 287
(a)
(b)
Fig. 7. RCWPT system in NCSLab (a) working at sweep-frequency mode (frequency at
1.8075 kHz ± 1000 Hz, x = 0.4) (b) working at 1.8075 kHz
6 Conclusion
In this paper, a WPT system is deployed into the NCSLab framework. The integration
of WPT system into NCSLab benefits from various monitoring and control widgets of
NCSLab. The optimum frequency and best working point can be easily obtained by
setting the WPT system in the sweep-frequency mode using widgets of the NCSLab,
which shows the results in a visual and intuitive interface. Thus, the system can be
adapted to the best working point by resetting the frequency obtained previously,
zamfira@unitbv.ro
288 Z. Lei et al.
which could make the system working at the best condition and achieve the highest
output power. The remote control platform provides flexibility for the experimenter to
remotely perform experiment anytime anywhere as long as the Internet is available,
which address the tuning issue as well as the safety issue at the same time. Using
NCSLab, the WPT system is able to be integrated into online laboratory for remote
experimentation for both classroom demonstration and experiment by students, which
brings state-of-the-art technology close to students.
Acknowledgement. This work was supported by the National Natural Science Foundation
(NNSF) of China under Grant 61374064.
References
1. Sample, A.P., Yeager, D.J., Powledge, P.S., Mamishev, A.V., Smith, J.R.: Design of an
RFID-based battery-free programmable sensing platform. IEEE Trans. Instrum. Meas. 57
(11), 2608–2615 (2008)
2. Kurs, A., Karalis, A., Moffatt, R., Joannopoulos, J.D., Fisher, P., Soljacic, M.: Wireless
power transfer via strongly coupled magnetic resonances. Science 317(5834), 83–86 (2007)
3. Inagaki, N.: Theory of image impendence matching for inductively coupled power transfer
systems. IEEE Trans. Microw. Theory Tech. 62, 901–908 (2014)
4. Kiani, M., Jow, U.-M., Ghovanloo, M.: Design and optimization of a 3-coil inductive link
for efficient wireless power transmission. IEEE Trans. Biomed. Circuits Syst. 5(6), 579–591
(2011)
5. Sample, A.P., Meyer, D.A., Smith, J.R.: Analysis, experimental results, and range adaptation
of magnetically coupled resonators for wireless power transfer. IEEE Trans. Industr.
Electron. 58(2), 544–554 (2011)
6. Beh, T.C., Kato, M., Imura, T., Oh, S., Hori, Y.: Automated impedance matching system for
robust wireless power transfer via magnetic resonance coupling. IEEE Trans. Industr.
Electron. 60(9), 3689–3698 (2013)
7. Deng, Q., Liu, J., Czarkowski, D., Kazimierczuk, M.K., Bojarski, M., Zhou, H., Hu, W.:
Frequency-dependent resistance of litz-wire square solenoid coils and quality factor
optimization for wireless power transfer. IEEE Trans. Industr. Electron. 63(5), 2825–2837
(2016)
8. Zhou, H., Zhu, B., Hu, W., Liu, Z., Gao, X.: Modelling and practical implementation of
2-coil wireless power transfer systems. J. Electr. Comput. Eng. 27, 1–8 (2014)
9. Hu, W., Zhou, H., Deng, Q., Gao, X.: Optimization algorithm and practical implementation
for 2-coil wireless power transfer systems. Am. Control Conf. (ACC) 2014, 4330–4335
(2014)
10. Kang, S.H., Choi, J.H., Harackiewicz, F.J., Jung, C.W.: Magnetic resonant three-coil WPT
system between off/in-body for remote energy harvest. IEEE Microwave Wirel. Compon.
Lett. 26(9), 741–743 (2016)
11. Moon, S.C., Kim, B.C., Cho, S.Y., Ahn, C.H., Moon, G.W.: Analysis and design of a
wireless power transfer system with an intermediate coil for high efficiency. IEEE Trans.
Industr. Electron. 61(11), 5861–5870 (2014)
12. Yin, J., Lin, D., Lee, C.K., Hui, S.Y.R.: A systematic approach for load monitoring and
power control in wireless power transfer systems without any direct output measurement.
IEEE Trans. Power Electron. 30(3), 1657–1667 (2015)
zamfira@unitbv.ro
Integrating a Wireless Power Transfer System 289
13. RamRakhyani, A.K., Lazzi, G.: Interference-free wireless power transfer system for
biomedical implants using multi-coil approach. Electron. Lett. 50(12), 853–855 (2014)
14. Lai, J., Zhou, H., Lu, X., Yu, X., Hu, W.: Droop-based distributed cooperative control for
microgrids with time-varying delays. IEEE Trans. Smart Grid 7(4), 879–891 (2016)
15. Lu, X., Yu, X., Lai, J., Guerrero, J.M., Zhou, H.: Distributed secondary voltage and
frequency control for islanded microgrids with uncertain communication links. IEEE Trans.
Indus. Inf. doi:10.1109/TII.2016.2541693
16. Nedic, Z.: Demonstration of collaborative features of remote laboratory NetLab. In: 2012 9th
International Conference on Remote Engineering and Virtual Instrumentation (REV), pp. 1–4
(2012)
17. Henke, K., Vietzke, T., Hutschenreuter, R., Wuttke, H.D.: The remote lab cloud
‘GOLDi-labs.net’. In: 2016 13th International Conference on Remote Engineering and
Virtual Instrumentation (REV), pp. 37–42 (2016)
18. Stefka, P., Zakova, K.: Displacement measurements versus time using a remote inclined
plane laboratory. In: 2016 13th International Conference on Remote Engineering and Virtual
Instrumentation (REV), pp. 435–439 (2016)
19. Hu, W., Liu, G.-P., Zhou, H.: Web-based 3-D control laboratory for remote real-time
experimentation. IEEE Trans. Industr. Electron. 60(10), 4673–4682 (2013)
20. Hu, W., Zhou, H., Liu, Z.-W., Zhong, L.: Web-based 3D interactive virtual control
laboratory based on NCSLab framework. Int. J. Online Eng. 10(6), 10–18 (2014)
21. Lei, Z., Hu, W., Zhou, H., Zhong, L., Gao, X.: A DC motor position control system in a 3D
real-time virtual laboratory environment based on NCSLab 3D. Int. J. Online Eng. 11(3),
49–55 (2015)
22. Hu, W., Liu, G.-P., Rees, D., Qiao, Y.: Design and implementation of web-based control
laboratory for test rigs in geographically diverse locations. IEEE Trans. Industr. Electron. 55
(6), 2343–2354 (2008)
23. Santana, I., Ferre, M., Izaguirre, E., Aracil, R., Hernández, L.: Remote laboratories for
education and research purposes in automatic control systems. IEEE Trans. Industr. Inf. 9(1),
547–556 (2013)
24. Maiti, A., Maxwell, A.D., Kist, A.A.: Features, trends and characteristics of remote access
laboratory management systems. Int. J. Online Eng. 10(2), 30–37 (2014)
25. Lei, Z., Hu, W., Zhou, H.: Deployment of a web-based control laboratory using HTML5. Int.
J. Online Eng. 12(7), 18–23 (2016)
26. Hu, W., Lei, Z., Zhou, H., Liu, G.-P., Deng, Q., Zhou, D., Liu, Z.-W.: Plug-in free web
based 3-D interactive laboratory for control engineering education. IEEE Trans. Industr.
Electron. doi:10.1109/TIE.2016.2645141
zamfira@unitbv.ro
Spreading the VISIR Remote Lab Along Argentina.
The Experience in Patagonia
Susana Marchisio , Sonia Beatriz Concari3, Federico Lerro3, María Isabel Pozzo4,
3
1 Introduction
The Virtual Instrument Systems in Reality (VISIR) is a well-known remote lab that has
been discussed many times in this conference and in many articles published in journals.
Being designed and developed by Prof. Ingvar Gustavsson in Sweden almost 10 years
ago [1], this remote lab has been set-up in different European institutions. University of
Deusto was the first institution that purchased and deployed the VISIR outside Sweden,
and it was followed by other universities in Spain, Austria and Portugal. After the
expansion of the remote lab platform, the VISIR Consortium created around it aimed at
sharing experiences and experiments using the VISIR as a learning tool which helps
students and teachers to achieve the learning outcomes of subjects related to analogue
electronics. With the goal of spreading the knowledge about the VISIR, the VISIR
zamfira@unitbv.ro
Spreading the VISIR Remote Lab Along Argentina 291
+ project was presented to ERASMUS+ European Union call, being finally accepted
in July 2015 [2]. To fulfil that goal, European universities that have the experience of
using the VISIR will transfer it to Latin American institutions, namely Higher Education
Institutions with engineering careers/courses.
VISIR+ has two well-differenced well distinctive stages: during the first one, insti‐
tutions from Latin America must deploy the physical elements, instruments and compo‐
nents of the VISIR remote lab. This stage is supported by staff from BTH (Sweden), the
developers of the remote lab. In a second stage, the other European Universities involved
in the project will help their Latin American partners to exploit the resources of the
VISIR remote lab as a learning tool, sharing with them their experiences along these
years.
This paper rather than being focused on describing the VISIR+ aims at exploring
the results of the first training action that was held in Rosario (Argentina) in September,
2016. During this training action, staff from University of Deusto introduced the VISIR
remote lab to more than 25 trainers, lecturers and professors from different parts of
Argentina that were interested in discovering the possibilities offered by the VISIR. The
sessions started with an introduction to remote lab as many of the attendants were novel
in these environments.
The goal of this paper is to show not only the experiences during this training action,
but also the first intensive use of the VISIR by lecturers and students from Universidad
Nacional de la Patagonia San Juan Bosco (UNPSJB).
One of the expected results of the project is a set of educational modules for engineering
courses comprising the use of hands-on, simulated and remote labs, following an
enquiry-based methodology. It implies the inclusion of the VISIR remote lab in theo‐
retical and practical lessons with students, within a variety of courses related to electric
and electronic circuits. In order to fulfill that objective, the project VISIR+ has two
training actions in associated Latin American institutions partners of the Project.
The first training action in the framework of VISIR+ project took place at Facultad
de Ciencias Exactas, Ingeniería y Agrimensura (FCEIA) from Universidad Nacional de
Rosario (UNR) in September 2016. The training was developed during three days,
combining oral presentations, workshops and practical activities with VISIR. The
training sessions were led by two research professors of the Universidad de Deusto, who
are experts in the use of VISIR, plus three UNR teachers who usually use remote labo‐
ratory practice in their subjects of Electronic Engineering courses. Also present was one
researcher of from Instituto Rosario de Investigación en Ciencias de la Educación
(IRICE), member of the VISIR+ project, with the aim of taking records of from the
training sessions.
This training action at FCEIA targets all teachers with lecture duties in Engineering
courses related to electric and electronic circuits, plus two representatives from each of
the two UNR associated partners: Facultad Regional Rosario of the Universidad Tecno‐
lógica Nacional and Instituto Politécnico Superior of the UNR. As this training action
zamfira@unitbv.ro
292 U. Hernandez-Jayo et al.
was also considered one key moment for dissemination at a regional level, there were
also invited academic authorities, PhD students and teachers from other institutions
nearby UNR.
Three teachers from different Argentine Universities were also invited to participate
in this training action. They were selected by Consejo Federal de Decanos de Ingeniería
(CONFEDI). The participation of CONFEDI as an associated partner provides the
conditions for creating an additional impact at the national level in Argentina. The three
teachers attended the training sessions as regional coordinators of a project that
CONFEDI is carrying out in Argentina, –to encourage the subsequent dissemination of
the use of VISIR in the Engineering faculties. Belonging to this last target group, is the
professor of the Universidad Nacional de la Patagonia San Juan Bosco, whose experi‐
ence using VISIR is presented in this paper.
During this training action, staff from Universidad de Deusto introduced the VISIR
remote lab to 26 trainers, lecturers and professors from different parts of Argentina that
were interested in discovering the possibilities offered by the VISIR (Fig. 1).
Due to some administrative delays related to the import process, UNR still lacked
the necessary equipment to support training. This inconvenience was overcome by using
the VISIR platform of Universidad de Deusto, via Internet.
The sessions started with an introduction to remote lab because many of the attend‐
ants were novels in these environments. The training program included aspects related
with the design, implementation and the evaluation of educational modules with VISIR.
In addition it included application examples selected from those available on Web Lab
Deusto, to prove the adaptability of VISIR to different institutional cultures and its
universality in terms of experiments with electric and electronic circuits. The teachers
focused on both, technical and didactic aspects, especially in order to scaffold student’s
learning and foster their autonomy, namely by allowing them to conduct real experi‐
ments over the Internet. Once the training was completed, and to encourage both the
teachers’ motivation on the use of VISIR and the immediate application of what was
zamfira@unitbv.ro
Spreading the VISIR Remote Lab Along Argentina 293
learned to the classroom context, attendees were asked to plan an educational activity
using VISIR contextualizing the plan in their own subject, career and institution.
Table 1. TA impact/outcomes
Workshop Technological Participants
equipment expectations
Excellent 48.17% 6% 47%
Highly satisfactory 43.83% 26% 43%
Above average 8.00% 68% 10%
Most participants scored the workshop as excellent (48.17%) and highly satisfactory
(43.83%). Only 8% found the workshop above average. The evaluation of the workshop
included the overt explanation of the TA objectives, the time allotted, the instructors’
participation and the extent to which technological equipment had enhanced the effec‐
tiveness of teaching and learning. As regards the actual use of the technological equip‐
ment, namely VISIR Remote Lab, the answers ranged from too easy to use (i.e. excel‐
lent) 6%, easy to use (i.e. highly satisfactory) 26%, and just right (i.e. above average)
68%. Finally, TA met participants’ expectations by 47% as excellent, 43% as highly
satisfactory and 10% above average.
An open question was also included in the SQ in order to provide a qualitative
perspective to the evaluation by eliciting reflection on positive and negative aspects of
the whole experience. Three main categories aroused from the reading of participants
answers: equipment potential, clear presentation, time. Most of the participants argued
that the training action raised awareness about the potential of VISIR equipment not
only by presenting the possibilities of actual use in the classroom but also by giving
participants the chance to experiment during the sessions. Secondly, most participants
pointed out the presentation approach facilitated their understanding of VISIR technical
and pedagogical use. Finally, participants referred to the need of more time to extend
the TA experience: the schedule was constrained to some slots for actual connection to
VISIR via University of Deusto WebLab which participants considered limited.
zamfira@unitbv.ro
294 U. Hernandez-Jayo et al.
One of the participants of TA which took place at UNR from Universidad Nacional de
la Patagonia San Juan Bosco (UNPSJB) in Comodoro Rivadavia (Argentina) imple‐
mented VISIR Remote Lab in his subject Theory of Circuits. The subject Theory of
Circuits is in the second year of Electronic Engineering at the Engineering College of
UNPSJB. VISIR Remote Lab learning tool was introduced to the subject to give students
more options on real circuit experiments. To the traditional lab activities, practice was
added to allow students to analyze and interpret the forced temporal response to a resis‐
tive, inductive and capacitive circuit (RLC). In this type of practice students had to
experiment on a real circuit, i.e. select components and instruments, make the connec‐
tions, set the instruments and carry out the measurement. Before the practice, students
made the modeling, calculus and simulation of the target phenomenon.
The modeling developed by students was based on the circuit theory from which set
physic magnitudes had to be calculated, expressions of variables obtained and results inter‐
preted. The behavior of the model was also simulated by means of appropriate software and
the results were compared. In the next stage, students carried out the experiments using
VISIR Lab and contrasted the results against calculus and simulation drawing conclusions
from results. To organize the tasks, a lab guideline was designed where the objective of the
practice, the activities preliminary to real circuit experiment and procedures were made
explicit. Students had access to the remote lab and all necessary information about VISIR
from the subject webpage (http://www.ing.unp.edu.ar/electronica/asignaturas/ee016/) and
links to WebLab-Deusto from University of Deusto, Spain.
The students carried out the activities individually in the computer room of the Electronic
Department. At the beginning of the activities, a professor guided students in the use of
VISIR Lab about how to access to the remote lab by means of assigned users’ names and
passwords. Then students carried out the selection of components, the wiring, the instru‐
ment configuration and the measurement following the procedure given and the objectives
set for the practice. During this process, students shared with classmates the results of each
individual experience, their learning and conclusions, this time being the role of professor
that of a moderator.
To analyze and interpret the behavior of electric variables of RLC circuits, the guide
suggested the model shown in the figure with R1 = 100 Ω, C1 = 2.20 nF, L1 = 10 mH
(Fig. 2).
zamfira@unitbv.ro
Spreading the VISIR Remote Lab Along Argentina 295
The procedure established that the circuit should be wired on the “protoboard”,
generate a square signal, 500 Hz frequency and 1 V PP amplitude, and obtain the signals
Vg and VL from the oscilloscope from which attenuation and resonance frequencies
should be measured (theoretical magnitudes are α = R1/2L1 ω0 = 1/(L1 * C1)1/2 respec‐
tively). To obtain the attenuation frequency students observed from the oscilloscope the
time τ = 1/α by which VL falls to a 37% of its minimum value. To determine the reso‐
nance frequency, they observed the period T of the sinusoid and calculated ω0 = 2π/T.
The results obtained from the experience using the VISIR Remote Lab were then
compared to the previous activities. Students submitted a report with the description of
the practice carried out and the conclusions drawn to the professor (Fig. 3).
3.2 Impact
Adopting the new tool VISIR Remote Lab to carry out the experiments turned out to be
an appealing option for both students and professors. The tool is accessible and has an
outstanding graphic interface. During the experience, there was an immediate adoption
to the remote lab and the tool resulted intuitive, especially to students who most of the
time anticipated teachers’ explanations about use. Probably, being familiar with similar
real instruments at the UNPSJB lab, students did not need to read manuals or additional
online information about VISIR.
Many aspects from the subject Circuit Theory syllabus were strengthened using a
remote lab, namely the teaching objectives, the management, the task organization, the
zamfira@unitbv.ro
296 U. Hernandez-Jayo et al.
accessibility and the relation and integration with other pedagogical means and
resources.
If the experience of most active user (unpsjb_1) is studied, we can obtain the
following information:
• The user with the account unpsjb_1 has accessed to the lab 11 times. The total time
spent by the user on the lab has been 11780.19 s.
• The day that the user spent more time on the lab was on October, 18, being 60 1 h
performing experiments on the lab. From this session, the following information can
be obtained:
– The user performed 31 experiments on the lab. This does not mean that the user built
31 experiments, but he/she executed 31 times one or different experiment on the lab.
– This session was before the last one, and he/she did not perform any work circuit.
This means that he/she did not try to build any not allowed circuit or measurement.
zamfira@unitbv.ro
Spreading the VISIR Remote Lab Along Argentina 297
– During the whole session, the circuit under test was the same and it was built by the
user in the same way. He/she only changed the configuration of the instruments to
obtain a better resolution of the measurement and then a better understanding of the
circuit behaviour.
4 Conclusion
The outcomes defined for VISIR+ project are the natural evolution of the use of the
VISIR remote lab during the last 10 years. This remote lab has been tested and used by
all the European partners involved in the project, so now it is high time it was deployed
in other regions as Latin American. Then, all the experiences and experiments developed
for ten years are going to be shared between all the institutions of the project. The Project
implemented Training Actions to bridge these experiences between European and Latin
American institutions. This paper shows how the VISIR instance deployed at University
of Deusto is being used by Universidad Nacional de la Patagonia San Juan Bosco
(UNPSJB) in Comodoro Rivadavia. However, this is only the first step of the VISIR
spreading in Latin American countries. According to the working plan of the project,
two VISIR platforms will be deployed on Argentina, making easier and faster its use by
other Argentinean institutions.
Acknowledgment. The authors would like to acknowledge the support given by the European
Commission to the VISIR+ project through grant 561735-EPP-1-2015-1-PT-EPPKA2-CBHE-
JP.
References
1. Gustavsson, I., Nilsson, K., Zackrisson, J., Garcia-Zubia, J., Hernández-Jayo, U., Nafalski, A.,
Nedic, Z., Gol, O., Machotka, J., Pettersson, M.I., Lago, T., Hkansson, L.: On objectives of
instructional laboratories, individual assessment, and use of collaborative remote laboratories.
IEEE Trans. Learn. Technol. 2(4), 263–274 (2009)
2. Alves, G.R., Fidalgo, A., Marques, A., Viegas, C., et. al.: Spreading remote lab usage. A
System – A Community – A Federation. In: CISPEE Conference, Vila Real, Portugal, 19–
21 October 2016
3. Orduña, P., Bailey, P.H., Delong, K., López-De-Ipiña, D., García-Zubia, J.: Towards federated
interoperable bridges for sharing educational remote laboratories. Comput. Hum. Behav. 30,
389–395 (2014)
zamfira@unitbv.ro
Educational Scenarios Using Remote Laboratory VISIR
for Electrical/Electronic Experimentation
Felix Garcia-Loro1, Ruben Fernandez2, Mario Gomez2, Hector Paz2, Fernando Soria2,
María Isabel Pozzo3, Elsa Dobboletta3, André Fidalgo3,4, Gustavo Alves4,
Elio Sancristobal1, Gabriel Diaz1, and Manuel Castro1 ✉
( )
1
UNED, Madrid, Spain
{fgarcialoro,elio,mcastro}@ieec.uned.es, gdiaz@ieee.org
2
UNSE, Santiago del Estero, Argentina
raf@unse.edu.ar, mariog76@hotmail.com, hrpazunse@yahoo.com.ar,
mfernandos80@hotmail.com
3
IRICE-CONICET, Rosario, Argentina
pozzo@irice-conicet.gov.ar, elsadobboletta@gmail.com
4
IPP, Porto, Portugal
{anf,gca}@isep.ipp.pt
1 Context
zamfira@unitbv.ro
Educational Scenarios Using Remote Laboratory VISIR 299
The emergence of remote laboratories has provided new horizons in the learning process
and has brought new challenges in teaching design. Remote laboratories are being used
in many different ways and with different strategies as in-person laboratories have been
used traditionally.
Remote laboratories are a new tool to complement in-person laboratory, simulators
and virtual laboratories. The pool resulting by all the available possibilities provides a
wide range of possibilities when designing a course in which experimentation plays a
key role.
Virtual Instruments System In Reality (VISIR) is a remote lab for wiring electric
and electronic circuits experiments that has been used, in the Electrical and Computer
Engineering Department (DIEEC) of the Spanish University for Distance Education
(UNED), within several subjects from different engineering degrees, master subjects,
expertise courses, Small Private Online Courses (SPOCs) and Massive Online Open
Courses (MOOCs); providing satisfactory results with regarding to either it’s perform‐
ance or skills acquired by students.
The whole system, formed by all the actors and all the strategies used in the diverse
scenarios, have been analyzed in order to define a new learning environment, with the
objective of achieving an improved system in which all the teaching/learning scenarios
must have room, solving the inconveniences experimented separately and in their inter‐
action between them.
3 Approach Used
The main advantage of remote laboratories versus in-person laboratories lies in its access
availability without temporal nor geographical restrictions; The main advantage of
VISIR, when comparing with other electronic remote laboratories, lies in his concurrent
access: multiple users interacting with the remote laboratory simultaneously, designing
the same or different circuits and monitoring the same or different signals in real time,
as in an in-person laboratory room with replicated workbenches.
The experience reached in the integration of remote laboratory VISIR, mainly in
distance education, and the data collected from students’ feedback, logs related with the
remote laboratory interaction, surveys, etc. have allowed identifying the needs for
improvement and/or redesign.
All the data gathered from the LMS (Learning Management System) platforms have
been obtained analyzing the different databases ((PostgreSQL, MongoDB). To evaluate
VISIR behavior (accurately of the measurements, response managing requests’ over‐
load, etc.) and to inspect students’ interaction in the laboratory (common mistakes
typical from VISIR, number of accesses, etc., its database (MySQL).
zamfira@unitbv.ro
300 F. Garcia-Loro et al.
VISIR, (number of accesses, etc.) has been analyzed its database (MySQL) and its
logs (over 51 million of lines from the logs).
Inside the VISIR+ and PILAR projects, as well as previously inside the use of the
remote laboratory VISIR in the distance and online learning courses at UNED and the
MOOCs delivered using the VISIR system, a wide use has been obtained and published
regarding this use, [1–9].
According to Ursutiu et al. [10] and its reference to Learning by Experience from
Haynes any experience for learning involves a number of steps:
• Experiencing/doing with the instructor’s help or not;
• Sharing/what happened?
• Processing/analyzing;
• Generalizing;
• Applying.
Using this experience, the process developed to be more effective inside the new
starting of new installations, [11–16], (hardware, software and educational uses inside
the High Educational Academic environments) sites with the VISIR remote laboratory
and software are:
1. Share publications and tutorials regarding the use of VISIR inside electrical and
electronics engineering courses.
2. Share the use of VISIR remotely to allow the new teachers access to start working
with the VISIR system.
3. Start a first time face-to-face experience with some of the decision making teachers
and academic administrators regarding the feasibility and best practice of the use of
VISIR inside the target institution.
4. Start several synchronous sessions (using some collaborative environment, like
Moodle, videoconference facilities, etc.) with the new teachers and personnel
involved in the new deployment to allow a fast starting access as well as the first
touch of the system. During these preliminary sessions the expert or monitor will
show the main functions and specifications of the VISIR system as well as some
starting simple examples and use in the same environments and working area of the
future implementations.
5. Develop the face-to-face delivery with all the people involved in the on-site imple‐
mentation as well as with some possible new target institution members in the area
of the local University to try to have a core users target that will have in the future
the local use.
6. Have a local experience on the use and development of the educational implemen‐
tation of the VISIR remote examples with the local students, inside the classroom
and as well as with remote access to extend the experience of the use and as comple‐
mentary use of the remote laboratory.
7. Develop and extend the teaching experience from the local institution to all the core
new institutions inside the local area to reinforce the knowledge and implementation
as well as to develop new local strategies and synergies.
zamfira@unitbv.ro
Educational Scenarios Using Remote Laboratory VISIR 301
8. Realize a formal evaluation and quality assurance of all the process involved during
the implementation of the previous new acquisition and development of the VISIR
remote laboratory deployment.
4 Outcomes
5 Conclusions
The results show the ductility of VISIR remote laboratory in different learning scenarios.
Together with VISIR, it is needed a well-designed course, contents and experimental
experiences in order to obtain satisfactory results since, not only VISIR, a remote labo‐
ratory is a tool: it is a means, not an end in itself.
A LMS platform with the necessary tools for a deeper analysis of the students’
learning process and that integrates both environments (courses’ platforms and remote
laboratory) seems necessary in order to evaluate the convenience of the supplementary
documentation (videos, documents, activities, etc.) and their relationship with learning
and disengaging.
All these findings led to a new and more inclusive structure for the whole system in
order to a better exploiting of the experimental resources and, mainly, to create a new
learning environment intended for the analysis of the learning process for further
improvements.
zamfira@unitbv.ro
302 F. Garcia-Loro et al.
Acknowledgments. The authors acknowledge the support of the eMadrid project (Investigación
y Desarrollo de Tecnologías Educativas en la Comunidad de Madrid) - S2013/ICE-2715, VISIR
+ project (Educational Modules for Electric and Electronic Circuits Theory and Practice following
an Enquiry-based Teaching and Learning Methodology supported by VISIR) Erasmus+ Capacity
Building in Higher Education 2015 nº 561735-EPP-1-2015-1-PT-EPPKA2-CBHE-JP and PILAR
project (Platform Integration of Laboratories based on the Architecture of visiR), Erasmus
+ Strategic Partnership nº 2016-1-ES01-KA203-025327.
And to the Education Innovation Project (PIE) of UNED, GID2016-17-1, “Prácticas remotas
de electrónica en la UNED, Europa y Latinoamérica con Visir - PR-VISIR”, from the Academic
and Quality Vicerectorate and the IUED (Instituto Universitario de Educación a Distancia) of the
UNED.
References
1. Tawfik, M., Sancristobal, E., Martin, S., Gil, R., Diaz, G., Colmenar, A., Peire, J., Castro, M.,
Nilsson, K., Zackrisson, J., Håkansson, L., Gustavsson, I.: Virtual Instrument Systems in
Reality (VISIR) for remote wiring and measurement of electronic circuits on breadboard.
IEEE Trans. Ind. Electr. 6(1), 60–72 (2013)
2. Haertel, T., Terkowsky, C.: Creativity versus adaption: a paradox in higher engineering
education. Int. J. Creativity Probl. Solving 26(2), 105–119 (2016)
3. VISIR+ Project. Educational Modules for Electric and Electronic Circuits Theory and
Practice following an Enquiry-based Teaching and Learning Methodology supported by
VISIR - Erasmus+ Capacity Building in Higher Education 2015 nº 561735-EPP-1-2015-1-
PT-EPPKA2-CBHE-JP. http://www2.isep.ipp.pt/visir/. Accessed 15 Nov 2016
4. PILAR Project. Platform Integration of Laboratories based on the Architecture of visiR -
Erasmus+ Strategic Partnership nº 2016-1-ES01-KA203-025327. http://ec.europa.eu/
programmes/erasmus-plus/projects/eplus-project-details-page/?nodeRef=workspace://
SpacesStore/2d88ecb1-3db1-4a29-93c1-dd2802eec4f6. Accessed 15 Nov 2016
5. Naef, O.: Real laboratory, virtual laboratory or remote laboratory: what is the most effective
way? Intl. J. Online Eng. 2(3), 1–7 (2006)
6. Hanson, B., Culmer, P., Gallagher, J., Page, K., Read, E., Weightman, A., Levesley, M.:
ReLOAD: Real Laboratories Operated at a Distance. IEEE Trans. Learn. Technol. 2(4), 331–
341 (2009)
7. Nedic, Z., Machotka, J., Nafalski, A.: Remote laboratories versus virtual and real laboratories.
In: 34th ASEE/IEEE Frontiers in Education Conference, T3E-1, pp. 1–6, November 2003
8. Coble, A., Smallbone, A., Bhave, A., Watson, R., Braumann, A., Kraft, M.: Delivering
authentic experiences for engineering students and professionals through e-labs. In: IEEE
EDUCON, pp. 1085–1090 (2010)
9. Sancristobal, E., Martin, S., Gil, R., Orduna, P., Tawfik, M., Pesquera, A., Diaz, G., Colmenar,
A., Garcia-Zubia, J., Castro, M.: State of art, initiatives and new challenges for virtual and
remote labs. In: IEEE 12th International Conference on Advanced Learning Technologies,
ICALT, pp. 714–715, July 2012
10. Ursutiu, D., Samoila, C., Jinga, V.: Remote experiment and creativity. Int. J. Creativity Probl.
Solv. 26(2), 47–80 (2016)
11. Potkonjak, V., Vukobratovic, M., Jovanovic, K., Medenica, M.: Virtual mechatronic/robotic
laboratory - a step further in distance learning. Comput. Educ. 55, 465–475 (2010)
zamfira@unitbv.ro
Educational Scenarios Using Remote Laboratory VISIR 303
12. Jara, C.A., Candelas, F.A., Puente, S.T., Torres, F.: Hands-on experiences of undergraduate
students in automatics and robotics using a virtual and remote laboratory. Comput. Educ.
57, 2451–2461 (2011)
13. Rojko, A., Hercog, D., Jezernik, K.: Power engineering and motion control web laboratory:
design, implementation, and evaluation of mechatronics course. IEEE Trans. Ind. Electron.
57(10), 3343–3354 (2010)
14. Vivar, M.A., Magna, A.R.: Design, implementation and use of a remote network lab as an
aid to support teaching computer network. In: Third International Conference on Digital
Information Management, ICDIM, London (UK), 13–16 November 2008
15. Gustavsson, I., Nilsson, K., Lagö, T.L.: The visir open lab platform. In: Azad, A.K.M., Auer,
M.E., Harward, V.J. (eds.) Internet Accessible Remote Laboratories: Scalable E-Learning
Tools for Engineering and Science Disciplines Engineering Science Reference, pp. 294–317
(2012). ISBN 978-1-61350-186-3
16. Sheridan, T.B.: Descartes, Heidegger, Gibson, and God: towards an eclectic ontology of
presence. Presence Teleoperators Virtual Env. 8(5), 551–559 (1999)
17. Ma, J., Nickerson, J.V.: Hands-on, simulated, and remote laboratories: a comparative
literature review. ACM Comput. Surv. 38(3), 1–24 (2006)
18. Lang, D., Mengelkamp, C., Jager, R.S., Geoffroy, D., Billaud, M., Zimmer, T.: Pedagogical
evaluation of remote laboratories in eMerge project. Eur. J. Eng. Educ. 32(1), 57–72 (2007)
19. Lindsay, E.D., Good, M.C.: Effects of laboratory access modes upon learning outcomes. IEEE
Trans. Educ. 48(4), 619–631 (2005)
20. Intecca.uned.es: INTECCA | ¿Qué es AVIP? (2016). https://www.intecca.uned.es/
inteccainfo/plataforma-avip/que-es-avip/. Accessed 15 Nov 2016
21. VISIR SIG: Online-engineering.org. VISIR Special Interest Groups (SIG). http://
www.online-engineering.org/SIG_visir.php. Accessed 15 Nov 2016
22. UNED-DIEEC: VISIR – Electronics Remote Lab « Research on Technologies for Engineering
Education (2016). http://ohm.ieec.uned.es/portal/?page_id=76. Accessed 15 Nov 2016
23. Openlabs: OpenLabs - ElectroLab (2016). http://openlabs.bth.se/index.php?page=
ElectroLab. Accessed 26 Jan 2016
24. UNED Abierta: UNED Abierta (2015). https://unedabierta.uned.es/wp/. Accessed 15 Nov 2016
25. IMS Global Learning Consortium: Imsglobal.org, IMS Global Learning Consortium. https://
www.imsglobal.org/. Accessed 15 Nov 2016
26. IMS Global Learning Consortium: Imsglobal.org. IMS Global Learning Tools
Interoperability Basic LTI Implementation Guide. https://www.imsglobal.org/specs/ltiv1p0/
implementation-guide. Accessed 15 Nov 2016
27. Coursera Technology: Tech.coursera.org. LTI Integration. https://tech.coursera.org/app-
platform/lti/. Accessed 26 Jan 2016
28. Oauth.net: OAuth Community Site. http://oauth.net/. Accessed 15 Nov 2016
29. IMS Global Learning Consortium: Imsglobal.org, Learning Tools Interoperability®
Background. https://www.imsglobal.org/activity/learning-tools-interoperability. Accessed
15 Nov 2016
30. Weblab-Deusto: WebLab-Deusto, documentation. Authentication. https://weblabdeusto.
readthedocs.org/en/latest/authentication.html. Accessed 15 Nov 2016
31. Moodle: Moodle - Open-source learning platform. http://www.moodle.org. Accessed 15 Nov
2016
32. Swope, J.J.: A Comparison of Five Free MOOC Platforms for Educators: EdTech. (2014).
http://www.edtechmagazine.com/higher/article/2014/02/comparison-five-free-mooc-
platforms- educators. Accessed 15 Nov 2016
zamfira@unitbv.ro
Use and Application of Remote and
Virtual Labs in Education
zamfira@unitbv.ro
Robot Online Learning Through Digital Twin
Experiments: A Weightlifting Project
1 Introduction
The view of robots as systems that repetitively perform preprogrammed behaviors under
automatic control is changing. The increasing complexity of robot hardware and
growing sophistication of robot tasks in unstructured dynamic environments make it
difficult to preprogram robots for every possible scenario, urging the development of
robots capable of adapting to changes, coordinate movements, and learn new behaviors
[1]. Robot intelligence technologies used to implement these capabilities are based on
methods of machine learning, simulation, and cloud computing.
The two machine learning approaches, mainly applied to teach new skills to robots,
are imitation learning and reinforcement learning [2, 3]. In imitation learning, the robot
records and imitates the target movement demonstrated by the instructor. In reinforce‐
ment learning (RL), the robot is not directly instructed, but through autonomous trial-
and-error search, it determines and records the action which optimizes the performance
zamfira@unitbv.ro
308 I. Verner et al.
criterion. The opportunities to teach robots new skills or adapt existing skills to new
situations by using the RL approach have been intensively investigated, while one of
the main challenges is to reduce the experimentation time and the wear and tear on the
robot. A recommended method to cope with this challenge is by means of an empirical
predictive model which is autonomously generated and used to guide robot actions [4].
Simulation in robotics is a software tool for design and testing robot behaviors on a
virtual robot before implementing them on a real robot. Some of the benefits of using
robotics simulations, listed in [5], directly relate to reinforcement learning. In particular,
performing robot trials in a virtual environment allows experimental data to be generated
faster, more easily, and in any desired quantity, thus significantly speed-up the learning
process. Modern computer aided design systems provide means for creating virtual
models which accurately resemble the geometric and mechanical characteristics of the
real robots.
Cloud robotics is a method to enhance functionality of a robot by using remote
computing resources of memory, computational power, and connectivity [6]. In robot
learning, connection to the intended cloud platform enables the accumulation, storage,
and processing of data of robot trials and other relevant information on the Web server,
and transmit the data to the robot. The platform can serve a hub of the Internet of Things
(IoT) network, through which robots can share between them the learned skills and
communicate with other systems.
The goal of our research project is to propose and explore an approach in which the
challenge of implementation of robot learning is used as a thread for teaching the
discussed robot intelligence technologies to high school and first-year engineering
students. In this approach, the student is assigned to implement a robot task in which the
desired behavior cannot be pre-programmed, but has to be learned by the robot. In such a
project the student teaches the robot to acquire the skill by implementing a reinforcement
learning process supported by simulation modeling and cloud communication.
Our research is an ongoing multi-case study conducted at the Technion Center for
Robotics and Digital Technology Education through collaboration with the PTC Israel
Office. Participants are 1st and 2nd year students from MIT doing summer internship
projects in our lab, high school (HS) students participating in our outreach activities,
and Technion students studying technology education. We utilize robots constructed by
students using the ROBOTIS Bioloid Premium kit (http://en.robotis.com) and software
tools by PTC, namely, the 3D modeling system Creo Parametric, and the IoT platform
ThingWorx.
Our project so far has passed three research phases. In the first phase, university
students implemented a RL scenario in which a humanoid robot, through a series of
trials, learns to adapt its body tilt angle for lifting different weights [7]. In the second
phase, a group of high school students, mentored by a faculty staff member and our
university students, constructed animal-like robots and implemented different RL
scenarios, utilizing the approach tested in the first phase. The focus of this paper is the
third phase, in which university students apply reinforcement learning, 3D modeling,
and cloud communication in order to implement a scenario in which a humanoid robot
learns to manipulate multiple joints to maintain its stability when lifting different
weights.
zamfira@unitbv.ro
Robot Online Learning Through Digital Twin Experiments 309
zamfira@unitbv.ro
310 I. Verner et al.
In the following section we will discuss the way in which the robot weightlifting
task has been implemented in our current study.
zamfira@unitbv.ro
Robot Online Learning Through Digital Twin Experiments 311
Then, we compared the balance characteristics of the physical robot and its digital twin.
We tested the balance of the physical and virtual robots in the same posture shown in
Fig. 1, for different values of mass of the weight. For the virtual robot, the calculation
was made using the “center of gravity analysis” and “sensitivity analysis” features of
Creo. The maximal mass that each robot can hold in this posture without losing stability
was determined. We calculated that the discrepancy between the physical robot and its
virtual twin was less than 3%.
Each optimal solution, generated through simulation analysis for each mass value,
contained three parameters (hip, knee and ankle angles). Database of these optimal
solutions has been uploaded to a data table in IoT platform ThingWorx.
We defined a function in ThingWorx which, upon getting the weight value as input,
utilizes the data table, and returns the corresponding angle values of the optimal solution.
zamfira@unitbv.ro
312 I. Verner et al.
When the physical robot has to lift a weight, it first measures the weight mass, and sends
its value to the ThingWorx server. In response, the robot receives the values of the three
angles suggested based on the simulation analysis. Then the robot executes the lifting
using those values.
To visualize the online communication between the robot and ThingWorx, we
created a mashup web page which serves as a dashboard for displaying parameters of
the robot weightlifting trial (the weight’s mass and the three angles). The mashup is
shown in Fig. 2.
4 Educational Implementation
As noted in the introduction, our research project explores an approach to teaching robot
intelligence technologies to high school and first-year engineering students by engaging
them in teaching robots to learn. The instructional design in the project is implemented
by the authors of this paper. A faculty staff member develops instruction in robot
construction and programming and conducts courses for school and freshman students.
Two more instructional designers are Technion mechanical engineering graduates
working at PTC and pursuing an additional degree in science and technology education
in our faculty. In the research project, which is part of their studies, they develop
instructional units on 3D modeling and internet communication in robotics and mentor
school and freshman students in these topics. The strategic planning, project coordina‐
tion and supervision of the instructional designers are made by a faculty member who
also guides school and freshman students in pedagogical concepts relevant to robot
learning. Consultancy regarding the modeling and communication technologies and
software systems was given by PTC.
zamfira@unitbv.ro
Robot Online Learning Through Digital Twin Experiments 313
5 Conclusion
The approach proposed in our research extends the scope of educational robotics, which
traditionally focuses on practices with preprogramed robots. Results of our research
indicate that the challenge of developing learning robots can engage novice engineering
students in experiential learning of innovative concepts and technologies, such as
machine learning, parametric design, digital prototyping and simulation, connectivity
and internet of things. We found that those concepts and technologies are within the
grasp of understanding of freshman and HS students. In the next phase of the research,
we will practically implement our approach in an outreach course and we anticipate that
the evaluation of this experience will lead to the development of strategies for learning
with learning robots.
References
1. Peters, J., Lee, D.D., Kober, J., Nguyen-Tuong, D., Bagnell, D., Schaal, S.: Robot learning.
In: Siciliano, B., Khatib, O. (eds.) Springer Handbook of Robotics, pp. 357–394. Springer
(2017)
2. Sutton, R.S., Barto, A.G.: Reinforcement Learning: An Introduction. MIT Press, Cambridge
(1998)
3. Kormushev, P., Calinon, S., Caldwell, D.G.: Reinforcement learning in robotics: applications
and real-world challenges. Robotics 2(3), 122–148 (2013)
4. Nguyen-Tuong, D., Peters, J.: Model learning for robot control: a survey. Cogn. Process.
12(4), 319–340 (2011)
zamfira@unitbv.ro
314 I. Verner et al.
5. Reckhaus, M., Hochgeschwender, N., Paulus, J., Shakhimardanov, A., Kraetzschmar, G.K.:
An overview about simulation and emulation in robotics. In: Proceedings of SIMPAR, pp.
365–374 (2010)
6. Kehoe, B., Patil, S., Abbeel, P., Goldberg, K.: A survey of research on cloud robotics and
automation. IEEE Trans. Autom. Sci. Eng. 12(2), 398–409 (2015)
7. Verner, I., Cuperman, D., Krishnamachar, A., Green S.: Learning with learning robots: a
weight-lifting project. In: Robot Intelligence Technology and Applications, vol. 4, pp. 319–
327. Springer (2017)
8. Harada, K., Kajita, S., Saito, H., Morisawa, M., Kanehiro, F., Fujiwara, K., Kaneko, K.,
Hirukawa, H.: A humanoid robot carrying a heavy object. In: Proceedings of the IEEE
International Conference on Robotics and Automation, pp. 1712–1717 (2005)
9. Arisumi, H., Miossec, S., Chardonnet, J.R., Yokoi, K.: Dynamic lifting by whole body motion
of humanoid robots. In: IEEE/RSJ International Conference on Intelligent Robots and
Systems, pp. 668–675 (2008)
10. Rosenstein, M.T., Barto, A.G., Van Emmerik, R.E.: Learning at the level of synergies for a
robot Weightlifter. Robot. Auton. Syst. 54(8), 706–717 (2006)
11. Kuo, C.H., Kuo, Y.C., Chen, T.S.: Process modeling and task execution of FIRA weight-
lifting games with a humanoid robot. In: Conference Towards Autonomous Robotic Systems,
pp. 354–365. Springer, Heidelberg (2012)
12. Michieletto, S., Tosello, E., Pagello, E., Menegatti, E.: Teaching humanoid robotics by means
of human teleoperation through RGB-D sensors. Robot. Auton. Syst. 75, 671–678 (2016)
13. Anderson, J., Baltes, J., Cheng, C.T.: Robotics competitions as benchmarks for AI research.
Knowl. Eng. Rev. 26(1), 11–17 (2011)
14. Ahlgren, D.J., Verner, I.M.: Socially responsible engineering education through assistive
robotics projects: the RoboWaiter competition. Int. J. Soc. Robot. 5(1), 127–138 (2013)
15. Verner, I.M., Cuperman, D., Cuperman, A., Ahlgren, D., Petkovsek, S., Burca, V.:
Humanoids at the assistive robot competition RoboWaiter 2012. In: Robot Intelligence
Technology and Applications 2012, pp. 763–774. Springer, Heidelberg (2013)
zamfira@unitbv.ro
Interactive Platform for Embedded Software Development
Study
Galyna Tabunshchyk1 ✉ , Dirk Van Merode2, Peter Arras3, Karsten Henke4, and
( )
Vyacheslav Okhmak1
1
Zaporizhzhya National Technical University, Zaporizhia, Ukraine
galina.tabunshchik@gmail.com slavas490@gmail.com
2
Thomas More Mechelen-Antwerpen, Mechelen, Belgium
dirk.vanmerode@thomasmore.be
3
Faculty of Engineering Technology, KU Leuven, Leuven, Belgium
peter.arras@kuleuven.be
4
Integrated Communication Systems Group, Ilmenau University of Technology, Ilmenau,
Germany
karsten.henke@tu-ilmenau.de
1 Introduction
Internet of Things (IoT), from just a definition nowadays transforms into a global
approach for the functioning/controlling of things in the world. Things refer to anything
which is used, controlled, measured and is connectable to the internet. IoT refers to all
kind of devices and (powerful) microcontroller systems that can read sensors, do some
preliminary digital signal processing and send output over the Internet to a variety of
users, being other machines or human users. The emergence of very powerful multi-
core microcontrollers, large sized working memories and a wide variety of commercial
of the shelf sensors enables this new, exciting and challenging market. It is projected
that the average amount of microcontrollers per person is rapidly growing and will
continue to grow in the next few years [1]. In their new Internet of Things report, Busi‐
nessinsider.com projects there will be 34 billion devices connected to the internet by
2020 [2].
It should be clear that there are great job opportunities for specialists in this specific
high-skilled field of expertise. These specialists should have a profound knowledge of
both hard- and software aspects of the system, in interfacing with sensors, in using
embedded operating systems or real-time operating systems, but also on networking.
zamfira@unitbv.ro
316 G. Tabunshchyk et al.
The task for higher educational institutes (HEI) is to deliver highly skilled engineers
and developers to the labor market who have the knowledge to design, build, operate
and problem solve these devices. Especially when it deals with Industrial Internet of
Things, where these systems are deployed in an industrial environment to run process-
critical applications, quality issues of the combined hard- and software become
extremely important.
It is the task of HEI’s to provide the most efficient and effective means for learning
the necessary skills. To acquire skills, technology education at any level requires exper‐
imenting and practicing in labs. For embedded systems and digital systems opportunities
for creating new laboratories as remote labs offer the possibility to learn about these new
features (IoT) by using it. The main characteristics for robust embedded system are
reliability, availability, failsafe and dependability [3]. For this reason it is obvious
important to teach students techniques not only to use but also to develop reliable soft‐
ware for embedded systems.
This paper describes the development of remote experiments for the study of
embedded systems and the reliability of the embedded software.
To come to a working and reliable embedded system (ES), there is a great variety of
tasks to be solved. Besides the hardware design there is the different software tasks which
need to be implemented in the embedded system. For this paper we only look at some
aspects of the software development.
We focus on two main tasks which a ES-engineer needs to solve: the manipulation
of data elements on different levels and the collection and processing of data.
Any lab – including remote labs - should offer enough possibilities for students to
experiment and offer measurable learning outcomes, associated with experimenting. In
other words, care should be taken that the remote lab is more than a demonstration lab,
but a real experiment – although controlled from a distance.
When developing remote experiments as teaching/learning aid, one should bear in
mind the same questions as when developing any other didactical method: namely think
carefully on the learning outcomes and teaching approach. The learning outcomes will
point out what and how students will need to learn and also point on how to evaluate [4].
These observations clearly show that remote labs not only have advantages but also
are the cause of many challenges when considering the construction of remote experi‐
ments. The advantages for students is clear: the 24/7 availability to experiment and
repeat experiments can motivate students to achieve a deeper learning on the topics. The
challenges for the construction of the lab are to make it user-friendly, efficient in
achieving the learning outcomes and motivating and attractive to students. Another
major challenge is potential distant evaluation and feedback for the students on mistakes/
good and bad practices they used [3].
In order to improve teaching in diagnostic methods for embedded systems, a remote
lab (Interactive Lab Platform (ILP)) that examines reliability problems in real time was
constructed [4] based on the Raspberry Pi platform which we named Informational
zamfira@unitbv.ro
Interactive Platform for Embedded Software Development Study 317
The heart of the ISRT design is a Raspberry Pi Model B. The ISRT is also equipped
with an expansion board [4] developed at Thomas More Mechelen Antwerpen (TMMA)
University College, a MI0283QT Adapter v2, Wi-Fi, BLE4 adapters for signal variation
tasks, camera for video transmission tasks, a display for online compilation of the
programs and results overview.
The program is designed on the principles of MVC (Model-View-Controller). For
software development the following elements were used: NodeJS platform with inte‐
grated auxiliary module [6]; framework Expresses; as programming language Java‐
Script; library bcm2835 [5].
The outcomes of the remote labs are reached with different exercises. For a first explo‐
ration of the system there are demo examples.
Next there are assignments on ISRT software development and checking.
ISRT Demo Examples. For a first exploration of the hardware architecture, which can
be built with a Raspberry Pi as server, four Demo modules were developed which realize
different ways of ISRT system usage:
zamfira@unitbv.ro
318 G. Tabunshchyk et al.
This experiment can also be used for the assessment of the knowledge of first year
bachelor students in the binary system.
Experiments for face recognition are developed with the standard OpenCV library.
Main study outcomes is the knowledge of the standard OpenCV library and influence
of the type and strength of light and the type of the camera on the time-delay in facial
recognition. Understanding delay times and times of execution is very important for real
time applications in general, and for using video-streaming of live pictures in particular.
zamfira@unitbv.ro
Interactive Platform for Embedded Software Development Study 319
For the GSM module manipulation the SIM900 was provided. Students can send an
SMS on a Ukrainian provider or system and can show the last SMS sent to the GSM
module (Fig. 2).
Main outcomes are that students understand the pipeline of communication which
they use and how this is affected by different components of the system.
ISRT Software Development. Students can practice their skills on the remote system
and have assignments on the platform for formal evaluation.
The assignment is to control different remote experiments by a self-generated
program.
This task is performed on the built-in editor on the Raspberry Pi. Programming mode
allows the user to create and run their applications developed on C ++ and Python on
the platform Raspberry Pi directly from the panel (Fig. 3).
Access to the software development interfaces of the ISRT is allowed only to regis‐
tered users. As for all processes there is provided a log file for control and check if
students participated.
After login, the user gets access to the programming page. At this page there is the
list of all stored programs of the user. As such students can edit previously developed
programs or can create new ones from scratch. After the user chooses a file, the code
editor will be shown and the user can write/edit the required code. When selecting the
“compile” button, the program is sent to the server. If compilation is successful the user
is forwarded to the output page. If there are errors, they are displayed on the code editor
zamfira@unitbv.ro
320 G. Tabunshchyk et al.
in a red frame. At the output page it is possible to execute the program, clear the output
screen, see the real time video of the experiment and return to the editing.
This ISRT-lab allows access to the laboratory equipment 24/7.
All developed experiments are used for the courses in bachelor and master study for
the Embedded System Software Development Modules in Zaporizhzhya National Tech‐
nical Universities.
Examples of the tasks are:
– to develop a program in Python for adding binary numbers and displaying the results
of the addition on the display of the TMMA expansion board;
– to create a system loop and measure the Mean Time to Failure (MTF) for the SIM900;
– to create a program for face detection and compare the response time to the same
algorithms from OpenCV library (Fig. 4).
For the master students the idea of re-usage of systems is implemented. Master
students have the task to develop a measurement system for a certain defined purpose
(e.g. ecological measurements, climate control measurements etc.). They can (re)use the
developed templates on the system for the development of a their own personal meas‐
urement system, providing reliability tests for the specified lab hardware.
4 Conclusion
Questions of software and hardware reliability are of great importance. For embedded
systems it is a challenge as soft- and hardware reliability should be solved simultane‐
ously. For the tasks of building reliable software a low-cost system was developed,
which allow students to remotely practice skills in the embedded software development
in C++ and Python for Raspberry Pi. The developed remote lab makes a variety of tasks
possible on embedded platforms from basics of embedded systems to calculation of the
reliability characteristics. Future work is to provide built-in solutions for automated
testing of the plugged-in components.
zamfira@unitbv.ro
Interactive Platform for Embedded Software Development Study 321
Acknowledgement. In work was done in the frame of international project Tempus 544091-
TEMPUS-1-2013-1-BE-TEMPUS-JPCR [DesIRE] [7]. We want to thank EmSys Group from
Thomas More Mechelen-Antwerpen University College for their support of our work with the
TMMA expansion board.
References
1. IDC (2016): Worldwide Internet of Things Forecast Update 2016–2020. Doc # US42082716.
http://www.idc.com. Accessed 17 Jan 2017
2. Camhi, J.: BI Intelligence projects 34 billion devices will be connected by 2020, Internet of
Things report, BI Intelligence (2015). www.busisnessinsider.com. Accessed 17 Jan 2017
3. Kozik, T., Simon, M., Arras, P., Olvecky, M., Kuna, P.: Remotely controlled experiments. In:
Noga, H., Cernansky, P., Hrmo, R. (eds.) Nitra, Slovacia: Univerzity Konstantina Filozofa v
Nitre (2016)
4. Tabunshchyk, G.: Remote experiments for reliability studies of embedded systems. In:
Tabunshchyk, G., Van Merode, D., Arras, P., Henke, K. (eds.) Proceedings of XIII
International Conference on REV2016, Madrid, Spain, 24–26 February 2016, pp. 68–71.
UNED (2016)
5. Raspberry Pi b/b+ compatible expansion board. http://emsys.pbei.be/?product=raspberry-pi-
bb-compatible-expansion-board
6. Platform NodeJS. https://nodejs.org/en/
7. Tabunshchyk, G., Van Merode, D., Petrova, O., Ochmak, V.: Multipurpose educational system
based on Raspberry Pi. In: Proceedings of the International Symposium on Embedded Systems
and Trends in Teaching Engineering, Nitra, Slovakia, 12–15 September, pp. 202–206 (2016)
8. Brahmbhatt, S.: Practical OpenCV (Technology in Action), 1st edn., 244 p. Apress, New York
(2013)
9. DesIRE Project website. http://tempus-desire.eu/. Accessed 17 Jan 2017
zamfira@unitbv.ro
Integrated Complex for IoT Technologies Study
1 Introduction
Today the IoT technologies greatly extend the possibilities of collecting, analysis and
distribution of data, which humanity can transform into information and knowledge.
The Internet of Things opens new perspectives and gives more opportunities to increase
economic efficiency by automating processes in various fields of activity [1]. At the
beginning of 2016 the main segments for IoT applying were Manufacturing, Energy and
Transportation [2]. The impact of the IoT on companies’ activities is increasing. Smart,
connected products and the data they generate are transforming traditional business
functions, sometimes significantly [3].
Of course, there are still many issues that must be solved: more and more new unique
IP-addresses, sensors’ autonomous power supply, IoT devices certification, security,
protection of personal information, etc. [1].
zamfira@unitbv.ro
Integrated Complex for IoT Technologies Study 323
But even today, thanks to IoT technologies, the world begins to interact with physical
and virtual “things” and devices in other way. For this reason, companies need more
and more appropriate experts for development and implementation of new technologies
for effective interaction of customers and “things”. A lot of companies have already
sharply felt a lack of such specialists.
Therefore, the task of IoT technologies teaching is relevant and focused on formation
of students’ knowledge in the field of IoT modern software and hardware, and practical
skills in application of existing platforms and devices.
IoT technologies’ teaching is not a simple task. On the one side, there is a lot of web-
sites, webinars, documentation and materials on this topic [4–12]. On the other side,
even the interpretations of the term «IoT» in various papers sometimes significantly
differs. In addition, a huge number of different devices and platforms for creation of IoT
systems are described but sometimes it is problematic for students and teachers to sort
out the necessary information in this variety. Therefore, the task of creation of IoT
teaching-learning environment for future professionals training is an urgent.
As well as the author of [13], we have accepted the following definition as a basis:
«The Internet of Things (IoT) is the interconnection of uniquely identifiable embedded
computing devices within the existing Internet infrastructure». In this case it is possible
to distinguish several practically-oriented educational tasks: learning of Embedded
Systems (ES) software and hardware, analysis of existing approaches to ES design,
studying of the principles of ES interaction and connection to the Internet [14, 15].
As known, ES can be used in conjunction with Sensors/Actuators for collecting
the information and turning the collected or received information into actions. Also
the ES can use a range of technologies for connecting with other devices or the
Internet (Wi-Fi, Bluetooth, RFID, Ethernet, GSM, CDMA and so on) [16]. More
often, the concept of IoT is inseparably connected with something smart: Smart
House, Smart Transport, Smart City, Smart Businesses and so on [17]. Technolo‐
gies of Smart House creating are interesting and useful for students, as they allow to
make our life more comfortable, safe and to provide resource saving. That is why,
REIoT complex for IoT technologies study and investigations includes two inte‐
grated parts - Smart House&IoT laboratory and laboratory RELDES (REmote Labo‐
ratory for Development of Embedded Systems) [18–20]. Integration of remote labo‐
ratory into educational process expands opportunities of distance learning and gives
the students all advantages of remote experiment [21–27].
REIoT complex architecture is shown in Fig. 1. The concept of Smart House&IoT
lab’s stand passed several stages of design (Fig. 2).
zamfira@unitbv.ro
324 A. Parkhomenko et al.
Eventually we have used two the most popular embedded platforms for IoT smart
devices - Arduino and Raspberry Pi [13, 17, 27], as well as OpenHAB (Open Home
Automation Bus). OpenHAB is the software for integrating different home automation
systems and technologies into one general solution that allows comprehensive automa‐
tion rules and offers uniform user interfaces [28].
zamfira@unitbv.ro
Integrated Complex for IoT Technologies Study 325
Minicomputer Raspberry Pi performs the role of the lab server with installed
OpenHAB platform. Additional libraries Modbus TCP Binding are used for connection
it with Arduino boards with USB interface and Modbus RTU protocol. Sequential line
RS-232 is used for communication between electronic devices.
Each Arduino board contains a program, which handles input and output data.
Arduino uses open Modbus Master- Slave library. With the help of this library the
holding registers of sign or non-sign type which available for recording and reading
were created in each board (subsystem). Registers contain 8 or 16 elements of 16 bits
length each. Thus, the structure for data exchange is created. Each Arduino board oper‐
ates as Master. OpenHAB platform reads or adds data to registers when it interrogates
the devices. Each element of the register is correlated to individual parameters of sensors
or actuators.
The laboratory includes IP camera D-Link DCS-2121 which transmits video
streaming for users to view the experiments’ current status. This IP-camera is a complete
system with a built-in CPU and Web-server that transmits high quality video with reso‐
lution of 1280 × 1024 pixels and speed 10 frames per second. IP-camera and computer
are connected via Ethernet cable and interact using protocol TCP/IP. Router D-Link
DIR-300 allows to connect to the laboratory via Wi-Fi and also to add devices using
network cables.
For the integration of two REIoT complex parts as well as for Smart House&IoT lab
administration OpenHAB REST API was used [29]. To access Smart House&IoT lab
experiments, RELDES administration system sends HTTP GET request to the OpeHAB
REST API and receives results in JSON format.
zamfira@unitbv.ro
326 A. Parkhomenko et al.
On receiving the list of available in the Smart House&IoT lab experiments, RELDES
include them into the total list of experiments (Fig. 4) and after that a queue, statistics
and other functions are available for them.
Subsequently, to carry out the experiment, RELDES refers REST methods to the
Smart House&IoT lab, for example HTTP PUT and HTTP GET requests are used for
illumination level change and result control.
In order to start the streaming broadcast, we have used the utility ffmpeg [30].
FFmpeg is a set of free libraries with open source code that allow record, convert and
transmit digital audio and video in various formats. Library ffmpeg starts to catch video
from our camera with resolution 1280 × 1024, codes it to MPEG format with 10 fps and
bitrate 800kbit/s, and after that uses HTTP for sending to local server, which sends this
video stream to the end user.
In order to divide video for blocks (to cut and select the part of video for each
experiment), the filter “crop” is used. As a result, we have got some video fragments for
each experiment or for group of experiments.
The experiment «Solar station» is intended for studying the basics of solar energy and
the principles of the solar battery power. The main components are solar panel (6 V,
70 mA), Li-Ion battery and the charger for Li-Ion batteries. The experiment «Climate
control» is intended to study the basics of climate control using digital temperature and
humidity sensors DHT-11 and air quality analog sensor MQ135. The experiment «Zone
control» based on optical pair with photodiode, allows controlling the status of the
perimeter and it reacts in case of the disturbed perimeter. The experiment «Presence
control» is intended for studying the principles of the lighting systems or systems that
zamfira@unitbv.ro
Integrated Complex for IoT Technologies Study 327
control the human presence in the room. The experiment is based on the presence control
devices which control the reactions using pyroelectric sensor and Fresnel lens (pyro‐
electric motion sensor). The experiment «Ventilation» allows studying the basics of
recovery units and air flow control. The experiment is built with using the electrical
loads driver L298E, and it controls the loads with wide-width modulation. The experi‐
ment «Light control» is intended for studying the basics of load objects remote control
using relay module and wide-width modulation in multiple channels. The main compo‐
nents are: LED strip, RGB LED strip and load driver L293E. The experiment «Illumi‐
nation control» allows the students to perform light level control of the different areas
using photo-resistors (Fig. 5). The experiment «Access control» is compounded with
RFID reader module cards and trinkets RC522. The experiment is intended for studying
the principles of access and authorization systems. The experiment «Safety control»
allows the students to study the principles of security systems that react to exceptional
situation as the motion in controlled zone. The subsystem can be in the state of zones’
control and the state of sensors’ indication. The experiment is based on pyroelectric
motion sensor with using the sound alarm. Several experiments can be performed simul‐
taneously. In this case, the students study the principles of interaction between subsys‐
tems, define the process logic, create effects, evaluate the reaction of the elements and
analyze the results.
zamfira@unitbv.ro
328 A. Parkhomenko et al.
The students can use standard and create original Actions within Scripts and Rules
for execution OpenHAB specific operations. For example, such Actions as Telegram,
my.openHAB and other can be used for Smart House&IoT lab events notification or
feedback. Thus, connection to the Telegram allows sending messages to Telegram
clients from a bot-client (for example - sending notifications to the user about the air
conditioner turn on/off). With my.openHAB, students can connect to OpenHAB from
any device from everywhere with the Internet connection, to provide access to other
users as well as to keep all activities and events in the cloud my.openHAB.
The administration of events executed by OpenHAB can be realized with MailCon‐
trol binding. It provides the possibility of receiving commands sent via email in JSON
format. The following types of commands can be sent: decimal, HSB, increase –
decrease, on – off, open – closed, percent, stop – move, string, up – down. Therefore,
one of the practical tasks for the students can be the development of desktop or mobile
applications for OpenHAB connection and control.
Also the integration with Google calendar for REIoT complex is possible. Students
can create events and manage the system on schedule (on/off lighting, air conditioning,
open/close the door for a predetermined time, etc.).
So, the students acquire the IoT technologies knowledge and practical skills by
performing remote control experiments, studying the descriptions of experiments,
carrying out various scenarios and developing scripts.
5 Conclusions
REIoT complex for students’ research and training brings together several subsystems
to create a true Internet of Things for Smart House. It is used for a variety of training
tasks in several modes. In the first mode the complex provides the possibilities of remote
experiments on each subsystem separately and with the entire system as a whole. The
descriptions of the experiments and measurement results are available for students in
this case. Another mode allows students to specify the logic of the system working,
programming and processes monitoring.
The implementation of real practical tasks, based on modern technologies, gives the
students valuable practical experience in the IoT engineering, the motivation to research,
to work in team, to communicate with the customer, to present the results of their work
to the audience.
Practically-oriented teaching methods based on REIoT complex usage will provide
students with the necessary knowledge and skills for implementation of their own
projects, as well as for the successful application of IoT technologies in the future
professional activities.
References
zamfira@unitbv.ro
Integrated Complex for IoT Technologies Study 329
3. Porter, M., Heppelmann, J.: How smart, connected products are transforming companies.
Harvard in Business Review Articles. http://www.ptc.com/internet-of-things/harvard-
business-review/download-article-2#sthash.L5wGKzcN.dpuf
4. Knowledge is Power. https://www.thingworx.com/resources/?topic=&type=white-papers
5. Chung, J.-M.: Internet of Things & Augmented Reality Emerging Technologies (in Russian).
https://ru.coursera.org/learn/iot-augmented-reality-technologies/
6. Purpose-built for the Internet of Things. http://www.thingworx.com/
7. Internet of Things Institute Learning Center. https://education.ioti.com/
8. Internet of Things Roadmap to a Connected World. http://web.mit.edu/professional/digital-
programs/courses/IoT/
9. IEEE Internet of Things Webinars. http://iot.ieee.org/education/webinars.html
10. Vermesan, O., Friess, P.: Internet of Things - From Research and Innovation to Market
Deployment. River Publishers, Aalborg (2014)
11. Qiu, F.: Overall framework design of an intelligent dynamic accounting information platform
based on the Internet of Things. iJOE 12(5), 14–16 (2016)
12. Cai, K., Tie, F., Huang, H., Lin, H., Chen, H.: Innovative experimental platform design and
teaching application of the Internet of Things. iJOE 11(6), 28–32 (2015)
13. Stage 1 - Introduction to the Internet of Things: What, Why and How. http://
www.codeproject.com/Articles/832492/Stage-Introduction-to-the-Internet-of-Things
14. Tabunshchyk, G., Parkhomenko, A.: New technologies in the training of embedded systems’
specialists. In: International Scientific Conference on Internet-Education-Science, pp. 248–
250. VNTU, Vinnitsa, Ukraine (2014) (in Russian)
15. Prytula, A., Tabunshchyk, G., Parkhomenko, A.: Practically oriented teaching methods in the
field of embedded systems. In: VII International Scientific Conference on Current Problems
and Achievements in the Field of Radio, Telecommunications and Information Technology,
pp. 216–218. ZNTU, Zaporizhzhya, Ukraine (2014) (in Russian)
16. Internet of Things – Overview. http://www.codeproject.com/Articles/833234/Internet-of-
things-Overview
17. IoT Path to Product: Smart Home. http://www.codeproject.com/Articles/1119436/IoT-Path-
to-Product-Smart-Home
18. Parkhomenko, A., Gladkova, O., Kurson, S., Sokolyanskii, A., Ivanov, E.: Internet-based
technologies for design of embedded systems. J. Control Sci. Eng. 13(2), 55–63 (2015)
19. Parkhomenko, A., Gladkova, O., Ivanov, E., Sokolyanskii, A., Kurson, S.: Development and
application of remote laboratory for embedded systems design. iJOE 11(3), 27–31 (2015)
20. RELDES. http://swed.zntu.edu.ua
21. Tabunshchyk, G., Van Merode, D., Arras, P., Henke, K.: Remote experiments for reliability
studies of embedded systems. In: International Conference on Remote Engineering and
Virtual Instrumentation, Madrid, Spain, pp. 68–71 (2016)
22. Poliakov, M., Larionova, T., Tabunshchyk, G., Parkhomenko, A., Henke, K.: Hybrid models
of studied objects using remote laboratories for teaching design of control systems. iJOE
12(09), 7–13 (2016)
23. Gilibert, M., Picazo, J., Auer, M., Pester, A., Cusidó, J., Ortega, J.: 80C537 microcontroller
remote lab for e-learning teaching. iJOE 2(4), 1–3 (2006)
24. Guerra, H., Cardoso, A., Sousa, V., Gomes, L.: Remote experiments as an asset for learning
programming in python. iJOE 12(4), 71–73 (2016)
25. Ozvoldová, M., Ondrusek, P.: Integration of online labs into educational systems. iJOE 11(6),
54–59 (2015)
zamfira@unitbv.ro
330 A. Parkhomenko et al.
26. Wuttke, H.-D., Ubar, R., Henke, K.: Remote and virtual laboratories in problem-based
learning scenarios. In: IEEE International Symposium on Multimedia, Taichung, Taiwan, pp.
377–382 (2010)
27. Cvjetkovic, V., Matijevic, M.: Overview of architectures with Arduino boards as building
blocks for data acquisition and control systems. iJOE 12(7), 10–17 (2016)
28. Open HAB. http://www.openhab.org/
29. REST API. https://github.com/openhab/openhab/wiki/REST-API
30. FFmpeg. https://www.ffmpeg.org/
zamfira@unitbv.ro
Incorporating a Commercial Biology Cloud
Lab into Online Education
Ingmar H. Riedel-Kruse(&)
1 Introduction
A new paradigm has recently emerged for providing access to biology experimentation
through a distributed online platform known as cloud biology labs (Hossain et al. 2015;
2016; Lee et al. 2014; Hayden 2014). The notion is similar to the well-established
framework of cloud computing (Fox 2011) and complements ongoing advances in
life-science technology (Kong et al. 2012; Melin and Quake 2007), which have focused
mainly on automation and parallelization but have largely ignored issues related to
remote or shared access. This cloud lab technology reduces access barriers to costly
biological lab equipment and also abates the need for maintenance and hands-on
preparation, allowing users to concentrate on experimental design and data analysis.
Further advantages include reduced training needs, improved biosafety, facilitated data
tracking, and increased standardization (Sia and Owens 2015).
Two of these biology cloud labs have been successfully deployed recently in
academic settings for educational purposes: interactive chemotaxis experiments with
the slime mold physarum over the course of a day (Hossain et al. 2015), and real-time
interactive phototaxis experiments with the protist Euglena over the course of one
minute (Hossain et al. 2016); the latter lab in particular promises to scale at low cost to
massive user numbers (millions of students per year with a cost of less than 1 cent per
experiment). Recent years have also seen the emergence of dedicated biology cloud lab
companies with initial efforts largely focused on industrial applications (Hayden 2014);
however, there is an opportunity for partnership between commercial cloud lab pro-
viders and biology educators to teach lab biology via the web, which has not yet been
achieved.
There is a growing literature regarding the utility of educational online experiments
and how such remote labs should be designed (Lowe et al. 2012; de Jong et al. 2013;
Heradio et al. 2016; Wieman et al. 2008; Bonde et al. 2014; Sauter et al. 2013). Key
design principles include that students feel a real presence (such as via a live video) and
that the user interface is intuitive enough to effectively abstract the logistics of
preparation so that students can focus on experimental designs and strategies. The
development of large-scale cloud biology platforms could be instrumental in bringing
true laboratories into online education. The above cited research also compares real
labs (remote or local) to simulations (virtual labs), with the conclusion that both have
their situation-dependent advantages, and that ideally both are used synergistically.
The goal of the presented work is to assess whether and how commercial biology
cloud labs could be utilized for education. We partnered with the start-up company
Transcriptic (https://www.transcriptic.com/) to develop a customized cloud-based
biological experimentation activity for educational use. Transcriptic has been devel-
oping the Workcell platform (Fig. 1), in which a robot shuttles biological specimens in
96-well plates between experimental instruments such as liquid-handling robots,
imaging devices, and incubators. Experiments can be fully programmed in Python.
This overall framework is under constant development; for example, some experi-
mental steps are still executed by hand, but will eventually be automated. The vision
and roadmap to full and flexible automation of cloud experiments is clear. To assess the
affordance for future scale-up in educational settings, we deployed this platform in a
graduate class user study where students designed and analyzed their own experiments
to model the effects of antibiotics on bacterial growth curves. The main categories in
which we evaluated the system were (1) logistical feasibility and cost and (2) student
responses and potential educational outcomes.
2 Methods
All experiments were handled off-site by our commercial partner (Transcriptic, Inc.,
Menlo Park, CA). Transcriptics has developed a Workcell that automatically executes
all experiments (Fig. 1A). These experiments are executed in disposable, standard
96-well plates. A liquid handling robot mixes the corresponding solutions and dis-
tributes them among the individual wells, achieving the specific antibiotic concentra-
tions requested by the remote users. The Workcell shuttles these plates between
incubators and a plate reader that executes the measurements. This platform is con-
trolled via a Python-based framework. In order to enable student access, we developed
a Python-based user interface that enabled students to enter their experimental
instructions, read-off their results in graphical form, and download their experimental
data in csv format (Fig. 2A).
zamfira@unitbv.ro
Incorporating a Commercial Biology Cloud Lab 333
Fig. 1. Commercial biology cloud labs supported by robotics allow remote and controlled
execution of cellular and molecular biology experiments. (A) Workcell (Transcriptic)
automates the execution of life-science experiments in order to increase speed, ease, accuracy,
and reproducibility. This Workcell contains several instruments and a robot that moves the
sample plates among instruments. Image adapted from https://www.transcriptic.com/. (B) Top,
electron micrograph of bacteria. Bottom, schematic of a typical bacterial growth curve in which
optical density is measured as a proxy for bacterial concentration. Generally, four phases of
growth can be distinguished: lag phase, exponential phase, stationary phase, and decay phase.
(Image: Public Domain, Credit Rocky Mountain Laboratories, NIAID, NIH; Scale: individual
bacterium *2 lm in length.)
Fig. 2. User interface, cloud lab technology, and experimental data on bacterial growth.
(A) Website displayed to students after login. On the left, six different amounts of Kan can be
entered. On the right, the currently selected batch of data is displayed, updated every *40 min
during the experiment to generate a new recorded data point. Current and previous batch data can
be selected for display. (B) Top: Plate reader used by the collaborating company; a robotic
platform shuttles the plate between the incubator and spectrophotometer. Bottom, transparent
96-well plates in which experiments are executed. Each student is assigned to six wells during
each run, enabling up to 16 students to execute experiments in parallel. (Images: Thermo Fischer
Scientific)
zamfira@unitbv.ro
334 I.H. Riedel-Kruse
Bacterial growth experiments were performed in 96-well plates (Fig. 2B). At the
beginning of an experiment, a robotic pipetting assembly was programmed to load each
well with 150 µL of a suspension of Escherichia coli. The E. coli suspension was
prepared by diluting an E. coli stock (optical density at 600 nm (OD600) = 0.9) into
Miller’s Luria Broth (LB) at a 1:150 ratio. To generate the stock, the DH5a strain of
E. coli (Zymo Research Corp, US) was transformed with pUC19 plasmid. The robotic
pipettor then added the user-specified amount (0–50 µL) of the antibiotic kanamycin
(Kan; 40 µg/mL in LB). Each well was then filled to 200 µL with LB, yielding final Kan
concentrations of 0–10 µg/mL. A robotic setup then shuttled the plate to a spec-
trophotometer (Figs. 1 and 2) to measure and record the bacterial concentration of each
well using OD600. The plate was covered and placed in an incubator at 37 °C, with
shaking at 180 rpm. Every 47 min, the plate was shuttled back to the plate reader and
new OD measurements were recorded. This cycle was repeated continually overnight
for a total of 20 times to generate a full bacterial growth curve (Figs. 1, 2 and 3).
zamfira@unitbv.ro
Incorporating a Commercial Biology Cloud Lab 335
3 Results
Among the many possible experiments available on this platform, we chose bacterial
growth in the presence of different antibiotic concentrations, as it suited the content of a
college-level class (and was likely suitable for middle and high school) and had the
potential for easy adaptation on the existing Transcriptic platform. We developed a
custom web interface (Fig. 2A) that allowed students to run and evaluate experiments
to test the effects of antibiotic concentration on the growth of bacterial populations.
Each user accessed six wells; wells from up to 16 users could be combined for a
parallel run on a 96-well. For each of the six wells, the user specified varying amounts
of an antibiotic to add to the medium (from 0 to 10 µg/L Kan), which affected growth,
primarily leading to delayed onset of growth, slower growth, and lower maximal OD600
(Methods). Experiments ran overnight; the next day, users logged on to the web
interface to view and download the resultant bacterial growth curves (Fig. 2A).
Prior to the user study, we performed approximately 10 test runs. Overall, these
initial test experiments were stable and satisfactory, with the typical experimental
outcome leading to Hill-type functions (Figs. 1B, 2A and 3A). The OD600 began at a
low and reasonably steady value (OD600 = 0.2; lag phase), increased over time (5–
10 h; exponential phase), and eventually plateaued (OD600 = 0.2 – 0.5; stationary
phase), yielding the familiar sigmoidal curve associated with bacterial growth. The
somewhat high initial OD600 reading of 0.2 reflects the initial bacterial starting con-
centration as well as the fact that reading were not normalized against initial readings or
blanks. Maximal OD600 is dependent on the total volume of media in the plate, the well
size, culture aeration during growth, and the bacterial strain. In the presence of Kan, the
entry into exponential phase was delayed, leading to a lower maximal OD600, con-
sistent with the literature (Faraji et al. 2006; Lin et al. 2000). At the maximum Kan
concentration (10 µg/mL), growth was completely inhibited, yielding a flat growth
curve at OD = 0.2.
The platform was deployed in an advanced undergraduate/graduate-level class on
modeling multicellular systems. Thirteen students of both genders with various back-
grounds related to engineering, physics, and biology took the class. Some of the
students did not have any hands-on biology lab experience beyond what is standard in
high schools, while others had taken extensive biology lab courses or had previously
worked in wet labs including performing equivalent bacteria sample preparation and
OD measurements on bacterial growth.
As part of a lecture module on biological growth and competition, students were
provided with the theoretical background and pointed to review papers on how to
model bacterial growth under the influence of antibiotics. The students were then
introduced to the commercial cloud platform and given an open-ended homework
zamfira@unitbv.ro
336 I.H. Riedel-Kruse
Table 1. Distinct initial experimental strategies and answers from the post questionnaire.
Strategy Description Examples # students # students post Incl. 0 Incl. Sweep Sweep Replicate
during exp exp 50 linear log
1a A linear sweep through [0 10 20 30 40 50] 5 5 X X X
full parameter ranges
1b A logarithmic sweep [0 1 3 9 27 50] 3 1 X X X
through full parameter
ranges
1c a single logarithmic [2 4 8 16 24 48] 1 0 X
sweep (w/o 0, 50)
2a 2 triplets at the extremes [0 0 0 50 50 50] 1 3 X X X
2b 2 triplets, one at 0 and [0 0 0 10 10 10] 1 0 X X
one low
3 mix of control and [0 0 0 2 10 50] 1 1 X X X X
spanning the space
4a 6 times same average [20 20 20 20 20 20] 1 0 X
4b full blank control [0 0 0 0 0 0] 0 3 X X
zamfira@unitbv.ro
Incorporating a Commercial Biology Cloud Lab 337
zamfira@unitbv.ro
338 I.H. Riedel-Kruse
Fig. 4. Example data set and fitting approach from one student. (Images are direct
replications of student’s homework reports.) (Left) Data for six Kan amounts (dots) with fitting
curves (solid) superimposed. For the 0 µL Kan condition, multiple data repeats were collected;
for 25 µL Kan, two were recorded; all others are single runs. (Right) The student chose a
sigmoidal model with four independent parameters—note that these four parameters could have
been defined differently by using a different notation for the sigmoidal curve. The table shows the
fit parameters for each condition. A linear dependency on Kan amount is assumed (fitted) for
each of these four parameters, leading to eight parameters shown in the table in panel D. One
component is found to be zero, leaving the student with a 7-parameter fit to explain bacterial
growth under the influence of Kan, which leads to the fit curves in A.
zamfira@unitbv.ro
Incorporating a Commercial Biology Cloud Lab 339
Noisyness / reliability of
Are these experiments
Was the activity liked?
Order of questions on
Disagree(1-3)
a real experiment
Neutral(4-6)
Agree(7-9)
questionnaire
too simple?
mean
class?
data
Question
15 I enjoyed this activity. x 2 5 6 5.9
The user online interface was intuitive to use / appropriate
9 x 0 2 11 7.9
for the task.
2 The activity was worth the time and effort. x x 2 6 5 5.4
experimentation, modeling, and data fitting. Future controlled studies would evaluate
the extent to which these outcomes are met.
This post-questionnaire also solicited general student feedback via 17 questions
summarized in Table 2. These questions can be loosely clustered into five aspects. The
first three aspects targeted general student impressions of whether they thought the
activity was useful/appropriate/engaging. The last two aspects focused on whether
there was substantive value in having run real experiments with real noise.
The majority of the class was neutral or positive as to whether they liked the
assignment, and also in whether they felt it was a valuable addition to the class and
reinforced course concepts. Students clearly found the web interface to be intuitive.
Overall, students were neutral in whether this activity was appropriate for the class
level, with many believing that the activity was too simple and lacked experimental
freedom. Some students reported that the noisiness of the real data provided an added
zamfira@unitbv.ro
340 I.H. Riedel-Kruse
benefit, while others did not. Students preferred the freedom in model choices and the
ambiguity of the data. Students had high trust in the system before starting the
experiment, but became more critical about the instrument throughout the assignment,
which is an important lesson for any experimentalist. Overall, students interpreted the
variability in their data as stemming primarily from issues with instrumentation rather
than biological noise. Overall, we conclude a subpopulation of students appreciated this
online lab and had a positive educational outcome.
4 Discussion
zamfira@unitbv.ro
Incorporating a Commercial Biology Cloud Lab 341
class, which was also indicated by student feedback. There was essentially only one
variable that students could choose: the amount of antibiotic. Students would have pre-
ferred more experimental freedom. We therefore believe that this set of activities would be
better suited to more introductory research-focused classes, perhaps at advanced
high-school or introductory college levels, to train students on experimental design and
data analysis while taking away the time consuming efforts for a hands-on lab.
Overall, the Transcriptic platform enables much more complex experiments for
future deployments given its professional target audience. These experiments could be
made more interactive and versatile, such as being able to choose from multiple
antibiotics or adding liquids to the sample at multiple times during the experimental
run. Many other experiments are possible, given that the ultimate vision of these cloud
lab companies is to enable (nearly) every possible experiment in the molecular and
cellular research domain. Finally, another interesting aspect is that students can access
research-grade equipment over the web.
Based on Transcriptic’s business model, the current cost of these of experiments
is *$70 per 96-well plate. This cost depends on the experiment type and is likely to
decrease in the future given advancements in the technology. Hence, running five
successive experiments of six wells per student might be considered reasonable for this
activity, which would cost *$20 per student, a price point that can be considered
reasonable in comparison to advanced hands-on lab classes in colleges, but is poten-
tially at the upper limit for K-12 education. One of the major advantages of such a
commercial cloud lab approach is that all the costs are already factored in, with no
additional logistics for the instructor. The 96-well experiment also demonstrates how
high-throughput experiments in general can be virtually partitioned among many users.
Given the size of the educational market, with millions of students in the US alone
going through the same curriculum each year, offering cloud biological experiments
could be of interest to commercial cloud lab companies.
We asked our collaborators at Transcriptic for their evaluation of the project. They
indicated that it was an important educational experience for them as well, especially as
the Workcell had just gone operational and was still in the debugging phase. The
variability that had emerged during the experiments was unintended, but also provided
valuable insights into where the system and the protocols needed improvements. These
issues were subsequently resolved through a combination of improved hardware and
specimen-handling protocols, such as purchasing more advanced liquid handling robots
and avoiding condensation on the plates.
It is also important and insightful to compare these results to the two other biology
cloud labs that were deployed in educational settings previously (Hossain et al. 2015),
where students performed chemotaxis experiments with the slime mold physarum over
the course of one day (Hossain et al. 2015) or phototaxis experiments with Euglena
cells over the course of one minute (Hossain et al. 2016). The bacterial growth
experiment investigated here provided the students with a significantly smaller design
space for their experiments (exhaustive exploration within five experiments vs. effec-
tively having an infinite space in the other two cases). Similarly, the result space was
much more limited (discrete data curves vs. rich image data, although the biological
variability in the growth curves added interesting elements). The Workcell approach of
shuttling experiments between different types of instruments provides a tremendous
zamfira@unitbv.ro
342 I.H. Riedel-Kruse
References
Bonde, M.T., Makransky, G., Wandall, J., Larsen, M.V.: Improving biotech education through
gamified laboratory simulations. Nature 32, 694–697 (2014)
Buchanan, R.L., Whiting, R.C., Damert, W.C.: When is simple good enough: a comparison of
the Gompertz, Baranyi, and three-phase linear models for fitting bacterial growth curves.
Food Microbiol. 14(4), 313–326 (1997). 1–14
de Jong, T., Linn, M.C., Zacharia, Z.C.: Physical and virtual laboratories in science and
engineering. Education 340(6130), 305–308 (2013)
zamfira@unitbv.ro
Incorporating a Commercial Biology Cloud Lab 343
Faraji, R., Parsa, A., Torabi, B., Withrow, T.: Effects of kanamycin on the macromolecular
composition of kanamycin sensitive Escherichia coli DH5a strain. J. Exp. Microbiol.
Immunol. (JEMI) 9, 31–38 (2006)
Fox, A.: Cloud computing–what’s in it for me as a scientist? Science 331(6016), 406–407 (2011)
Hayden, E.C.: The automated lab. Nature 516, 131–132 (2014)
Heradio, R., de la Torre, L., Galan, D., Cabrerizo, F.J., Herrera-Viedma, E., Dormido, S.: Virtual
and remote labs in education: a bibliometric analysis. Comput. Educ. 98(C), 14–38 (2016)
Hossain, Z., Chung, A.M., Riedel-Kruse, I.H.: Real-time and turn-based biology online
experimentation. In: Remote Engineering and Virtual Instrumentation (REV) (2015)
Hossain, Z., Blikstein, P., Riedel-Kruse, I.H., Jin, X., Bumbacher, E.W., Chung, A.M., et al.:
Interactive cloud experimentation for biology. Presented at the the 33rd Annual ACM
Conference, pp. 3681–3690. ACM Press, New York (2015). http://doi.org/10.1145/2702123.
2702354
Hossain, Z., Chung, A.M., Baumbacher, E., Lee, S.A., Honesty, K., Walter, A., et al.: A real-time
interactive, scalable biology cloud experimentation platform. Nature Biotechnol. 34(12),
1293–1298 (2016)
Kong, F., Yuan, L., Zheng, Y.F., Chen, W.: Automatic liquid handling for life science: a critical
review of the current state of the art. J. Lab. Autom. 17(3), 169–185 (2012)
Lee, J., Kladwang, W., Lee, M., Cantu, D., Azizyan, M., Kim, H., et al.: RNA design rules from
a massive open laboratory. Proc. Natl. Acad. Sci. 111(6), 2122–2127 (2014)
Lin, J., Lee, S.-M., Lee, H.-J., Koo, Y.-M.: Modeling of typical microbial cell growth in batch
culture. Biotechnol. Bioprocess Eng. 5(5), 382–385 (2000)
Lowe, D., Newcombe, P., Stumpers, B.: Evaluation of the use of remote laboratories for
secondary school. Sci. Educ. 43(3), 1197–1219 (2012)
Melin, J., Quake, S.R.: Microfluidic large-scale integration: the evolution of design rules for
biological automation. Ann. Rev. Biophys. Biomol. Struct. 36, 213–231 (2007)
Sauter, M., Uttal, D.H., Rapp, D.N., Downing, M., Jona, K.: Getting real: the authenticity of
remote labs and simulations for science learning. Distance Educ. 34(1), 37–47 (2013)
Sia, S.K., Owens, M.P.: Share and share alike. Nat. Biotechnol. 33, 1224–1228 (2015)
Swinnen, I.: Predictive modelling of the microbial lag phase: a review. Int. J. Food Microbiol.
94(2), 137–159 (2004)
Wieman, C.E., Adams, W.K., Perkins, K.K.: PHYSICS: PhET: simulations that enhance
learning. Science 322(5902), 682–683 (2008)
Zwietering, M.H., de Koos, J.T., Hasenack, B.E., de Witt, J.C., van’t Riet, K.: Modeling of
bacterial growth as a function of temperature. Appl. Environ. Microbiol. 57(4), 1094–1101
(1991)
zamfira@unitbv.ro
Learning to Program in K12 Using a Remote Controlled
Robot: RoboBlock
1 Introduction
The promotion of STEM among the young people is one of the objectives of different
countries and institutions like EU. To increase the interest of the young people to science,
engineering and technology the schools are including subjects that combine program‐
ming and robotics [1, 2]. The main effect of these initiatives is in K12.
In general, the teachers in the classrooms have to deal with different problems.
Firstly, they need financial resources to buy several robots. Secondly, they have to
maintain them in perfect conditions (even after being used by students). These two issues
reduce the time that they have to teach the students how to program. It is exciting for
the students to program and control real robots, but at the same time it is frustrating for
the students to see that their robot is not running properly but the robot of other group
is running well, with the same program!! At the same time the teacher has to manage
this situation, and he could also be frustrated or he could be very busy reviewing the
boards, the sensors, etc. This problem is especially remarkable at the beginning of the
learning process because the students have more problems with the equipment and with
the programming language. The learning curve is in Fig. 1.
zamfira@unitbv.ro
Learning to Program in K12 Using a Remote Controlled Robot 345
At the same time, and at least in Spain, robotics use to be taught in the Technology
subject and the associated teacher does not use to be a computer or electronics engineer.
The common situation is that the Science teacher must to teach programming/robotics
because he needs to fill his timetable. Currently it is not common to have a specific
teacher for technology and programming.
To solve this situation or at least to help teachers and institutions to manage this
situation a robot can be designed and used as a remote experiment. RoboBlock will help
the community to solve these problems using a web tool to complete all the teaching
process. There are several remote laboratories offering robots around the world but not
with the same characteristics that RoboBlock [3, 4].
A remote laboratory allows the user/student to experiment using the Internet as his eyes
and hands. Instead of using the hands to manipulate the equipment and instead of using
the eyes to see the evolution of the experiment, the user will use the Internet. To make
an experiment the user only needs an Internet connection, so he can experiment every‐
where and every time he has an Internet connection.
To deploy a remote laboratory into a K12 classroom there are some requirements [5]:
• Universality: the remote laboratory can be accessed using any OS and any web
browser.
• Device: the remote laboratory can be accessed using any kind of device, including
tablets and smartphones [6].
• Security: the remote laboratory cannot affect the IT security of the school, so it is
based only in standard ports (80, …) and without problems with firewalls.
• LMS: the remote laboratory can be connected to the LMS of the school.
• Queue: the remote laboratory manages the users’ queue.
If the previous requirements are not fulfilled by the remote laboratory, then it could
not be implemented successfully in the classroom. RoboBlock fulfills these technical
requirements. But there are also some functional requirements.
zamfira@unitbv.ro
346 J. García-Zubía et al.
zamfira@unitbv.ro
Learning to Program in K12 Using a Remote Controlled Robot 347
Fig. 2. ArduBlock
Fig. 3. ArduLab
In the previous section, RoboBlock has been described in general terms, and this section
is devoted to describe what can be done with RoboBlock and how it can be done.
Universality, security, etc. are key factors of any IT learning tool in the school, but
they are not important for the teacher if the tool is not powerful enough. RoboBlock can
be explained using the programming blocks and to the programs themselves.
zamfira@unitbv.ro
348 J. García-Zubía et al.
In a first group (see Fig. 5), in Loops the student can select blocks to implement:
loop, while, count and break. In the Logic area the available blocks are: if, numerical
comparison, logical comparison and other blocks. In the Mathematics area we can find:
arithmetical operations (basic and advanced), constants, and other functions. There is a
block for timing functions: wait, measure, etc. There are other blocks for text, commu‐
nications, etc.
For programming RoboBlock, it is important to have powerful blocks for sensoring.
It is important to read from sensors the current situation of the robot, and it is important
to write in the actuators to modify the situation of the robot, everything in real time.
There are three proximity sensors placed in the center of the robot, in the right side and
in the left side, but also every sensor has two parts: left and right; by this way the control
of RoboBlock can be very accurate. The same situation occurs with line detector sensor:
there is one sensor in the center of RoboBlock but it reads five signals. RoboBlock has
zamfira@unitbv.ro
Learning to Program in K12 Using a Remote Controlled Robot 349
also three buttons to perform different tasks, in this case the programmer can scan the
buttons to read pulses and levels.
The student can also control RoboBlock using the actuators of the two motors, left
and right. Each motor has an associated speed from 0 to 100.
This is the current state of RoboBlock, but additional sensors or actuators can be added
to it, and in this case new programming blocks should be added. The design of web
programming tool, ArdulabBlockly, is intended to allow its modification easily to promote
the adaptation of RoboBlock to each scenario in different schools and events (Fig. 6).
Fig. 6. Blocks for programming of RoboBlock: loops, time, and numerical and logical functions.
zamfira@unitbv.ro
350 J. García-Zubía et al.
The student, after writing/drawing the code, can validate his code clicking on Vali‐
date (Validar) to see if there is any problem with his program. After doing this, the
student will open the Zumoline environment to download his program in RoboBlock.
Zumoline knows what is the last program validated by the user (called blocks), and this
program will be downloaded in RoboBlock. The student can also access and test to a
variety of running examples.
RoboBlock is designed to develop the basic programming skills using a robot. The
student can create programs combining basic blocks like loop, if-then-else, assignation,
read, write, etc. It is up to the teacher to decide the complexity level of the programs
focusing to the curricula of the subject (Fig. 8).
zamfira@unitbv.ro
Learning to Program in K12 Using a Remote Controlled Robot 351
For example, the program in Fig. 9 controls the robot increasing its speed one by
one from 0 to 50 every 10 ms. In this case the student practice the for structure using
the robot to see the effect of it.
In Fig. 10 the algorithm shows the student how to use the if structure. RoboBlock
will go faster (one by one, from 0 to 50 every 10 ms) until detecting a line in the floor
using the sensor.
zamfira@unitbv.ro
352 J. García-Zubía et al.
As final project the teacher can purpose the students a competition like “Win in a
Formula one circuit”. The circuit made with plastic in Fig. 11 is a copy of the Montmelo
Circuit in Spain, and the students should write the best program to obtain the lowest
time to tour the circuit.
Every time the student writes a new program, he can test it in the remote laboratory,
in RoboBlock. At this moment to download the code it is needed to open a new tab in
zamfira@unitbv.ro
Learning to Program in K12 Using a Remote Controlled Robot 353
the web browser, the Zumoline environment. When a student accesses to Zumoline he
has 5 min (it can be increased) to select his program, download it in RoboBlock and see
the evolution of the robot. After this, maybe he can modify the program to be just a bit
faster in the competition. In Fig. 12 there is a simple program to win in Momtmelo that
the student can improve adding a better control.
The development of the laboratory can be divided into three stages. Firstly, a Python
module performs remotely the compilation of source files to generate the binary files to
be loaded into the program memory of the microcontroller. Secondly, a programming
environment has been developed to capture algorithms with Blocky. Finally, a system
has been developed that allows the programming of the mobile robot and the interaction
with its inputs, so that these can be accessed remotely.
zamfira@unitbv.ro
354 J. García-Zubía et al.
zamfira@unitbv.ro
Learning to Program in K12 Using a Remote Controlled Robot 355
The main element of the experimentation platform is Pololu Zumo robot 32u4. This
robot is based on the microcontroller compatible with Arduino Atmega32u4. It includes
the following electronic and mechanical components.
Electronic components:
• 4 wide-angle infrared sensors
• 2 DC motors
• 2 encoders
• 5 line sensors matrix
• 3-axis Gyroscope
• 3-Axis Accelerometer
• Magnetometer
• Buzzer
• Bootloader
Chassis:
• 4 sprockets
• 2 tires for crawler type wheels
• Main Chassis
• Metal shovel
Raspberry Pi embedded platform has been used to develop the controller. This device
connects to the robotic platform performing the following functions:
• Bootloader activation
• Programming the robot
• Access to serial communication port
• Activation of buttons remotely
• Reading the status of the outputs
The Zumo robot 32u4 includes several expansion ports that allow to replace the
sensors and actuators of which it has natively by others. In this case, this expansion port
has been used to connect certain outputs of the Raspberry Pi in it. Since the connections
are made directly, it has not been necessary to develop an extra electronic circuit.
At the software level, an object has been implemented that contains the methods
necessary to perform the actions listed above. These methods have been developed in
Python language and are invoked directly from the web application. The methods
included in the driver are as follows:
• ledChecker (): This routine continually checks the status of the microcontroller
outputs that activate/deactivate the robot LEDs.
• startLedChecker (): Method to start the status check routine of the LEDs.
• stopLedChecker (): Method to finalize the status check routine of the LEDs.
• startSerial (): Method to initialize the serial port.
• connectSerial (): Opens the serial port and starts the serialTask () routine.
• serialTask (): This routine checks if data has been received by the serial port and
transmits it to the connected client.
zamfira@unitbv.ro
356 J. García-Zubía et al.
• sendSerial (message): This method is invoked when the client sends data through the
serial port to the robot.
• enableBootloader (): This method is used to activate the bootloader and load a
program.
• eraseMemory (): This method erases the entire program memory of the microcon‐
troller.
• loadBinary (path): This method is used to load a binary on the robot. It is necessary
to mark the path of the binary as a parameter.
• buttonOn (button): This method activates any physical button on the robot remotely.
• buttonOff (button): This method disables any physical button on the robot remotely.
• Queues. At this moment there is only one copy of RoboBlock, so if a student wants
to access the RoboBlock but it is being used by other student, the first student must
be in a queue until RoboBlock is released. If there are several schools using
zamfira@unitbv.ro
Learning to Program in K12 Using a Remote Controlled Robot 357
RoboBlock more situations like this can happen. WebLab-Deusto manages the queue
and can assign different priorities to each school or student, or maybe the teacher will
have a higher priority than the students to grant him to access the robot in a short
time.
• Learning analytics. WebLab-Deusto tracks all the activity of the student, so the
teacher can see for every student how many times and when he entered the platform.
Even more WebLab-Deusto records the files that the student downloaded during the
semester. An additional tool allows the teacher to see if there are files copied among
the students.
• Scalability and load balance. If there was different copies of the RoboBlock,
WebLab-Deusto would manage the situation sharing the effort among the different
copies of the robot.
• Federation. A school can implement a copy of RoboBlock in its building, and in this
case it can be federated in WebLab-Deusto. In this situation, instead of having only
one copy of RoboBlock, we will have two copies. The federation mechanism imple‐
mented in WebLab-Deusto allows the school to fix the conditions of use of its own
RoboBlock, and even the school can earn some money from this federation.
RoboBlock is designed and implemented to help the schools and teachers to teach
programming and robotics to K12 students. This tool will help to promote STEM among
young students.
RoboBlock is a remote experiment and it can be controlled remotely in ZumoLine
by the student using the algorithm written by him in ArduBlocks using a graphical tool
based in Blockly. All the tasks are performed in the web, so if the student has an Internet
connection he can complete his practical activity using any computer in any place at
every time. RoboBlock is a novelty in the K12 scope.
The design of RoboBlock has been recently finished and it has been deeply tested,
and it is time to implement it in several schools in a pilot. The results of the pilot will
help the designers to improve the original design of RoboBlock.
References
1. Roscoe, J.F., Fearn, S., Posey, E.: Teaching computational thinking by playing games and
building robots. In: International Conference on Interactive Technologies and Games (iTAG),
pp. 9–12 (2014)
2. Merrill, M.D.: First principles of instruction. Educ. Technol. Res. Dev. 50(3), 43–59 (2002)
3. Islamgozhayev, T.U., et al.: IICT-bot: educational robotic platform using omni-directional
wheels with open source code and architecture. In: International Siberian Conference on
Control and Communications (SIBCON), pp. 1–3 (2015)
4. Antonio, C.P., et al.: Remote experiments and 3D virtual world in education. In: 3rd
Experiment@, International Conference, exp’at 2015, pp. 65–70 (2015)
5. García-Zubía, J., Orduña, P., López-de-Ipiña, D., Alves, G.: Addressing software impact in
the design of better remote labs. IEEE Trans. Ind. Electr. 56(12), 4757–4767 (2009)
zamfira@unitbv.ro
358 J. García-Zubía et al.
6. García-Zubía, J., López-de-Ipiña, D., Orduña, P.: Mobile devices and remote labs in
engineering education. In: Proceedings of 8th IEEE International Conference on Advanced
Learning Technologies, ICALT 2008, pp. 620–622 (2008)
7. Guimaraes, E., Cardozo, E., Moraes, D., Coelho, P.: Design and implementation issues for
modern remote laboratories. IEEE Trans. Learn. Technol. 4(2), 149–161 (2011)
8. Sarik, J., Kymissis, I.: Lab kits using the Arduino prototyping platform. In: 2010 IEEE
Frontiers in Education Conference, FIE 2010, Washington, DC, pp. T3C-1-T3C-5 (2010)
9. Woodring, I., El-Said, M.: An economical cluster based system for detecting data leakage
from BYOD. In: 2014 11th International Conference on Information Technology, pp. 610–
611. New Generations, Las Vegas (2014)
10. Perate, C.: Ardublockly (2016). https://github.com/carlosperate/ardublockly
11. Pimentel, V., Nickerson, B.G.: Communicating and displaying real-time data with
WebSocket. IEEE Internet Comput. 16(4), 45–53 (2012)
zamfira@unitbv.ro
Spatial Learning of Novice Engineering Students Through
Practice of Interaction with Robot-Manipulators
1 Introduction
The ways to increase the efficiency of learning practices in Robotics and Computer
Integrated Manufacturing (RCIM) laboratories are widely discussed [1]. When
educating unprepared students, the recommended lab practice is that which combines
training technical skills with learning the principles of robot operation and development
of generic skills required in different workplaces [2]. Among the most important of these
is the ability of spatial vision. Industrial robotics laboratories generally implement three
types of learning scenarios [3]: setting up a robot system, programming different indus‐
trial robots, and performing advanced robot-handling tasks. The laboratories offer
learning practice in hands-on, virtual, and remote environments.
To perform robot system setup, programming and operation assignments, the student
needs immediate and detailed visual information from the robot workspace. In the hands-
on environment the student is near the robot system and so all needed information is
acquired directly through observation. In the remote control system visual feedback is
transmitted from video cameras via a computer screen, and so is incomplete and delayed.
In the virtual environment the student works with a graphic simulation of the robot
system on the computer screen under limitations of the given software. The advantages
and constraints of the hands-on, virtual, and remote learning practices have been
discussed and compared in the literature [4]. Less attention has been paid to the analysis
of difficulties that students face while performing tasks in robotic environments, and to
the impact of this practice on the development of fundamental engineering skills,
including spatial skills [5].
zamfira@unitbv.ro
360 I. Verner and S. Gamer
The current paper reports on the results of our study conducted in the RCIM Labo‐
ratory of the Faculty of Industrial Engineering and Management (IEM) at the Technion–
Israel Institute of Technology. Over four academic years (2011–2015) we ran in the
laboratory robotics workshops for IEM first-year undergraduates and, separately,
outreach robotics courses for 10th-grade students in an underprivileged vocational high
school. Both sets of courses offered learning practice in programming and operation of
robot manipulators, while the tasks focused on training spatial skills. Details of our study
are presented in [6].
The RCIM Laboratory in the Technion’s IEM Faculty conventionally supports courses
and research activities for industrial engineering majors by enabling hands-on experi‐
mentation in the design, control and operation of automated manufacturing systems. The
laboratory facilities include nine semi-industrial robots. In terms of software, the lab is
equipped with the RoboCell.
zamfira@unitbv.ro
Spatial Learning of Novice Engineering Students 361
zamfira@unitbv.ro
362 I. Verner and S. Gamer
This robotics course was designed at the request of a vocational high school to help 10th
graders majoring in mechanical engineering who were having spatial difficulties
mastering technical drawing. The 16-hour course consisted of eight two-hour sessions.
The curriculum was divided into three parts, where each part focused on a certain aspect
of robot programming and operation, and on training one of the main categories of spatial
ability: spatial perception, mental rotation, and visualization.
The first three sessions focused on robot pick-and-place operations and spatial
perception tasks. In the first session the students learned about the structure of the robotic
arm and its motion in the workspace. In the second and third sessions they studied the
robot control language ACL, learned to define robot positions by coordinates, and prac‐
ticed programming simple pick-and-place manipulations with cubic parts. The next three
sessions dealt with rotation of objects by the robot. In the fourth session the students
learned about rotations around coordinate axes and how to perform them using the
robotic arm. In the fifth and sixth sessions, they learned to use the RoboCell software
and to operate a robot in the virtual environment. They completed this module by
assembling a six-cube picture puzzle from identical cubes with geometrical symbols
drawn on their sides (Fig. 2). The seventh and eighth sessions were devoted to
performing three assembly tasks with real robots. The first task was to assemble a six-
cube picture puzzle through teleoperating the robotic arm based on visual feedback from
zamfira@unitbv.ro
Spatial Learning of Novice Engineering Students 363
two digital cameras. In the second task the students were required to assemble a puzzle
from six identical cubes with geometric figures drawn on their sides. The puzzle was
presented using three orthographic projections (front, top, and side views) and a sketch.
The students were asked to use the sketch to depict a three-dimensional view of the
puzzle by drawing appropriate geometric symbols on the sides. They then had to
assemble the puzzle using the robot.
5 Robotics Workshop
The 6-hour workshop was delivered to first-year students as part of the Introduction to
Industrial Engineering and Management course. The workshop included a 2-hour lecture
and two 2-hour robotics lab classes. The lecture “Principles of Robot System Operation”
introduced the students to the concepts of CIM, robot programming, and robot operation.
The lecture also presented the lab assignments. The first laboratory class was devoted
to practice in the RoboCell virtual environment. The students were assigned to program
a 5 degrees-of-freedom robot to assemble a structure from different blocks. In the second
laboratory class the students operated real robots. The task required to operate the robot
so as to pick up an oriented cube, move it from the storage area to the buffer, rotate it to
the final orientation, and place it in the destination position at the assembly area. The
students planned and operated robot movements using predefined positions of the
mechanical arm and subroutines implementing basic pick-and-place operations (written
in the Advanced Control Language).
The evaluation study involved twenty high school students participated in the course
and 248 university students participated in the workshop. We evaluated whether the HS
students improved their performance in spatial tasks following the laboratory practice
in operating robot manipulators. The objective of the university workshop was to expose
first-year students to industrial robotics and foster awareness about spatial challenges in
programming and operating robots. Therefore, in this case our evaluation addressed the
development of spatial awareness.
Evaluation of the outcomes of the university workshop was in line with its objective:
to expose first-year students to industrial robotics and foster their interest and awareness
about spatial challenges in programming and operating robots. Awareness is defined as
individual’s consciousness of something to the degree that it can influence her/his
behavior [13]. Raising interest in industrial engineering and fostering awareness of its
professional requirements, particularly spatial awareness, is one of the core missions in
educating novice IEM students. Therefore, in the evaluation our interest was whether
the practice in operating robot manipulators improve the students’ awareness of spatial
skills in industrial robotics.
zamfira@unitbv.ro
364 I. Verner and S. Gamer
Students note the advantages of the spatial practice in the virtual environment:
The virtual lab lets you perform operations with the robot without fear that something will break
or go wrong.
zamfira@unitbv.ro
Spatial Learning of Novice Engineering Students 365
The virtual lab made it easier to understand considerations in planning robot operations: calcu‐
lating angles, heights and positions.
Evaluations of the spatial practice with real robots were even higher:
The physical lab was much more interesting, since it was a new work environment. The challenge
was to think how to accomplish the task in the most effective way.
The difficulties noted by the students related to the following spatial tasks: determination
of the height of the robot gripper above the working surface, use of coordinates of the
robot arm and their calculation, and collisions the arm with objects in the workspace,
while performing manipulations. From students’ reflections:
It was difficult to estimate distances between objects in the virtual environment.
Cube rotation tasks were complex and required spatial thinking
In response to our request to compare the contributions of the virtual and physical labs
the students did not strongly favor one over the other. Rather, their responses suggest
that both platforms serve important functions:
In the virtual lab it is easier to understand the thinking behind operating the robot, calculating
angles, heights and locations.
The physical lab better demonstrates the robot workspace and gives an idea of the production
process.
7 Conclusion
In this paper we presented our experience in adapting the Technion Robotics and
Computer Integrated Manufacturing Laboratory for introductory engineering courses.
We engaged first year IEM students in robotics activities and opened the laboratory to
high school students majoring in mechanical engineering.
We build the courses on the educational opportunities afforded by placing students
in the loop of robotic system, focusing their practice in the RCIM lab on understanding
the principles of robot operation, fostering spatial skills and awareness of their impor‐
tance in industrial robotics. This practice is crucial for novice engineering students who
are choosing the future profession. The key features of our approach are:
• Customizing the robot workspaces to enable performance of spatial operation tasks.
• Combining practice in direct, virtual and remote robot operation.
• Extending the robotic environment to enable the manipulation of oriented blocks.
• Directing robot operation tasks to train spatial skills.
We implemented and evaluated the developed approach in our RCIM lab for engi‐
neering novices of two categories: high school students majoring in mechanical engi‐
neering and first-year IEM students. As found, the high school students in the course
improved significantly in perception, mental rotation, and visualization tests. In the case
of IEM students, the workshop provided the first-hand experience in operation of real
and virtual robots, helped to understand the spatial problems dealt with by industrial
engineers and recognize the skills needed to cope with them.
zamfira@unitbv.ro
366 I. Verner and S. Gamer
Based on the results of our study, obtained under specific conditions, we argue for
further exploration of the proposed approach and call for using robotic environments
for training spatial skills that are highly demanded in engineering education and practice.
Acknowledgment. This study is supported by the Israel Science Foundation. We appreciate help
of Technion and school instructors: Dr. Assaf Avrahami, Niv Krayner, Elena Baskin and Ronny
Magril.
References
1. Munro, D.: Development of an automated manufacturing course with lab for undergraduates.
In: IEEE Frontiers in Education Conference, pp. 496–501 (2013)
2. Bien, Z.Z., Lee, H.-E.: Effective learning system techniques for human–robot interaction in
service environment. Knowl.-Based Syst. 20(5), 439–456 (2007)
3. Chang, G.A., Stone, W.L.: An effective learning approach for industrial robot programming.
In: ASEE Annual Conference and Exposition, Atlanta, US (2013)
4. Ma, J., Nickerson, J.V.: Hands-on, simulated, and remote laboratories: a comparative
literature review. ACM Comput. Surv. 38(3), 1–24 (2006)
5. Goldstain, O., Ben-Gal, I., Bukchin, Y.: Evaluation of telerobotic interface components for
teaching robot operation. IEEE Trans. Learn. Technol. 4(4), 365–376 (2011)
6. Shin, D., Wysk, R.A., Rothrock, L.: A formal control-theoretic model of human-automation
interactive manufacturing system control. Int. J. Prod. Res. 44(20), 4273–4295 (2006)
7. Lathan, C.E., Tracey, M.: The effects of operator spatial perception and sensory feedback on
human-robot teleoperation performance. Presence 11(4), 368–377 (2002)
8. Menchaca-Brandan, M.A., Liu, A.M., Oman, C.M., Natapoff, A.: Influence of perspective-
taking and mental rotation abilities in space teleoperation. In: Proceedings of the 2nd ACM/
IEEE International Conference on Human-Robot Interaction, pp. 271–278. ACM (2007)
9. Stanney, K.M., Cohn, J., Milham, L., Hale, K., Darken, R., Sullivan, J.: Deriving training
strategies for spatial knowledge acquisition from behavioral, cognitive, and neural
foundations. Mil. Psychol. 25(3), 191–205 (2013)
10. Intelitek: RoboCell for Controller USB. http://www.intelitekdownloads.com/Manuals/
Robots/ER-4u/100346-G%20RoboCell-USB-v56.pdf. Accessed 22 Mar 2015
11. Uttal, D.H., Meadow, N.G., Tipton, E., Hand, L.L., Alden, A.R., Warren, C., Newcombe,
N.S.: The malleability of spatial skills: a meta-analysis of training studies. Psychol. Bull.
139(2), 352–402 (2013)
12. Verner, M.: Robot manipulations: a synergy of visualization, computation and action for
spatial instruction. Int. J. Comput. Math. Learn. 9(2), 213–234 (2004)
13. Bower, G.: Awareness, the unconscious, and repression: an experimental psychologist’s
perspective. In: Singer, J. (ed.) Repression and Dissociation, pp. 209–231. University of
Chicago Press, Chicago (1990)
14. Eliot, J., Smith, I.: An International Directory of Spatial Tests. NFER-Nelson, Atlantic
Highlands (1983)
zamfira@unitbv.ro
Concurrent Remote Group Experiments in the Cyber
Laboratory
A FPGA-Based Remote Laboratory in the Hybrid Cloud
Nobuhiko Koike ✉
( )
Abstract. With the advent of M2M and IoT, it becomes important for the educa‐
tional remote laboratory to realize group M2M/IoT experiment environments,
where a number of group experiments are concurrently carried out by making use
of LAN-connected FPGA devices. Docker containers are employed to realize
separate FPGA-Run service environments, corresponding to every FPGA
devices. The Cyber laboratory can contain hundreds of FPGA evaluation boards
and FPGA-Run service containers. Each of those pairs is allocated to one of
twenty laboratory servers together. The Docker Swarm is also adopted to realize
multi FPGA group experiments by allocating a set of FPGA board and FPGA-
Run service container pairs. Each FPGA-Run Service container consists of a Web
server application, a Web-camera motion, the FPGA-run application and associ‐
ated individual FPGA device driver. A combination of the container and the
corresponding FPGA-board pair realized a separate FPGA run service virtual
machine. Newly designed gang scheduler issues a set of the Web services to start
a group experiment together. By making use of Docker volume plugins, FPGA-
run results and recorded videos can be sent to the common faculty data base for
post experiment analysis. The use of inexpensive public cloud enables to offload
most private cloud side workloads and to be migrated to public cloud. It realizes
an easy scale out or shrinking functionalities. The hybrid cloud organization and
the use of many FPGA-boards together with associated containers realized an
efficient sharing of servers and the FPGA-devices. The use of the Web services
and the Docker Swarm manager allow a flexible and easy device allocation/gang
scheduling and initiation of group experiments. The paper showed the Cyber
laboratory’s applicability for M2M and IoT kinds of remote experiments.
1 Introduction
Recent advancement in IoT and M2M technologies motivated the author to implement
the concurrent remote group experiment functionalities for the existing Cyber Labora‐
tory [2, 3], which is an educational FPGA-based remote laboratory in the hybrid cloud.
zamfira@unitbv.ro
368 N. Koike
The new configuration allows students to conduct IoT or M2M related remote experi‐
ments independently or collaboratively among students by making use of available
FPGA boards. If the number of available FPGA boards is insufficient, those group
experiments are kept in the waiting queue until sufficient FPGAs become available.
Those experiment services are handled in both space division and time division fashions.
Although, the former work [4, 5] already realized an online FPGA device remote
experiment, the exclusive use of FPGA board resulted in poor use of FPGAs and labo‐
ratory platforms. It also has no support for IoT or M2M kind of experiments, where
number of IoT/FPGA devices as well as edge/cloud servers should work together. In
order to realize such IoT kind of remote experiments, the remote laboratory should
realize a group remote experiment environment that contains number of FPGA devices
and quite a few edge or cloud servers connected by the internet.
Thanks to the advancement in device technologies, an affordable remote laboratory,
which contains hundreds of FPGA evaluation boards with Web cameras allocated over
twenty laboratory servers, can be constructed at a nominal cost. As the use of the Docker
containers associated with each FPGA board/Web camera pairs [1] contributed to realize
a lightweight virtual machine, each laboratory server can easily handle quite a few these
containers. It is also possible to add additional IoT edge server containers, these mimic
IoT or M2M server side applications in the experiments. The use of the Docker Swarm
makes it possible to construct a light weight on premise private cloud, where hundreds
of FPGA experiment platform and container pairs become available as remote experi‐
ment platforms.
The Docker Swarm manager/job scheduler can allocate FPGA experiment platforms
over twenty laboratory servers both in space division and time division fashions. For an
IoT experiment, edge server containers can also join the group experiment to realize a
network-connected group remote experiment environment,
The design has been completed and the prototype of single FPGA-evaluation board
version is operational. The job entry/dashboards, rendezvous/gang scheduling for
students and administrators are under construction. The hybrid cloud organization and
the use of many FPGA-boards together with associated containers realized an efficient
sharing of servers and the FPGA-devices. The use of Docker containers contributed to
realize separate lightweight FPGA experiment virtual machines. The Docker Swarm
enables to allocate them over the twenty laboratory servers to organize as the on premise
private cloud. The use of the Web services and the Docker Swarm allow realizing a
flexible and easy device allocation/gang scheduling and initiating the experiments
mechanisms. The paper showed the Cyber laboratory’s applicability for M2M and IoT
kinds of remote experiments.
The author has been engaged in the development of the former Cyber Laboratory in the
hybrid cloud [1–3], which consists of laboratory servers as a private cloud and a public
cloud. The use of the public cloud was employed to cope with the dynamically changing
student workload by scaling out or shrinking the number of servers in the public cloud.
zamfira@unitbv.ro
Concurrent Remote Group Experiments in the Cyber Laboratory 369
On the other hand, the use of the on premise private cloud should be inevitable, as on
premise laboratory servers should be equipped with FPGA-evaluation devices. The use
of FPGA boards and associated device dependent tools make it difficult to be migrated
to a public cloud. Thus, the hybrid cloud organization became an unavoidable solution.
Fortunately, most design automation tasks except for the FPGA experiment service can
be migrated to the public cloud. So, the laboratory servers can concentrate on FPGA
experiment services and the twenty laboratory servers were more than enough for
handling 80 student uses. Although the former Cyber laboratory can handle a number
of single FPGA use experiments both in parallel and time-shared fashions by making
use of available laboratory servers, it could not properly support group experiments by
making use of plural FPGAs and edge servers such as the cases for IoT experiments. If
students should start such remote group experiments simultaneously by making use of
plural FPGA devices and edge servers, the system should become overloaded. To over‐
come such shortcomings, up to eight FPGA devices have been added to each laboratory
server. As those workloads are rather lightweight, one server can handle quite a few
FPGA evaluation boards simultaneously. The proposed system could handle a group
remote experiment which contains hundreds of FPGA devices as edge devices and quite
a few edge servers connected by the internet. So, even for the IoT or M2M kinds of
experiments, an efficient sharing of laboratory experiment platforms can be realized by
the new Cyber Laboratory as shown in the Fig. 1.
The key technologies for implementing the new Cyber laboratory are summarized
as follows:
– The use of a public cloud for handling most Design Automation tasks to offload the
private cloud side workloads and realizing the scale out ability
– The use of the on premise private cloud for implementing the FPGA device dependent
services and connecting them with the public cloud counterpart to organize as a
hybrid cloud
– The use of the Docker containers for realizing isolated lightweight virtual environ‐
ments of FPGA experiments in the on premise cloud
– The use of the Docker Swarm for realizing a lightweight multi-host virtual machine
network to organize as the on premise private cloud
– The gang scheduler handles collection of FPGA containers to perform a group
experiment
– The rendezvous mechanism realizes the scheduling for group of group experiments
In this way, Each FPGA experiment service which is associated with each FPGA
evaluation board can be handled in a separate operating environment. As an average,
one laboratory server contains more than eight such virtual environments at the same
time. The use of Docker containers is promising as it can handle number of such virtual
environments with little overhead in the form of containers. It can provide users with
separate and isolated virtual environments. On the other hand from the operating system
point of view, each container seems to be a collection of processes and thus quite a few
containers can reside on a server and run more efficiently without virtual machine change
overhead, which is much heavier than process change overhead. Since each container
zamfira@unitbv.ro
370 N. Koike
works like a standalone machine, communications can be realized in the form of the
Web Services and the Web Sockets via the http ports.
In order to setup twenty laboratory servers to form as a private cloud, the Docker
Swarm [6] is useful. The laboratory server cluster can easily be organized as a Docker
Swarm multi-server network configuration. The Docker Swarm provides the system
zamfira@unitbv.ro
Concurrent Remote Group Experiments in the Cyber Laboratory 371
manager/scheduler with easy and centralized management means for controlling the
FPGA service containers in the private cloud. In order to guard the FPGA service
containers against outside service accesses, additional Swarm manager server is neces‐
sary to run a container for the Swarm management and service scheduling. A key value
store is necessary as well to manage the multi host network cluster, where all Docker
daemons in all laboratory servers have accesses to the key value store to join the network.
After the join operations, the Swarm manager becomes accessible to all joined laboratory
servers and can deploy containers on any laboratory server by making use of the remote
Docker APIs. Users can only send their FPGA service requests to this Swarm manager/
scheduler. Then, the scheduler find out idle FPGA devices and assigns corresponding
FPGA service containers by issuing a set of Docker swarm deploy commands to the
laboratory servers in the Docker swarm network.
The FPGA service containers can be allocated in either spread mode or gang sched‐
uling mode. After FPGA service containers are deployed in the laboratory servers, the
scheduler forwards the FPGA service requests to the designated FPGA service
containers by making use of the Web services via the http port. In case of a group
experiment request, namely a gang scheduling, it is translated into a set of FPGA service
requests to be deployed together and put into the wait queue specified as a gang sched‐
uling mode. When sufficient number of idle FPGAs becomes available, the scheduler
deploys participating containers by making use of the Docker remote APIs and issues a
set of FPGA run commands to them all by making use of the Web services.
If IoT edge servers are also specified, appropriate server containers are allocated as
well and the Docker image of them are pulled from the repository and deployed. It is
easy to add such Docker containers as the ones, each containing http daemon and user
defined server side applications in the laboratory servers. They mimic IoT edge server
functionalities.
If a request contains number of differently designed FPGAs, then it is handled as a
group of group experiments request. The request is first put into the rendezvous pool to
wait for remaining group experiment requests to arrive, because the experiment cannot
be started until all participating FPGA designs become ready. When all designated
number of group experiment requests become available, they are sent to the scheduler
and the group of group experiments should be carried out in the same way as the group
experiment.
In the Docker Swarm networking, the allocations of FPGA service containers are
automatically handled by the scheduler and the users usually do not care of the physical
container allocations. However, the users should care of their FPGA experiment results
and would like to observe the actual FPGA behaviors by making use of their recorded
videos. Web cameras are convenient to take shots of such experiments in videos, which
become useful for post experiment inspections. So, all Docker containers associated
with individual FPGA boards have been also installed motion application programs,
those control the attached Web cameras and take videos during the experiment and
transfer them to the faculty database.
The Docker volume plugins are employed to mount the faculty database network
directory in place of local directory for sharing files and achieving the file persistency. As
the faculty database stores all student designs, result files and recorded videos, a
zamfira@unitbv.ro
372 N. Koike
convenient sharing of the experiment information among students and teachers can be
realized. It also contributes to minimize the file transmissions among the on premise
private cloud, the public cloud and student PCs.
As all communications in the Cyber laboratory are realized in the form of the Web
services and the Web sockets, messages are exchanged in the form of XMLs in the
packets. They contain only directory information for the faculty database. So, a number
of unnecessary transmissions for the raw design data can be minimized.
The public cloud handles the remaining services other than the FPGA experiment
services. The organization of the public cloud can be done in a straightforward way.
Free FPGA design automation tools except FPGA setup/run tools have been installed
in every student virtual machine. As for the FPGA setup/run parts, they have to be
installed in the on premise cloud side servers. So, the newly designed programs instead
of the original tool generate the FPGA setup/run requests and communicate with the
Swarm manager/scheduler in the on premise private cloud by means of the Web services
and the Web sockets. Allocations of these student VMs to public side servers are deter‐
mined according to the amount of student workloads.
Aside from the student VMs, limited number of physical servers in the public cloud
is also allocated for servicing licensed Verilog-HDL high level logic synthesis software.
As the license limits the number of platforms, installing them in all student VMs is not
affordable. Students may access to these licensed Verilog-HDL Synthesis services via
the Web services in order to share these tools in both space and time shared fashions. A
VPN is setup among the on premise cloud, the public cloud, the faculty database and
student PC/Laptops to guard against unauthorized accesses. In order to use the cyber
laboratory, the student has to join the VPN and login the system. After the successful
authentication, one student VM is allocated to handle entire experimental services in the
public cloud. The student can have an access to the allocated VM by making use of
either the remote desktop service or http Web applications.
Figure 2 shows the general overview of the proposed remote group experiment envi‐
ronments as well as the simple on-demand FPGA experiments. Users can perform such
group experiments or standalone single FPGA-use experiments concurrently by making
use of idle FPGA devices. Up to a hundred and sixty FPGA experiment container/FPGA
device pairs over twenty laboratory platforms can be organized for FPGA experiment
services. In this way, plural IoT remote experiments can be carried out by making use
of groups of FPGAs and server containers simultaneously. The cyber laboratory
manager in the Swarm manager/scheduler container performs the system management
tasks, such as experiment request receptions, obtaining file location directories in the
faculty database, container deployments and initiating FPGA experiments. Each student
VM in the public cloud setups the connection with the Swarm manager/scheduler
container in the On-Premise private cloud to request FPGA experiment services in on-
demand basis. The dash board service lets students know of current workload status and
select appropriate containers.
zamfira@unitbv.ro
Concurrent Remote Group Experiments in the Cyber Laboratory 373
For a group experiment, a set of FPGA experiment service requests and IoT server
service requests have been wrapped into a JSON file and sent to the scheduler. It is un-
wrapped and put into the experiment request queue specified as the gang scheduling
mode. In case when a group of group experiments request has been received, it is un-
wrapped and sent to the rendezvous pool. It should be kept waiting until all group
experiment requests have arrived. Then these group experiment requests are put into the
experiment request queue together in the same way as the group experiment request
case.
4 Conclusion
The new Cyber laboratory has been presented. It showed the feasibility of conducting
remote group experiments concurrently, applicable for IoT or M2M related experiments.
The use of Docker containers and Docker Swarm technologies contributed to realize a
network of hundreds of independent FPGA experiment containers together with quite a
few server containers. It can contain various kinds of subnetworks, where group
zamfira@unitbv.ro
374 N. Koike
experiments utilizing FPGA devices and edge servers can be conducted concurrently.
The newly designed Swarm manager and scheduler have been developed to realize the
gang scheduling and rendezvous mechanisms for handling incoming experiment
requests. A group of group experiments, which contains different designs of FPGA
devices and IoT edge servers, can be handled by making use of the rendezvous pool.
Although the VPN and authentication process assure the minimum security
level, much higher level of security protection mechanisms should be introduced for
practical use.
Acknowledgments. The author would like to thank Mr. Yuichi Toyoda for his zealous efforts
in realizing the test bed for the semi-automatic experiments in the hybrid cloud.
References
1. Toyoda, Y., Koike, N., Li, Y.: An FPGA-based remote laboratory: implementing semi-
automatic experiments in the hybrid cloud. In: 14th International Conference on Remote
Engineering and Virtual Instrumentation (REV2016)
2. Koike, N.: Cyber laboratory: migration to the hybrid cloud solution for device dependent
hardware experiments. In: International Conference on Information Technology Based Higher
Education and Training (ITHET 2014), Kent, U.K., June 2014
3. Koike, N.: A cyber laboratory for device dependent hardware experiments in a hybrid cloud.
In: 6th International Conference on Computer Supported Education (CSEDU 2014),
Barcelona, Spain, April 2014
4. Morgan, F., Cawley, S., Kane, M., Coffey, A., Callaly, F.: Remote FPGA lab applications,
interactive timing diagrams and assessment. In: Irish Signals & Systems Conference 2014 and
2014 China-Ireland International Conference on Information and Communications
5. Jethra, J.S.T., Patkar, S.B., Datta, S.: Remote triggered FPGA based automated system. In:
11th International Conference on Remote Engineering and Virtual Instrumentation
(REV2014), pp. 309–314, 26–28 February 2014
6. https://docs.docker.com/swarm/networking/
zamfira@unitbv.ro
The VISIR+ Project – Preliminary Results
of the Training Actions
1 Introduction
As previously stated, VISIR system is a widespread remote lab used mainly in the
study of electrical and electronic circuits, with increased popularity in the last 5 years
[13], mainly due the intrinsic advantages of being a remote lab: accessibility, avail-
ability and safety - since the users are not exposed to any electrical signal, and in turn
they are not able to damage the physical equipment, due to a series of protection layers
that prevent this. The only physical interaction between the user and the real equipment
zamfira@unitbv.ro
The VISIR+ Project – Preliminary Results of the Training 377
zamfira@unitbv.ro
378 M.C. Viegas et al.
(HEI), mainly in Europe, for the past years, the technical and didactical experience
gained by the five Europeans Institutions partners (EU partners) in the project, is now
being shared with five Latin American Institutions (LA partners). Their characteristics
and main role within the Project are described in Table 1.
zamfira@unitbv.ro
The VISIR+ Project – Preliminary Results of the Training 379
Apart from spreading VISIR usage, the project includes the VISIR system purchase
by each HEI in LA, fostering their sense of ownership and contributing to enlarge the
VISIR community. BTH is the EU partner in charge of technical support during
installation and organizing a training technical workshop for each technical team.
Overall, the project purpose is to enlarge the VISIR usage community by progressively
enlarging its coverage: firstly, through a one-to-one relation between EU and LA
partners, where the EU partner acts as the tutor; and secondly each LA HEI partner
with their AP, working closely with LA HEI, also implementing VISIR in their
courses. These AP involved serve different education levels (higher, secondary and
professional).
In order to guarantee the implementations success in all LA HEI partners and AP,
three TA’s were defined in different project stages. The first two have been performed
by the EU partners and the third one will be carried out by each LA HEI partner in their
AP. The objective was to replicate and enlarge the community of usage, share expe-
riences, render its advantages and contextualize their implementations. In order to
better understand the outcomes obtained from the different approaches and the insights,
an external observer was present in all TA’s. TA1 took place in Europe and its goal was
to introduce VISIR and its capabilities, where each EU partner shared their experience.
The TA2 meant to specifically address teachers’ needs (in each institution), particularly
to those implementing VISIR in their classes. TA2 took place in each target LA HEI.
The TA3 was designed to be delivered by LA HEI teachers who used VISIR and to
take place in their AP, with the objective of sharing their own contextualized experi-
ences and involving more teachers. The TA’s are sequential and intended to support the
different implementation phases, where the 1st implementation phase is meant to be
unique - one course per LA HEI, and the 2nd phase is meant to spread into several
implementations. The 3rd phase is meant to occur both on LA HEI partners as in AP. In
sum, the major outcomes of the VISIR+ project will necessarily be: trained local
technicians, trained local teachers, educational modules development and enlargement
of the VISIR facilitators group.
At the present stage of VISIR+ Project development, not all planned actions for each
LA HEI partner took place, namely the VISIR acquisition. In most cases economic and
administrative constrains delayed the acquisition procedure. Still, and due to the remote
lab characteristics, the project actions “TA2” and “1st implementation” could be per-
formed successfully by using the EU partner’s VISIR system. In order to better
understand TA impact in those circumstances and preview steps to corrective/redirect
the development process, the action results were assessed.
3.1 Focus
This paper presents the preliminary results of two Training Actions (TA1 and TA2). As
in most didactical implementation, teachers’ perception of different tools, their recep-
tivity and motivation to change their classes, strongly conditions the outcomes. So, the
zamfira@unitbv.ro
380 M.C. Viegas et al.
global impact of these TA is probably a good indicator of teachers’ interest and the
success of subsequent implementations. The goal of this study is to assess each LA HEI
implementation and analyze their differences in order to adjust the following phase of
implementations. The research questions are: Which factors can be considered
important in terms of conditioning the TA and the didactic implementations using
VISIR? Is there any relation between TA characteristics and the implementations
designs?
3.2 Approach
The research methodology used is a Multi-Case Study [22], in which five cases (LA
HEI Partners) will be presented and assessed. Due to the diversity of contexts, back-
grounds and experience, there were natural differences between TA’s, even though a
common base had been established. In order to characterize these differences three
categories were defined according to their timeline: pre-TA; during TA and post-TA
(Table 2).
zamfira@unitbv.ro
The VISIR+ Project – Preliminary Results of the Training 381
Regarding the followed activities, data also included teachers’ schedule imple-
mentations of VISIR in their classes, the number of teachers and students involved in
each case and the kind of VISIR’s usage interaction that would be asked from them.
zamfira@unitbv.ro
382 M.C. Viegas et al.
Regarding VISIR’s acquisition, only PUC-Rio was able to perform the planned
sequence of actions: TA1 ! VISIR acquisition ! Technical Workshop ! TA2 ! 1st
Implementation. In all other cases, the administrative constrains within each Institution,
Governmental and European directives, forced them to resort to an alternative plan. This
plan was made possible due to the resourcefulness of remote labs: each European tutors
made available their own VISIR system in order to allow to plan TA2 and didactical
implementations. In this case, the sequence was altered to: TA1 ! TA2 ! 1st Imple-
mentation ! VISIR acquisition ! Technical Workshop.
Concerning the Project and TA2 dissemination, there were cases where the partners
assumed a general dissemination to all potential teachers and, especially in TA2, profit
from the EU partners visit to enlarge the bounds with LA institutions and associated
partners. This was the case for instance in UFSC and UNSE. Others, like PUC-Rio or
IFSC interpreted that TA2 was meant for the teachers already motivated to use VISIR
and did not centered their efforts on encouraging more teachers to attend.
Finally, in terms of past experience with ICT tools, remote labs or VISIR, some
differences are worth mentioning. UFSC uses remote labs since 1997 and were
responsible for the development of the RexLab project [23]. IFSC and UNR have
already used VISIR together with their tutors (IPP-ISEP and UDEUSTO, respectively)
in the past [24]. Another example of previous experience is PUC-Rio, that have been
using ICT tools in education for 21 years with their Maxwell platform [25]. And even
though they did not have past experience with VISIR, PUC-Rio was the only HEI who
actually performed a pre-implementation (within the VISIR+ Project scope), using
their tutors’ VISIR system, before the project-planned implementations.
zamfira@unitbv.ro
The VISIR+ Project – Preliminary Results of the Training 383
Fig. 1. Distribution of TA2 participants (HEI Fig. 2. Distribution of TA2 time duration (in
partner and AP) between HEI holder. days) in each HEI.
zamfira@unitbv.ro
384 M.C. Viegas et al.
comments were made from the start (more meaningfully in UNR and UNSE where
the presentations became more interactive). This distribution is showed in Fig. 3
relatively to the five cases. The questions and queries from the audience facilitated
the observation of attendees’ attention and interest. The practical activities, such as
accessing lab, designing circuits, measuring and analyzing results, got attendees
involved in the lab use straightaway, and their questions and queries were readily
answered by the lecturers.
• Regarding TA’s attendees’ perception (quantitative assessment)
The global feedback in both TA was highly positive which evidences the satis-
faction of LA HEI. Even though EU partners were present in TA and some also
answered the satisfaction questionnaire, the results shown in Fig. 4 only refers to
LA’ answers. In general, the global average level of satisfaction even grows in
IFSC, UFSC an PUC Rio. About presentation interaction with the audience (Q2),
the level of satisfaction maintains or increases in TA2. On the other hand, partic-
ipants’ expectations (Q7) and difficulties in practical use (Q8), maintains or
decreases in almost every case. The answers in this last question had a more
notorious decrease in UFSC, UNSE and UNR. PUC-Rio was the only one who had
a slight increase in this question.
• Regarding TA’s attendee’s perception (qualitative assessment)
The purpose of the open question of the TA Satisfaction Questionnaire aimed at
eliciting qualitative information about positive and negative aspects of the TA. Four
main categories about aspects of the TA1 became salient after analyzing partici-
pants’ answers: content of presentations, VISIR practice, time management and
sharing experiences. As regards content of the presentations, most answers referred
to their relevance and clarity while some pointed out the fact that the content of each
presentation was discrete and failed to reach common objectives (“[…] the training
session is not the addition of few sessions, this must be a common session, with a set
of objectives. Each of these objectives will be reached by each presenter, and so
on”; “Maybe first pedagogy and after technology”). Most participants agreed that
more practice with VISIR Lab equipment could have been introduced: “We had no
practice hands-on”; “hands-on activity is mandatory to understand better the
possibilities”; “the time allotted for practice/hands-on was null”; “I would have
liked to have real practice on the setting up of components in the lab, not just using
it” (our translation); “Maybe a training session with PC’s doing
circuits/experimenting in VISIR could be really interesting”. Timing was the aspect
of the presentation which most participants referred to, although it was considered
from multiple perspectives: time assigned for each presentation slot (“Speakers did
not fit to their time slots, this disturbed the following speakers”; “Not enough time
for all presentations and questions”; “Time allocation was uneven, so some
speakers ended up with little time to explain their results” and “The time for the
conference was not enough for all”); time lost (“The time to set up the presentations
could be avoided by using the same computer for the entire session”) and time for
more actual practice with VISIR. Finally, most participants found the presentation
of EU HEI experiences an asset in the training action (“[positive] Present
experiments and experiences at different institutions using VISIR”), although some
zamfira@unitbv.ro
The VISIR+ Project – Preliminary Results of the Training 385
argued more opportunities for open interaction could have been present (“Every-
thing was clearly explained, however we should have kind of round table to discuss
more about the experience the colleagues had had”). TA1 also had virtual
streaming. Virtual attendees found the videoconference positive (“interesting”,
“excellent”) although when answering Question 9, they referred mostly to technical
problems: sound problems; questions asked by participants were not heard; only
slides were shown during the presentation.
Fig. 3. Distribution of TA2 presentation Fig. 4. Distribution of TA1 and TA2 satisfac-
approach in each HEI. tion questionnaires results for each case.
zamfira@unitbv.ro
386 M.C. Viegas et al.
making the online use slow”). Finally, some recommendations for extension of the
experience were given: “I hope VISIR could be taken to Angola, my country” and “The
lab has to be promoted to many departments of electric engineering careers”.
zamfira@unitbv.ro
The VISIR+ Project – Preliminary Results of the Training 387
integration of VISIR in the course, using their material which was already designed to
accommodate several different resources, in which students can complete their tasks (in
a similar way as the VISIR project stimulates teachers to use simultaneously hands-on,
simulators, remote labs and calculus). PUC-Rio also implemented a Complementary
Activity using VISIR, open to all engineering students from various backgrounds: an
online course, covering up basic electricity concepts. So, even though not in a large
number, the quality of these implementations cannot be underestimated. Several AP are
also already using VISIR, but mainly to test it and implement it next academic year.
5 Analysis
zamfira@unitbv.ro
388 M.C. Viegas et al.
The challenge endured in this work, was to assess in which terms external and internal
factors to the VISIR+ project was affecting the ongoing actions. In particular, at what
extent the pre-project experience of the LA HEI partners and particular aspects that
made TA different in each case were significant in terms of affecting teachers
involvement and the developing VISIR’s harmonious integration in their course
curricula.
Since the main turning point in each case was the TA, the analysis was divided into
three chronological stages in which cases could be differentiated: pre TA; during TA
and post TA. Pre TA characterization showed some differences between cases, namely:
PUC-Rio was the only one who managed to acquire VISIR system on schedule; UFSC,
UNSE and UNR performed a larger dissemination of the Project and TA among their
HEI fellows and including their AP; PUC-Rio shows a high level of performance while
ICT tools users; UFSC shows a vast experience using remote labs; UFSC, IFSC and
UNR shows some experience with VISIR.
TA characterization was made regarding attendees’ participation, the language used
by EU partners, as its duration and presentation approach. TA where assessed through a
satisfaction questionnaire in which quantitative and qualitative data was collected. The
major results are now summarized: UFSC, UNSE and UNR had a large number of
participants, including their AP; A significant difference in terms of language used by
EU partners was possible from TA1 to TA2. Due to their affinities, EU partners
performed their presentations (or in case of PUC-Rio, their discussions) in participants’
native language; The time allocated in each case to TA2 varied from one day (in UFSC
and IFSC) and four days in UNSE; In UFSC and IFSC the presentation approach was
less interactive (questions were mostly postponed to the end); Global feedback in both
TA was highly positive; The global participants’ perception of the presentations
interaction with the audience level (Q2) maintains or even grows in some cases from
TA1 to TA2; Regarding the level of achievement of participants’ expectations (Q7) and
the sensed difficulty in using VISIR (Q8), the results shows a maintenance (of the lower
zamfira@unitbv.ro
The VISIR+ Project – Preliminary Results of the Training 389
zamfira@unitbv.ro
390 M.C. Viegas et al.
Acknowledgment. The authors would like to acknowledge the support given by the European
Commission through grant 561735-EPP-1-2015-1-PT-EPPKA2-CBHE-JP.
References
1. Jara, C., Candelas, F., Puentes, S., Torres, F.: Hands-on experiences of undergraduate
students in Automatics and Robotics. Comput. Educ. 57, 2451–2461 (2011)
2. Feisel, L., Rosa, A.: The role of the laboratory in undergraduate engineering education.
J. Eng. Educ. 94, 121–130 (2005)
3. Brinson, J.: Learning outcome achievement in non-traditional (virtual and remote) versus
traditional (hands-on) laboratories: a review of the empirical reserach. Comput. Educ. 87,
218–237 (2015)
4. Corter, J., Nickerson, J., Esche, S., Chassapis, C., Im, S., Ma, J.: Constructing reality: a study
of remote, hand-on and simulated laboratories. ACM Trans. Comput. Hum. Interact. 14(2),
1–27 (2007)
5. Bochicchio, M., Longo, A.: Hands-on remote labs: collaborative web laboratories as a case
study for IT engineering classes. IEEE Trans. Learn. Technol. 2(4), 320–330 (2009)
6. Corter, J., Esche, S., Chassapis, C., Ma, J., Nickeson, J.: Process and learning outcomes from
remotely-operated, simulated and hands-on student laboratories. Comput. Educ. 57, 2054–
2067 (2011)
7. Ma, J., Nickerson, J.: Hands-on, simulated and remote laboratories: a comparative literature
review. ACM Comput. Surv. 38(3), 1–24 (2006)
8. Marques, A., Viegas, C., Costa-Lobo, C., Fidalgo, A., Alves, G., Rocha, J., Gustavsson, I.:
How remote labs impact on course outcomes: various practises using VISIR. IEEE Trans.
Educ. 57(3), 151–159 (2014)
9. Gustavsson, I.: On remote electronics experiments. In: Zubía, J.G., Alves, G.R. (eds.) Using
Remote labs in Education: Two Little Ducks in Remote Experimentation, pp. 157–176.
Universiy of Deusto, Bilbao (2011)
10. Gomes, L., Bogosyan, S.: Current Trends in Remote Laboratories. IEEE Trans. Industr.
Electron. 56(12), 4744–4756 (2009)
11. [IAOE] Winners of the GOLC Online Laboratory Award, 11 February 2015. http://lists.
online-lists.org/pipermail/iaoe-members/2015-February/000120.html. Accessed 2016
12. Alves, G., Fidalgo, A., Marques, A., Viegas, C., Felgueiras, M., Costa, R., Lima, N.,
Garcia-Zubia, J., Hernández-Jayo, U., Castro, M., Díaz-Orueta, G., Pester, A., Zutin, D.,
Kulesza, W.: Spreading remote labs usage: A System – A Community – A Federation. In:
Proceedings of the 2nd International Conference of the Portuguese Society for Engineering
Education (CISPEE2016), Vila Real, Portugal (2016)
13. Lima, N., Viegas, C., Alves, G., Garcia-Peñalvo, F.: VISIR’s usage as a learning resource: a
review of the empirical research. In: Proceedings TEEM 2016 - Fourth International
Conference on Technological Ecosystems for Enhancing Multiculturality (TEEM 2016),
Salamanca, Spain (2016)
14. Gustavsson, I., et al.: The VISIR project - an open source software initiative for distibuted
online laboratories. In: Remote Engineering & Virtual Instrumentation (REV 2007), June
2007
15. Gustavsson, I., Zackrisson, J., Nilsson, K., Garcia-Zubia, J., Hakansson, L., Claesson, I.,
Lago, T.: A flexible electronics laboratory with local and remote workbenches in a grid. Int.
J. Online Eng. (iJOE) 4(2), 12–16 (2008)
zamfira@unitbv.ro
The VISIR+ Project – Preliminary Results of the Training 391
16. Alves, G., Viegas, C., Lima, N., Gustavsson, I.: Simultaneous usage of methods for the
development of experimental competences. Int. J. Hum. Cap. Inf. Technol. Prof. 7(1), 48–63
(2016)
17. Claesson, L., Hakansson, L.: Using an online remote laboratory for electrical experiments in
upper secondary education. Int. J. Online Eng. (iJOE) 8(S2), 24–30 (2012)
18. Fidalgo, A., Alves, G., Marques, A., Viegas, C., Costa-Lobo, C., Hernadez-Jayo, U.,
Garcia-Zubia, J., Gustavsson, I.: Adapting remote labs to learning scenarios: case studies
using VISIR and RemotElectLab. IEEE Revista Iberoamericana de Tecnologias del
Aprendizage 9(1), 33–39 (2014)
19. Garcia-Zubia, J.: Using VISIR experiments, subjects and students. Int. J. Online Eng. (iJOE)
7(2), 11–14 (2011). REV2011
20. Lima, N., Alves, G., Viegas, C., Gustavsson, I.: Combined efforts to develop students
experimental competences. In: Proceedings Exp. at 2015 3rd International Experimental
Conference, Ponta Delgada, Azores (2015)
21. Viegas, C., Lima, N., Alves, G., Gustavsson, I.: Improving students experimental
competences using simultaneous methods in class and assessments. In: TEEEM 2014
Proceedings of the second International Conference on Technological Ecosystems for
Enhancing Multiculturality, Salamanca, Spain (2014)
22. Cohen, L., Manion, L., Morrison, K.: Research Methods in Education, 6th edn. Routledge,
Taylor & Francis Group (2007)
23. UFSC. http://rexlab.ufsc.br. Acedido em 11 2016
24. Lerro, F., Orduña, P., Marchisio, S., García-Zubía, J.: Development of a remote laboratory
management system and integration with social networks. Int. J. Recent Contrib. Eng. Sci.
IT (iJES) 2(3), 33–37 (2014)
25. PUC-RIO. http://www.maxwell.vrac.puc-rio.br. Accessed Nov 2016
zamfira@unitbv.ro
Laboratory Model of Coupled Electrical Drives
for Supervision and Control via Internet
1 Introduction
Control of electrical drives offers insight into electric drives and their usage in motion
control environment [1, 2]. It provides links among electrical machine and control theory,
practical hardware aspects, programming issues, and application-specific problems [1].
Most of the machine centers, industrial robots, servomechanisms and other rotating
machinery have the geared reduction mechanisms between output shafts of motors and
driven machine parts. The insufficiency of the torsional stiffness of the geared reduction
mechanism often induces transient vibrations mainly related to eigenvalues of the
mechanical parts in the lower-frequency range when the motor starts or stops [3]. Elastic
couplings and joints within the machine system are major impediments to the
© Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6_37
zamfira@unitbv.ro
Laboratory Model of Coupled Electrical Drives for Supervision and Control via Internet 393
performance enhancement, since high loop gains often destabilize torsional resonance
modes associated with the transmission flexibility. The presence of torsional resonance in
motion control systems limits the maximum achievable performance and causes unde-
sirable oscillations in the control system response [3, 4]. Vibration suppression of
rotating machinery is an important engineering problem [3, 4].
This paper presents a laboratory model which should help users to understand the
key elements of motion control systems, introduce them in hands on practice with
industrial servo drives, analyze and design discrete-time speed and position controllers,
set adjustable feedback parameters, and evaluate closed-loop performances including
of suppression of torsional resonance phenomena [4, 5]. The laboratory model covers
wide span applications for problem based learning and research [5].
Laboratories are inherent part of engineering education. Good designed student
experimental work bridges the gap between theoretical analysis and industrial practice
[5–11]. But, many universities worldwide, especially in poor countries, cannot afford
adequate laboratories for engineering education because of high cost of laboratory
equipment. One solution is blended learning approach and the development of WEB
laboratories where laboratory resources can be shared among a lot of users from any
place in any time [5, 6, 11].
This paper presents WEB pages of designed laboratory model integrated within
Internet mediated laboratory with purpose of support for teaching/learning in electrical
and servo drives use in mechatronics applications.
zamfira@unitbv.ro
394 M. Matijević et al.
Actual
Velocity
Resonance
Reference
Time
In order to solve these problems, system designers will sometimes attach a damping
load, such as an inertial damper, to the back of the motor. However, such a load has the
undesired effects of decreasing overall performance, and increasing system cost [3]. On
the other side, to overcome the problem, various control strategies have been proposed,
that may be divided into the following three groups [4]: (1) control strategies based on
the direct measurement of motor- and load- side variables, (2) strategies involving only
one feedback device attached to the motor and the observer that estimating remaining
states, and (3) vibration suppression strategies based upon the notch filtering and
phase-lead compensation applied in conventional control structures.
Designers of the control part of a servo system, usually use the simplest motor/load
models that haven’t information about resonance modes and fast dynamics. The more
realistic model of an AC motor with load is illustrated on Figs. 2 and 3, as a two-mass
motor/load system with flexible coupling [4].
ΔΘ Δω
Mem cs
Motor Load
Θl
Jm Jl
Θ m ωm bv ωl Ml
Fig. 3. Block diagram of the servo system’s plant with flexible coupling
zamfira@unitbv.ro
Laboratory Model of Coupled Electrical Drives for Supervision and Control via Internet 395
The electromagnetic torque Mem is control variable, and the torque on loaded shaft
Ml presents disturbance. The motor inertia Jm and load inertia Jl are coupled by the
shaft or the transimission system having a finite stiffness coeficient cs. The friction
coefficient bv generally assumes very low values, giving rise to weakly damped
mechanical oscillations [4]. The torsional torque Mo equals the load torque Ml only in
the steady state. During transients, the speeds of motor and load differ, and torsional
torque Mo is given by
Mo ¼ cs Dh þ bv Dx ð1Þ
Contrary to the traditional model Wm(s) = 1/(Jl + Jm)s, if the shaft sensor is
mounted on the motor, the transfer function of the mechanical subsystem is defined by
2fz
xm ðsÞ 1 1þ xz s þ x2z s
1 2
Wm ðsÞ ¼ ¼ ð2Þ
Mem ðsÞ ðJm þ Jl Þs 1 þ 2fp
xp s þ x2p s
1 2
where undamped natural frequencies (xp, xz) and relative damping coefficients (fp, fz)
are given by
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi rffiffiffiffi
cs ð Jm þ Jl Þ cs
xp ¼ ; xz ¼ ;
Jm J l Jl
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi sffiffiffiffiffiffiffiffiffiffi ð3Þ
b2v ðJm þ Jl Þ b2v
fp ¼ ; fz ¼
4cs Jm Jl 4cs Jl
Undamped natural frequency xp and xz of the pole- and zero-pairs in (2) are
referred as the resonance and antiresonance frequencies [4], and their quotient is known
as the resonance ratio
rffiffiffiffiffiffiffiffiffiffiffiffiffiffi
xp Jl
Rr ¼ ¼ 1þ ð4Þ
xz Jm
In the case under consideration, a low value of resonance ratio reduces the influence
of torsional load on dynamics of the speed control loop. With Jm Jl, oscillations of
torsional torque are filtered by a large motor inertia Jm and their influence on the control
of the motor speed becomes smaller. A damped control of Hm and xm is favorable, but
most applications require fast and precise control of the load variables Hl and xl [3].
Also, in that case, the estimation of resonance modes from detected signals (Hm and
xm) is not possible, and the load speed and position might exhibit weakly damped
oscillations that cannot be disclosed and compensated from the feedback signals [3].
In the case that sensor is mounted on the load shaft, the mechanical subsystem of
the drive has the transfer function Wl(s) given by
zamfira@unitbv.ro
396 M. Matijević et al.
2fz
xl ðsÞ 1 1 þ xz s
Wl ðsÞ ¼ ¼ ð5Þ
Mem ðsÞ ðJm þ Jl Þs 1 þ 2fp s þ 12 s2
xp xp
where undamped natural frequencies (xp, xz) and relative damping coefficients (fp, fz)
are given by (3), too.
Many controllers already exist in the field of motion control, but all most of them are
designed by assuming an ideal, rigid transmission train [1–4]. However, the desired
speed-loop bandwidth in modern machining centers approaches the frequency of tor-
sional resonance and coincides, at the same time, with most disturbing statistical and
deterministic noises [3].
Under these conditions, PI control laws are not suitable. Standard improvement of
conventional motion control laws and structures is based on anti-resonant compensator
inclusion as it is shown on Fig. 4.
Ml
Conrolling structure
ω ref ω m ,θ m
Anti-resonant Me m
Control law compensator OU
ωl, θl
as antiresonance comensator (Fig. 4) is the most frequently used in practice. The notch
filter zeros cancels critical poles (of the torsional load), while the poles of the filter
become
a new pair of conjugate complex poles with increased relative damping
fpp fzz . Digital implementation of notch filter fpp ¼ fp ; fzz ¼ fz ; xnf ¼ xp is
given by discrete transfer function
qffiffiffiffiffiffiffiffiffiffiffiffiffi
xT xT
ðfp fz Þ fp xp T
cos xp T 1 f2z z1 þ eðfp þ fz Þ z2
p p
e 2e
Wnotch z1 ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffi ð7Þ
T
1 2efp xp T cos xp T 1 f2p z1 þ e2fp xp z2
zamfira@unitbv.ro
Laboratory Model of Coupled Electrical Drives for Supervision and Control via Internet 397
For an exact cancellation of resonance modes, both the resonance frequency and
damping factor must be known while tuning all parameters of the notch filter [4]. But,
the exact location of critical poles is unknown and, thus, the cancellation is generally
imprecise. The notch filter (6) suppresses the resonant mode by the ratio fpp =fzz . Since
a low damping coefficient of zeros increases greatly the snesitivity to parameter vari-
ations, the ratio fpp =fzz is limited. Hence, the excitation of resonance modes can be only
reduced, but not eliminated completely, by the notch serial compensator. The notch
compensator is very sensitive to parameter variation, and it presents a serious problem
in tuning and implementation the notch filter [4].
Ωm Me m
PI K Torsional ωl
ωm
Ml load
1-K 1 Jm s
τs+1 τs+1
Ml
ωr -1 1 Me m ωl
P r( z ) -1 Torsional load
+ - + - R( z )
o
z-1Puo (z-1) - Q (z-1)
+
+ cp
+
z-1 1-cp
Py( z-1)
In the literature, as antiresonant control strategies are proposed model based control
approaches, control techniques based on the disturbance observers (Fig. 5), approaches
based on IMPACT structure (Fig. 6) [3], two freedom structure based on H2 control,
approaches based on antiresonant compensators (Fig. 4) [4], RST controller [12],
robust control and vibration suppression control in two-mass drive systems [13–16].
A good review of control strategies of vibration supression is given in [4]. In [4] is also
proposed a new antiresonance compensator, practically, FIR filter
1 þ zn Tosc
WNF ðz1 Þ ¼ ;n ¼ ð8Þ
2 2T
where the n stands for the ratio between the resonance mode half-period (Tosc/2) and
zamfira@unitbv.ro
398 M. Matijević et al.
the sampling time (T) of the discrete time controller. The oscillation period of the
resonance mode Tosc is given by
2p
Tosc ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffi ð9Þ
xp 1 fp
and it is adjustable parameter of the FIR filter (8), that could be experimentally defined.
The idea of the synthesis of filter (8) was elaborated in [4]. The conceived cascade
anti-resonance compensator is simpler, less sensitive to parameter changes, and
requiring a setting of only one parameter, but parameter n have to be identified pre-
cisely. The theoretical value of the suppression at xosc ¼ 1=Tosc frequency is infinite,
rather then a finite f/fp notch filter suppression value.
The resonance ratio control is proposed as an improvement of model-based control
techniques (i.e. model following control, application of disturbance observer, time
derivate feedback, state feedback control). The resonance ratio is defined by relation (4),
pffiffiffi
and should be about 5 because of effective vibration suppression [3]. In [3], as a
simple and practical strategy, it is mentioned the resonance ratio control based on the
pffiffiffi
fast disturbance observer (see Fig. 5), with optimal resonance ratio 0.8 5. In conven-
tional disturbance observer applications 100% of the estimated disturbances is feed
back. In the case on Fig. 5, 1 − K of the estimated disturbances is used. Parameter
K (0 < K 1) and time constant s (which defines observer’s cutoff frequency) are
adjustable parameter for vibration suppression. But, as it is previous commented, this
control strategy cannot efficiently provide vibration suppression on the load side.
Synthesis of IMPACT structure starts from following plant model (see Fig. 2 –
flexible coupling is neglected)
1 1 T z1
Wou ðsÞ ¼ ¼ ; Wou ðz1 Þ ¼ ð10Þ
Js ðJm þ Jl Þs Jm þ Jl 1 z1
Tosc p
T¼ ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffi ð11Þ
8 4xp 1 f2p
Namely, in the IMPACT structure, the control plant is given by its nominal discrete
model
which is included into the control part of the IMPACT structure as a two-input internal
plant model. According to the standard procedure of IMPACT structure synthesis, for a
zamfira@unitbv.ro
Laboratory Model of Coupled Electrical Drives for Supervision and Control via Internet 399
where
qffiffiffiffiffiffiffiffiffiffiffiffiffi
s1=2 T
z1=2 ¼ e ; s1=2 ¼ fxn jxn 1 f2 ð13Þ
The role of the RLSN (Recursive Linear Smoothed Newton) predictive filter with
user friendly parameter cp < 1 in inner loop of the IMPACT structure at Fig. 6 is
explained in [3]. Adjustment of parameters cp simply influences on efficiency of
absorption of disturbance effects or expansion of area of robust stability and sup-
pression of torsional oscillations. The presented structure is simple with small number
of adjustable parameters that could be easily set to achieve the desired robust, filtering,
and dynamic properties of the system.
zamfira@unitbv.ro
400 M. Matijević et al.
4 4
ωm [rad/s] ωl [rad/s]
3 3
2 2
1 1
0 0
-1 -1
-2 -2
0 0.05 0.1 0.15 0.2 0 0.05 0.1 0.15 0.2
t[s] t[s]
2 1.5
Me [Nm] Mo [Nm]
1 1
0 0.5
-1 0
-2 -0.5
0 0.05 0.1 0.15 0.2 0 0.05 0.1 0.15 0.2
t[s] t[s]
Fig. 7. Operation of IMPACT structure for cp = 0.2, xn = 400 rad/s, f = 0.7, xr(t) = 3h(t−0.05)
[rad/s], Ml(t) = 1h(t−0.1) [Nm]
Mechanical resonance is a current problem in servo systems, and falls into two cate-
gories: low-frequency and high-frequency. High-frequency resonance usually causes
instability at the natural frequency of the mechanical system, typically between 500 Hz
and 1200 Hz. Low-frequency resonance occurs more often in general industrial
machines, at the first phase crossover, typically between 200 Hz and 400 Hz. Standard
servo control laws are structured for rigidly coupled loads. However, when a such
control law is applied to a flexible coupled servo system, the system performance can
be unsatisfactory. Typical experimental set-ups for testing of control laws for electrical
drives with flexible coupling are consisting of identical motors interconnected by
elastic shaft. For example, in number papers of T. Orlowska-Kowalska et al. in this
area, like in [14], the laboratory model on Fig. 8. is used.
The laboratory model, presented at Fig. 8, is composed of a DC motor driven by a
static converter/four-quadrant chopper/and a DC loading machine. The motor is cou-
pled to a load machine by an elastic shaft (a steel shaft of 5 mm diameter and 600 mm
length). The moment of inertia can be varied by flywheel, where the inertia ratio of the
motor to the load machine varies from 0.125 to 8. Both motors had the nominal power
of 500 W each. The speed and position of both motors were measured by incremental
encoders (5000 pulses per rotation). The mechanical system has a natural frequency
approximately 9.5 Hz. The control and estimation algorithms were implemented using
a digital signal processor with dSPACE software.
In [4], the laboratory model for experimental verification of anti-resonant control
laws in servo systems with flexible coupling, is also arranged according to the block
diagram shown in Fig. 3. The laboratory model consists of two synchronous permanent
magnet servomotors connected by a flexible hollow shaft. Torsional oscillations upon
rapid load steps are found at 156 Hz, decaying to zero with the time constant of
approximately 200 ms [4]. Both motors are fed and controlled by a digital servo
amplifier capable of both the torque- and the speed-control operating modes. The motor
M1 (Fig. 3) is speed controlled, while the motor M2 is set in the torque-control mode
zamfira@unitbv.ro
Laboratory Model of Coupled Electrical Drives for Supervision and Control via Internet 401
Fig. 8. The mechanical part of the laboratory model (a); The general view of the lab. model (b);
Schematic diagram of experimental setup in [14] (c);
and used as a programmable load. The motors are equipped with electromagnetic
resolvers. The resolver signals are decoded into the motor- and the load-side positions
and in (Fig. 1) within the servoamplifier’s control section. The R/D converter’s
bandwidth is 1 kHz and its resolution is 12 b. Having two shaft sensors at each end of
the flexible coupling, it was possible to perform experiments with motor-side feedback
and load-side feedback [4].
Based on described laboratory model in [4], at Faculty of Engineering at University
of Kragujevac, based on industrial automation components is developed the laboratory
model of coupled electrical drives shown on the Fig. 9.
The laboratory model (Fig. 9, [5]) consists of
• Desktop computer equipped with internet/ethernet connection; College Teaching
License for LabVIEW (NI Academic Site License), Matlab/Simulink and NI PCI
6229 interface for control, measurement and data acquisition.
• Two servo drives (i.e. servo amplifiers) Yaskawa Omron SGDH-04AE-0Y [17, 18].
• Two AC synchronous permanent magnet servomotors Yaskawa Omron SGMAH
04A1A61D-OY [17, 18].
• Shaft or flexible shaft and inertial masses on the shaft because of coupling of two
motors like in the experimental setup described in [4].
zamfira@unitbv.ro
402 M. Matijević et al.
Fig. 9. Laboratory model of servo system with flexible coupling at Centre for Applied
Automatic Control at Faculty of Engineering at Univ. of Kragujevac
The laboratory model (Fig. 9, [5]) is equipped with web published practicum for
students [5]. These practices are aimed at introducing the users in motion control
systems [1, 5, 17, 18]. Similar to [11], all practices have several parts that users have to
do in a logical sequential order: A theoretical description of the equipment and its
operation mode, a scheme of the practice that has to be done and an explanation of the
tasks that has to be developed during the laboratory practice. The practices are detailed
briefly below:
• Practice 1: Configuration of the laboratory model. Introduction to basic servo and
motion control technology concepts. Components of motion control systems and
AC servos are discussed within the context of Yaskawa components and applica-
tions: servomotors, servo amplifiers, wiring, trial operation [5, 17, 18]. Students
should be introduced in basic principles and technologies of connection of used
components. Safety aspects of design of servo systems are encompassed.
• Practice 2: Safely and properly procedures for commissioning of a servo system.
A practical overview and operation aspects of AC servo systems and their imple-
mentation. Introduction in servo amplifier’s operation modes: How to tune different
operation modes related to referent torque or referent angular velocity? Students in
small groups, up to 4 persons, have hands on trainings about seting of desired an
operation mode of servoamplifier and turning on the laboratory model wihtout
coupling between servomotors at desired angular velocity (two decoupled ser-
vosystems are used indipendently and command/referred signal of desired angular
velocity per servomotor is settled by potenciometer) [5, 17, 18].
• Practice 3: Procedures for servo amplifier settings [5, 17, 18]. Introduction in list of
CN1 terminals, parameters, I/O signal names and functions. Parameter configura-
tions for 3.1. Function selection constants, 3.2. Servo gain and other constants, 3.3.
Position control constants, 3.4. Speed control constants, 3.5. Torque control con-
stants, 3.6. Sequence constants, 3.7. Reserved parameters. Servo amplifier settings
and speed, position and torque control loop performances.
zamfira@unitbv.ro
Laboratory Model of Coupled Electrical Drives for Supervision and Control via Internet 403
Software for this Web laboratory is written using Node.JS JavaScript framework.
Real-time communication between user and the server is ensured by Socket.IO library.
Node.JS is acting like a middleman between user and the laboratory model, where user
is communicating to the Node.JS server through Socket.IO protocol, and Node.JS is
communicating to the Matlab/Simulink via TCP/IP sockets (Fig. 10).
zamfira@unitbv.ro
404 M. Matijević et al.
One part of mentioned hands-on laboratory exercises are available for realization
via Internet. Servo amplifiers are tuned and servomotors are decoupled. A user can ran
an experiment via Web Lab GUI and change some controller parameters from
Matlab/Simulink. The experiment with coupled servomotors can be used via Web Lab
also. Because of possible hazard effects of this experiment, its use via Internet is
organized only from time to time.
In the first experiment, a user can pass an arbitrary step input to the laboratory
model and see achieved speed of the motor. With information gathered from this
experiment, user can identify model via time-domain data. Simulink model is shown on
Fig. 11.
In the second experiment, user must tune gain for P controller for the position
servomechanism.
There are two opened TCP/IP connections from Node.JS, one for each parameter.
Simulink model is shown in Fig. 12.
zamfira@unitbv.ro
Laboratory Model of Coupled Electrical Drives for Supervision and Control via Internet 405
Fig. 12. Simulink model for position control of the servo system
Last online exercise is setting up parameters for PI speed servo controller. Simulink
model is given on Fig. 13, and Web laboratory GUI is on the Fig. 14.
Fig. 14. Web lab GUI for speed control of the motor
7 Conclusions
In the majority of practical cases of servo systems with flexible coupling, where the
range of applicable gains is limited due to mechanical resonance, the problem reveals
itself in the form of sustained oscillations. The audible noise and excessive tracking
zamfira@unitbv.ro
406 M. Matijević et al.
error reach unacceptable levels as gains are increased. This phenomenon depends upon
the drive-train wear, temperature, control-loop bandwidth, offset of stator current
controller, number of motor poles, shaft sensor characteristics, etc. The most of the
aforementioned secondary effects could not be modeled and covered by the simulations
and adequate laboratory model is necessary for research and verification of servo
systems with flexible coupling.
This paper presents a laboratory model for education, training and research in servo
drives applications including servo drives with flexible coupling. This paper presents
several methods for dealing with torsional resonance in servo drives, and proposes a
special case of the IMPACT structure with simple parameters adjustment. The
antiresonant feature of the structure is not based on the exact cancellation of resonance
poles. Due to the simplicity and robustness of the proposed structure, it can be easily
applied to various flexible systems with different regulator combinations. Didactic span
of possibilities of the laboratory model (with and without flexible coupling) is a special
part of this paper. In special situation of use, the laboratory model can be used via
Internet. Its integration in the WEB laboratory and possibilities of remote laboratory
practice are described. Students have opportunity to use laboratory model for experi-
ments conducting via Internet concerning with servo drives monitoring and control
(speed, torque and position of motor shaft are supervised and controlled). Web pages of
the laboratory model are dedicated to both theoretical and practical aspects of problem
based learning concerning with this laboratory model.
Acknowledgment. Work on this paper was partly funded by the SCOPES project
IZ74Z0_160454/1 “Enabling Web-based Remote Laboratory Community and Infrastructure” of
Swiss National Science Foundation.
References
1. Vukosavic, S.N.: Digital Control of Electrical Drives, Power Electronics and Power
Systems. Springer, New York (2007)
2. Vukosavic, S.N.: Electrical Machines, Power Electronics and Power Systems. Springer,
Heidelberg (2012)
3. Matijević, M.S., Vukosavić, S.N., Schlacher, K.: Eliminating instabilities in computer
controlled motion control systems caused by torsional resonance. Electronics 10(1), 35–40
(2006)
4. Vukosavic, S.N., Stojic, M.R.: Suppression of torsional oscillations in a high-performance
speed servo drives. IEEE Trans. Ind. Electron. 45, 108–117 (1998)
5. Milanovic, M.: Development of laboratory model of coupled electrical drives for supervision
and control via Internet (in Serbian). M.Sc. thesis, Faculty of Engineering at University of
Kragujevac (2016)
6. The Go-Lab Project and the Go-Lab Portal (2016). http://www.golabz.eu
7. Tawfik, M., Salzmann, C., Gillet, D., Lowe, D., Saliah-Hassane, H., Sancristobal, E., Castro,
M.: Laboratory as a service (LaaS): a novel paradigm for developing and implementing
modular remote laboratories. iJOE 10(4) (2014)
8. Nedic, Z., Nafalski, A.: Development of online power laboratory with renewable generation.
iJOE 11(3) (2015)
zamfira@unitbv.ro
Laboratory Model of Coupled Electrical Drives for Supervision and Control via Internet 407
9. Krein, P.T., Sauer, P.W.: An integrated laboratory for electric machines, power systems, and
power electronics. IEEE Trans. Power Syst. 7, 1060–1067 (1992)
10. Huang, T.C., El-Sharkawi, M.A., Chen, M.: Laboratory set-up for instruction and research in
electric drives control. IEEE Trans. Power Syst. 5, 331–337 (1990)
11. Dom´ınguez, M., Fuertes, J.J., Reguera, P., Mor´an, A., Alonso, S., Prada, M.A.: Remote
laboratory for learning of AC drive control. In: Proceedings of the 18th IFAC World
Congress, Milano (Italy) (2011)
12. Matijević, M.S., Sredojević, R., Stojanović, V.M.: Robust RST controller design by convex
optimization. Electronics 15(1), 24–29 (2011)
13. Khan, I.U., Dhaouadi, R.: Robust control of elastic drives through immersion and invariance.
IEEE Trans. Industr. Electron. 62(3), 1572–1580 (2014)
14. Szabat, K., Orlowska-Kowalska, T.: Vibration suppression in a two-mass drive system using
PI speed controller and additional feedbacks – comparative study. IEEE Trans. Ind. Electron.
54, 1193–1206 (2007)
15. Szabat, K., Tran-Van, T., Kaminski, M.: A modified fuzzy luenberger observer for a
two-mass drive system. IEEE Trans. Ind. Inform. 11(2), 531–539 (2014)
16. Li, Q., Xu, Q., Wu, R.: Low-frequency vibration suppression control in a two-mass system
by using a torque feed-forward and disturbance torque observer. J. Power Electron. 16(1),
249–258 (2016)
17. Yaskawa America Training in Servo Basic Concepts. https://www.youtube.com/playlist?
list=PLNAENlyEDCkw0gUDF1BMu3brEJArhD8Kt, YouTube Channel, 2014 (21.11.
2016.)
18. Yaskawa eLearning Curriculum (eLearning Modules and eLearning Videos), 21 Nov 2016.
https://www.yaskawa.com/pycprd/training/elearning-curriculum/tab0/link00
19. Web Laboratory Aggregator Service http://cpa.fin.kg.ac.rs/weblab/index from the SCOPES
project I37430/160454 “Enabling Web-based Remote Laboratory Community and Infras-
tructure” of Swiss National Science Foundation, at Faculty of Engineering at University of
Kragujevac (2016)
zamfira@unitbv.ro
Online Course on Cyberphysical Systems
with Remote Access to Robotic Devices
1 Introduction
This work was supported by a grant from NASA through University of Central Florida’s NASA-
Florida Space Grant Consortium (UCF-FSGC 66016015). Partial support was provided by a WIDER
grant from the National Science Foundation, Award No. DUE-1347640.
engineering to chemical engineering to mechatronics, two related issues have not been
covered sufficiently well: (a) teaching cyberphysical systems in software engineering
programs, and (b) designing online laboratories for teaching cyberphysical systems in
such programs. There are not that many reports in the literature on dealing with related
problems or proposed solutions [11].
Consequently, the objective of this paper is to present an insight into the devel-
opment of a lab and related projects within this lab, which could help establish similar
laboratories in undergraduate software engineering programs and related disciplines.
The rest of the paper is structured as follows. Section 2 discusses the Cyberphysical
Systems course itself, Sect. 3 outlines the need for and essence of development pro-
jects, and Sect. 4 presents the contents of the lab and some results from first experi-
ences. Section 5 concludes the paper with discussion of some findings.
The Cyberphysical Systems course discussed in this paper is a new course based on a
previously offered Embedded Systems Programming course, with added networking
component and changed focus. The learning objectives for this new course have been
formulated as follows. The students will acquire:
• an awareness of the interactions of a cyberphysical system with the environment
• the ability to address sensor and control operations in cyberphysical systems
• the ability to understand the software lifecycle for cyberphysical systems
• the ability to design, analyze and document software for cyberphysical systems
• the ability to work effectively in teams to address collectively software issues in
cyberphysical systems
• the ability to complete an integrated project in the cyberphysical systems domain
• the ability to present project related material in a variety of forms
• an awareness of non-functional requirements for cyberphysical systems, such as
safety, security and reliability.
The course is offered online and its principal structure includes two major components:
lecture modules offered via the web and software development projects. There are
twelve lecture modules offered in the following order, according to the top-down
principle of software development, from high-level design to low-level hardware issues,
which proved effective in software engineering education, with selected application
topics and coverage of non-functional system and software properties at the end:
1. Introduction to Cyberphysical Systems
2. Design of Embedded Real-Time Software
3. Designing Software Architecture
4. Programming Languages for Cyberphysical and Embedded Applications
5. Real-Time Kernels
6. Advanced Real-Time Kernels Concepts
7. Timing in Embedded and Cyberphysical systems
8. Hardware Issues in Programming
zamfira@unitbv.ro
410 J. Zalewski and F. Gonzalez
The second major component of the course is the software development project, with
extensive lab activities. The essential question faced by the instructor, in this regard, is:
How to organize the software engineering lab in an online course on cyberphysical
systems, to facilitate software development? Two critical and inter-related issues are to
be considered here:
• access to the lab, or more broadly, how the lab work is to be done, whether cen-
tralized inside the lab, individual at the student’s location, or remotely online, and
• the selection of project topics, whether proposed by students, solicited from
industrial partners, a pool to choose from provided by instructor, or a single topic
across the board generated by instructor.
Regarding the lab access, the ultimate goal is to provide students with such access to
use respective devices. It must be made clear, however, that it is not meant to be only a
remote control of these devices, which is common [12] in courses in disciplines, such as
control engineering, electronics or mechanical engineering, simply to test certain mea-
surement methods or control algorithms by choosing various device parameters. In other
words, it is not only the online use of remote devices, even the most sophisticated
medical, chemical or physical instruments, which has to be provided. In software
engineering, a new qualitative step is needed, which is consistent with the mission of this
profession. Namely, software engineers develop software, so their access to remote
devices must be provided for a substantially different reason: to be able to upload soft-
ware to the target device online and test the software and debug it on the remote target.
Perhaps a word of explanation is worthwhile here, since this goal may be perceived
differently by different stakeholders. The most prominent example of what is
zamfira@unitbv.ro
Online Course on Cyberphysical Systems 411
considered is the case of the NASA’s Pathfinder mission to Mars, in 1997, when the
rover control software had a glitch involving its real-time kernel, VxWorks, and had to
be analyzed back on Earth at the mission control in Jet Propulsion Laboratory. Once
the bug was fixed the software was uploaded back to the rover on Mars [13]. A more
contemporary, but also spectacular, example that can be mentioned is remote access to
the experiments at Large Hadron Collider (LHC), in Geneva, where physicists and
engineers around the world can program their data acquisition and control systems over
the Internet (Fig. 1). In today’s terms, with the widespread proliferation of the Internet
of Things, this issue is no longer so spectacular, but gradually becomes a matter of
everyday life, and it is the responsibility of educators to adequately prepare the soft-
ware engineering workforce for this task.
With respect to the choice of project topics, a whole spectrum of lab projects has
been pursued in previous courses on Embedded Systems Programming and are
described in separate publications [14, 15]. They involve an entire array of hardware
platforms, real-time kernels and input-output devices, including single board com-
puters, microcontrollers, game boards, FPGA boards, wireless sensor networks,
Atmel/Arduino/Raspberry Pi platforms, and simple robotic devices (such as Lego,
IntelliBot, Parallax Boe-bot and multiple others). While all were suitable for previous
courses, they are not well suited for projects in this specific course, which is focused on
online access and requires more intelligence on the part of the device, especially
network connectivity. Therefore, the decision was made to attempt the use of a single
class of devices, with multiple “incarnations”, which would exhibit a wide range of
sensing and actuating elements. This led us to focusing on various kinds of robotic
devices, which additionally have networking capabilities.
zamfira@unitbv.ro
412 J. Zalewski and F. Gonzalez
Just like in case of integrating cyberphysical systems into software engineering edu-
cation, where very few papers have been published, yet [11], there are only a few
attempts to introduce robotics into software engineering curricula [16, 17]. To the best
of the authors’ knowledge, none of these publications make any recommendations on
creating such labs, except of reporting on using Lego robots, for example. Thus, there
is not much guideline material to base a software engineering robotics lab on.
Consequently, the lab created for this course was based on somewhat unstructured
principles and relied on focusing to achieve course objectives according to the fol-
lowing five assumptions:
• make sure the emphasis is on one category of equipment, that is, robotics devices,
as opposed to a different equipment, such as sensor networks, FPGA, etc.
• provide a wide variety (diversity) of robotics devices to view multiple aspects of the
software development process
• realize projects with a full but simple software development cycle, to focus on
online access to remote equipment
• ensure possibilities of invasive remote labs, that is, allowing student developers to
change robotic software by uploading updates and modifications
• address a variety of software requirements, including non-functional ones, such as,
safety, security, and reliability.
These assumptions resulted in a somewhat eclectic aggregation of robotic devices,
assembled over a period of several years, with required functionality as shown in
Table 1. Around a dozen robotic units are accessible in the lab, exhibiting an array of
zamfira@unitbv.ro
Online Course on Cyberphysical Systems 413
different properties. What is important to this course is that each device operates with at
least one networking protocol suitable for handling remote connections.
Software development projects assigned to use these different robotic devices
varied from using an existing native network connectivity (for instance, in case of
NAO) to developing network connectivity in case it was not provided by the vendor (as
for Lego EV3). Projects followed strictly the software development lifecycle, using (for
simplicity) the waterfall model divided into four phases: requirements, design,
implementation and testing. The documentation included project reports from all four
phases.
zamfira@unitbv.ro
414 J. Zalewski and F. Gonzalez
5 Conclusion
The main objective, to enhance software engineering education by the use of online
robotics lab was accomplished. Additionally, the projects were successful in shedding a
light how diversity in a specific category of equipment (robotic devices) affects
understanding of the subject matter? In this regard, various views of essentially the
same problem, as presented in project discussions, converged to a better understanding
of the necessary software mechanisms. An unexpected issue arose due to different
levels of students’ familiarity with robotic devices (a prerequisite course in robotics
was not required [18]). This was addressed by gradual introduction of complexity in
meeting software project goals. An unanswered question remains, to what extent this
kind of online labs can be invasive, that is, allow remote access to the robot to develop
software (with uploading new versions and debugging), not only test it.
Of the 3 kinds of issues always facing instructors in remote labs, administrative,
technical and pedagogy, the pedagogy outcomes can be summarized as follows:
• remote interaction and software design for robotic devices enhances understanding
of a functionality of cyberphysical systems by the use of physical inputs/outputs
• since not all students had familiarity with robots, enforcing knowledge acquisition
was diversified in a sequence: demo, exercise, assignment, experiment and project
• there was insufficient time in this edition to fully address the professional knowl-
edge of non-functional requirements, such as reliability, safety, and security.
References
1. Interim Report on 21st Century Cyber-Physical Systems Education, National Research
Council, Washington, DC (2015)
2. 3rd International Workshop on Cyber-Physical Systems, IWCPS 2016, Gdansk, Poland,
11–14 September 2016. https://fedcsis.org/2016/iwcps
zamfira@unitbv.ro
Online Course on Cyberphysical Systems 415
3. Third NSF/Georgia Tech Summer School on Cyber-Physical Systems, Atlanta, Georgia, 27–
29 June 2011. http://www.ece.gatech.edu/research/labs/esl/Activities/CPS-2011/index.html
4. International Summer School on Cyberphysical Systems, Toulouse, France, 5–9 September
2016. https://www.laas.fr/public/en/1st-summer-school-cyber-physical-systems-cps2016
5. Marwedel, P.: Course on Cyber-Physical System Fundamentals. Youtube video. Technische
Universität Dortmund (2016). http://www.youtube.com/user/cyphysystems
6. Törngren, M., et al.: Education and training challenges in the era of cyber-physical systems:
beyond traditional engineering. In: Proceedings of WESE 2015, Workshop on Embedded
and Cyber-Physical Systems Education, Amsterdam, 8 October 2015. Paper 8
7. Peter, S., Momtaz, F., Givargis, T.: From the browser to remote physical lab: programming
cyber-physical systems. In: Proceedings of Frontiers in Education Conference, FIE 2015,
El Paso, Texas, 21–24 October 2015
8. Grega, W., Kornecki, A.J.: Real-time cyber-physical systems: transatlantic engineering
curricula framework. In: Proceedings of Federated Conference on Computer Science and
Information Systems, FedCSIS 2015, Lodz, Poland, 13–16 September 2015, pp. 755–762
(2015)
9. Wade, J., et al.: Systems engineering of cyber-physical systems: an integrated education
program. In: Proceedings of 123rd Annual ASEE Conference, ASEE 2016, New Orleans,
LA, 26–29 June 2016. Paper No. 17162
10. Lee, E.A., Seshia, A.A.: Introduction to embedded systems: a cyber-physical systems
approach, 2nd edn. (2016). http://leeseshia.org/
11. Laird, L., Bowen, N.: A new software engineering undergraduate program supporting the
Internet of Things (IoT) and Cyber-Physical Systems (CPS). In: Proceedings of 123rd
Annual ASEE Conference, ASEE 2016, New Orleans, LA, 26–29 June 2016. Paper
No. 16378
12. Zubía, J.G., Alves, G.R. (eds.): Using Remote Labs in Education. University of Deusto,
Bilbao (2011)
13. Reeves, G.E.: What really happened on Mars? (1997). https://trs.jpl.nasa.gov/handle/2014/
19020
14. Zalewski, J.: Web-based labs for cyberphysical systems: a disruptive technology. In:
Proceedings of 10th IFIP World Conference on Computers in Education, Torun, Poland, 2–5
July 2013
15. Zalewski, J.: Cyberlab for cyberphysical systems: remote lab stations in software
engineering curriculum. In: Proceedings of 4th International Conference on e-Learning,
ICeL 2013, Ostrava, Czech Republic, 8–10 July 2013, pp. 1–7 (2013)
16. Göbel, S., Jubeh, R., Raesch, S.L.: A robotics environment for software engineering courses.
In: Proceedings of 25th AAAI Conference on Artificial Intelligence, San Francisco,
California, 7–11 August 2011, pp. 1874–1878 (2011)
17. Gonzalez, F., Zalewski, J.: Online robotic labs in software engineering courses, Research
Papers of the Faculty of Electrical Engineering and Automation of Gdansk Polytechnic, no.
37, pp. 15–18 (2014)
18. Gonzalez, F., Zalewski, J.: A robotic arm simulator software tool for use in introductory
robotics courses. In: Proceedings of IEEE Global Engineering Education Conference,
EDUCON 2014, Istanbul, Turkey, 3–5 April 2014
zamfira@unitbv.ro
Models and Smart Adaptive Interfaces
for the Improvement of the Remote Laboratories
User Experience in Education
1 Introduction
Remote laboratories normally have been implemented as isolated systems, which means
that they are not integrated to any other systems. In the educational context it is important
to interchange information back and forth between the remote lab and the learning
management systems (LMS), the curricular manager or at least be able to send results
of the student experimentation experience to the grading systems. This current lack of
integration can negatively impact the motivation for using remote laboratories a part of
the education processes.
zamfira@unitbv.ro
Models and Smart Adaptive Interfaces 417
Remote laboratories definitions, taxonomies, general architectures and models for the
integration of remote laboratories with academic environments have been developed
during the last couple of years [4, 5], showing the importance and the benefits of inte‐
grating this technology in the education context. To have a clear understanding about
which are the requirements of a software system, specifically a software tool, the design
of uses cases is well known as a good technique.
The following diagram describe the basic operations made by different roles such
as: students, teachers and administrators during the interaction with a remote laboratories
platform integrated with the LMS or any other virtual learning environment. This section
describes operations such as user sessions, authentication and authorization, scheduling,
user and resources management, learning objects integration, results management and
more important the student interaction process with the remote laboratory.
The laboratory reports data to the platform based on the educational standard Tin
Can Api, from which is possible for students, teachers or administrators gather infor‐
mation about the students results and general information about the use of the labora‐
tories (Fig. 1).
zamfira@unitbv.ro
418 L.F. Zapata Rivera and M.M. Larrondo Petrie
The following diagrams show the details of each use case, they include either oper‐
ation from the point of view of students, teachers and administrators as well. The first
one is the process of the user session creation in which the user validate its credentials
and is authenticated and authorized according with its role, see Fig. 2.
zamfira@unitbv.ro
Models and Smart Adaptive Interfaces 419
The teacher schedule an appointment or work session in the laboratory for a group
of students or for a specific student. The process use the student profile to take infor‐
mation about the student level, and is going to restrict the access to some content and
will suggest the type of view for the student. Once the profiles are validated the system
will check the availability and reserve the requested remote experiment, see Fig. 3.
Additionally the teacher can improve the student learning experience wrapping the
remote lab into a learning objects that can include for example: original content, or
integrate external tools such as: animations, videos or simulations. To help in the eval‐
uation process, the teacher can also integrate rubrics for the remote laboratory evalua‐
tion. To evaluate the student results during the experimentation process the platform
provides an interface to post the grades and also to give feedback to the students, see
Figs. 4 and 5.
Fig. 4. Use case for the learning experience based on learning objects.
zamfira@unitbv.ro
420 L.F. Zapata Rivera and M.M. Larrondo Petrie
The administrator role will manage the user accounts and the remote experiments
available in the platform, this include actions such as: add new resources, edit the experi‐
ments configuration of a specific experiment or even remove one of the experiments. The
following figures show the user and resources management, see Figs. 6 and 7.
zamfira@unitbv.ro
Models and Smart Adaptive Interfaces 421
The results view process will provide adapted reports according with the rights of
each role. This results will be available from the learning environment and will be inte‐
grated with the grades books of the courses, see Fig. 8.
The interaction with the experiment will be made through an adaptive interface that
adapt the view of the laboratory providing the student more or less control of the equip‐
ment according with his level of knowledge, see Fig. 9.
zamfira@unitbv.ro
422 L.F. Zapata Rivera and M.M. Larrondo Petrie
applications, among others. Some advances have been done in the user interfaces for
Virtual Laboratories [6] and Remote Laboratories [7–10] that provide alternatives to
improve the user experience with the remote laboratory.
The following simplified architecture shows the remote laboratory which informs
on the student level and according to the complexity of the experiment, the laboratory
provides one of the views of the experiment, these views can block or hide some of the
controls, restricting the view of part of the controls or blocking the option of configuring
some others (view 1, view 2 and view 3).
One of the new concepts proposed is to integrate the remote lab and its interface as
a learning object into the learning environment. The students start using a very simple
version of the remote laboratory based on the difficulty level and their knowledge level,
as their knowledge and mastery level increase, the remote laboratory can turn on more
controls, increasing the complexity of the experiment. Information must flow in both
directions, from the student profile to the remote laboratory and vice versa, keeping
updated the student level and the remote laboratory interface (Fig. 10).
Fig. 10. Different user visual interfaces and the integration with the academic platform
4 Conclusions
The paper describes the basic use cases and some details the behavior of the smart
adaptive interfaces for remote laboratories, based on the integration of the platform of
the laboratories with the learning environments.
Smart adaptive interfaces will create a more meaningful experience for the students,
guiding them in a process that will change the difficulty level by changing the interface,
this creates a whole new learning experience in which the teacher and administrators
will be participants in the process.
By the implementation of this design it is expected to have, in the near future, wider
use of remote laboratories adapted to the requirements of the curriculum.
zamfira@unitbv.ro
Models and Smart Adaptive Interfaces 423
The development of the IEEE P1876 Networked Smart Learning Objects for Online
Laboratories Standard, will generate a unified view about the meaning of a learning
object that includes remote experimentation as a main component.
References
zamfira@unitbv.ro
Empowerment of University Education
Through Internet Laboratories
Abdallah Al-Zoubi(&)
1 Introduction
Internet-based remote and virtual laboratories are positioned in the heart of the edu-
cational reform agenda worldwide. This emerging paradigm has already had a pro-
found impact on engineering education and in the process of advancement and
consolidation of universities and their role in socio-economic development [1]. There is
ample evidence that online laboratories, particularly remote labs, have a positive effect
on students’ learning, practical experience and engineering skills, and may offer a great
potential for cooperation and collaboration among teachers, researchers and profes-
sionals as well as higher education institutions all over the world. In addition, the use of
remote laboratories as a complementary distance-learning tool has already refuted the
misconception of any possible threat to replace conventional ones [2].
The state of utilization, advancement and profiteering of remote labs in the Middle
East has been timid and prudent, to say the least. Limited attempts, mainly as individual
efforts, have been made to promote the implementation of remote labs as an important
educational tool in universities across the region. Most of the pioneering researchers in
the field are based in institutions outside the region and no serious indigenous and
sustainable education infrastructure can be detected in the literature. In addition, the
patterns of cooperation and collaboration among universities in the Middle East are
limited, modest and fragile. The research community is therefore urgently invited to
establish collaboration links in order to strength its capacity of influencing the adoption
of online labs in the region.
In this paper, an initiative to disseminate a culture of online engineering through the
utilization and integration of a communication engineering lab at Princess Sumaya
University for Technology is presented. First, a brief on a number of collaboration
projects with universities in Europe and USA that demonstrate the process of capacity
building in the field is described. The design choice, methodology and associated IT
infrastructure is then discussed in details and results presented. The authors aim to
serve students and teachers in the Middle East to benefit greatly from using the lab
without restrictions and to increase the quality of teaching and learning process by
allowing them to work remotely. This endeavour may also serve as an attempt to
establish science partnerships between countries and can give researchers the oppor-
tunity needed to help boosting the educational systems of their countries.
zamfira@unitbv.ro
426 A. Al-Zoubi
The design of the remote lab system architecture has taken several criteria into con-
sideration in order to widen the use and implementation of the technology in the
country. The first criterion was ensuring participation of a large number of students
from different universities in performing experiments easily and freely. Cost was also
an important factor which was decisive in the design preference due to limited
resources. The system consequently has followed the scheme shown in Fig. 1.
The actual hardware experiments were accommodated by the NI ELVIS work-
station and an EmonaFOTEx board. The workstation is a comprehensive tool for
teaching subjects such as circuit design, instrumentation, controls and telecommuni-
cations as it offers the flexibility of virtual instrumentation and the ability of cus-
tomizing application. It is a hands-on design and prototyping platform that integrates
the most commonly used instruments such as oscilloscope, digital multi-meter, func-
tion generator and bode analyzer into a compact form ideal for the lab or classroom.
The EmonaFOTEx board, on the other hand, has been specifically devised for teaching
introductory fiber-optic telecommunications topics in electrical and computer engi-
neering curricula. The board allows students to create communications systems by
wiring together system components with plug-in fiber-optic patch cables. This is
facilitated with detailed user documentation and a lab manual which contains
step-by-step detailed procedures to build, experiment and measure communications
systems. With this particular board, students could perform up to 11 experiments on the
theory of fibre optics communication. Other boards can also be used to conduct
experiments on analogue and digital communication circuits.
zamfira@unitbv.ro
Empowerment of University Education Through Internet Laboratories 427
The workstation was connected to a virtual instruments (VI) server where all
simulated instruments and equipment such as oscilloscope, function generator and
various fibre optics communication components were programmed and accommodated.
The VIs were programmed by LabVIEW while the VI server acted as an intermediary
between the ELVIS workstation and web server which hosted the learning management
system (LMS), the Moodle open-source platform in this particular case, through which
students could access the experimental hardware.
Moodle was preferred to proprietary platforms due to limited financial resources
and its powerful features such as ease of use and the wide supports it provides for a
large community of users. It was used as an entrance point for students to facilitate and
ease the provision of online classrooms by means of integrated features and tools such
as administrative, synchronous and asynchronous communication, assessment and
tracking, and multimedia sharing tools. One can apply all such services provided by
platform which accommodates the remote practical laboratory sessions. The platform is
actually based on PHP language and operated under an open source Apache web
server. In particular, Moodle 2.4.3+, PHP 5.4.7 and MySQL 5.5.27 version was
selected. It was supported by plugins and other software tools to manage e-content,
virtual class rooms, assignments, task submissions and grading, quizzes, exams, tasks
queue, lab booking, and scheduling online experiments. The VPN and server IP
addresses were located at 193.188.67.34 and 172.31.0.53 respectively, and physically
hosted at Princess Sumaya University for Technology servers which can be accessed at
the link http://eqtel-vle.psut.edu.jo.
Suitable educational and teaching material were selected to suite students and
professors while care was taken to integrate the remote lab to the LMS such that
experiments were delivered with full interactivity. The learning tools included syn-
chronous mode such as chat rooms and video conferencing for instant real-time
interaction; and asynchronous modes for students and teachers’ communication
through email and forums, assessment and evaluation, as well as cooperation and social
zamfira@unitbv.ro
428 A. Al-Zoubi
media tools, thus offering flexibility in delivery and providing the opportunity to share
social learning material amongst several institutions. In addition, four types of accounts
were created for administrator, teacher, student and guest. Appropriate privileges were
given to these accounts according to their roles. Students could access the content of
the lab, experiment, by visiting the main page. A description of each topic was given
prior to accessing the actual content of each experiment to enable students to famil-
iarize themselves with the scientific content. The interactive electronic content could be
accessed by clicking on “eContent” link and student could navigate easily using pre-
vious and next buttons and a table of content designed as a menu on to easily allow
access of a specific lesson.
In addition, a Microsoft SQL server was used to host tables of data for the purposes
of access and authentication, storage and retrieval of experimental data, and students’
feedback. The first table included a login student’s ID and name. In addition, for each
experiment, a table was created to allow students to answers questions in every single
experiment by filling in data into fields reserved specifically for this purpose as the
student was performing the experiment and executing step-by-step procedures. The
fields were created using available components in Microsoft Visual Studio ASP.NET
framework, and their dimensions such as height and width were adjusted carefully by
editing layouts using a web development tool called the inspect element that allow
designers to test, debug, and mainly edit the code for each online experiment. An
additional table was reserved for the feedback questionnaire at the end of each
experiment. In total, 13 tables were created in the database.
4 System Operation
The operation of the system is explained following the steps depicted in the flowchart
shown in Fig. 2.
zamfira@unitbv.ro
Empowerment of University Education Through Internet Laboratories 429
Students should then choose the desired experiment from amongst the list of 11
experiments, and read the appropriate step-by-step procedure.
A brief description on the experiment appear such as its name, documentation,
number of sessions that is allowed for each student, duration available to carry out the
experiment and server time. Student have to book a specific time slot on a specific date
to enter the experiments as shown in Fig. 4, one student at a time in order to avoid
conflicts in requesting and using VIs.
A student would then wait for her/his reserved time before being able to enter the
experiment. When the session starts, the experiment login page will appear as an iframe
inside the experiment page and the student logs in again using .Net+SQL server
authentication. An SQL connection object will check if the entered student ID and
name were already existing in the database.
zamfira@unitbv.ro
430 A. Al-Zoubi
A third segment was the user’s interface that included the experiments forms in
ASP.NET components for input fields and HTML pages. Since various types of input
fields such as submit button, survey page, and students’ answers, a solution to connect
all together as a one system was sought. The ‘view state’ method has consequently
been used in the code to reserve the entered answers in case of any unpredictable error,
or in case of page reloading. However, this method could not be applied to the
screenshots answers since images were clients’ property.
An example for a sample experiment is Nyquist in PCM which is illustrated below
to showcase the experimental procedure a student would encounter while performing
the experiment. The definition of PCM modulation was first given as a kind of source
coding which was a conversion of a signal from analog to digital. The Nyquist theorem
states that a signal must be sampled at a rate greater than twice the highest frequency
component of the signal to accurately reconstruct the waveform. In performing this
experiment, the student would first be asked to install the run time engine program,
only once at the beginning of undertaking the lab, to operate each VI.
zamfira@unitbv.ro
Empowerment of University Education Through Internet Laboratories 431
The next step would be to activate the functions generator, scope and DSA, and the
VI page appears as shown in Fig. 5. Then, adjustment of the function generator using
its soft controls to produce a sinusoidal wave output with specific frequency and
amplitude were performed. The oscilloscope could be adjusted for amplitude and
frequency in each of its two channels with appropriate DC offset using the knobs for
amplitude and time/division.
Fig. 5. The VIs of the DSA, scope and function generator as they appear during the experiment.
zamfira@unitbv.ro
432 A. Al-Zoubi
The student would then follow a step-by-step adjustment and measurement pro-
cedure by setting the DSA with values as shown in the tables below:
Units to dB
Mode to RMS
Frequency
Display
Scale to Auto
Mode to RMS
# of averages to 3
The student would then be asked to run the oscilloscope at Channel (0) and to take
a screenshot for the input signal and upload into the system database. The peak-to-peak
voltage of the signal was reordered, the DSA was run and a screenshot for the input
power signal was taken and uploaded. The student would then be asked to determine
the power signal for the input signal with the following values where the required input
field have been created for students to fill with the recorded data obtained.
zamfira@unitbv.ro
Empowerment of University Education Through Internet Laboratories 433
15
20
25
30
35
The student was consequently instructed to change to the output channel (1), to take
a screenshot for the output signal an upload, record the peak-to-peak output voltage,
take and upload a screenshot for the output power signal and determine the output
power signal at given frequency values as below:
15
20
25
30
35
Screenshots of both the input and the output signal from channels (0) and (1) would
then be taken and uploaded, and the frequency of the first six aliases in the sampled
message were determined and recorded in the field as below.
zamfira@unitbv.ro
434 A. Al-Zoubi
1.
2.
3.
4.
5.
6.
Finally, a screenshot for the first aliasing waveform was taken and uploaded and the
results submitted. After the students submitted the answers in the experiment page, a
survey page will appear with seven questions that must be answered and a suggestions
block to be filled as well to enable the design team to enhance or apply advantageous
marbles in future modifications to the system. The answers were eventually stored at
the survey’s database. Full details of operation and experimental procedure could be
viewed at https://youtu.be/VAdJJ454_C8.
The remote lab was first operated at the university in the second semester of the
academic year 2015/2016, were 30 students registered in the traditional communication
engineering lab offered as part of the requirements of the bachelor degree programme
on communication engineering. The students were requested to perform only 3 online
experiments out of 11 available which could be open at any time upon request. The
remaining required experiments were physically conducted by students as part of the
regular traditional lab. This approach was followed in order to gradually introduce the
online lab into curriculum with the minimum of ramifications and subsequently fertilize
a paradigm shift and disseminate a culture of technology-enhanced learning.
The feedback survey consisted of 7 questions related to the quality of the online
operational manual and the information it provided to set up and run the experiment,
easiness of conducting the experiments, understanding practical aspects of communi-
cation system, sense of operating the virtual equipment, flexibility to fit the laboratory
into schedule, future preference and choice of using remote experiments over traditional
ones and overall rating of performance of the remote lab. Ratings were based on a scale
from 1 to 5 (1 very low to 5 very high). The survey was actually aimed at investigating
students’ comprehension, perception and views, and satisfaction rather than seeking
acceptance or consent or even trying to reach a verdict on integrating remote labs into
the curriculum at this early stage of development. However, initial results, shown in
Fig. 6, have already shown promising prospects with an average of 78% of students
expressed satisfaction with the manual of the remote lab providing enough information
for easy setup of the experiment, 96% found it was easy to conduct the experiment,
zamfira@unitbv.ro
Empowerment of University Education Through Internet Laboratories 435
72% understood the practical aspects of communication systems, 76% felt like they
were operating real experiments, 87% were allowed to fit the laboratory work into their
schedule, 83% rated favourably the overall performance of the remote lab.
When asked to provide suggestions and observations to improve and enhance the
process, procedures and operation of the remote lab and experiments, 18 students had
no suggestions while 5 positively and favourably commented on its flexible, great
experience and time and efforts saving, and 12 students required the need to have more
trails and time to perform the experiment. The trail at hand may offer the university,
and higher education institutions in Jordan and beyond, the opportunity to enhance its
capacity to productively engage in development challenges and contribute to the
indigenization of technology-enhanced education through a viable context-based
approach. It is indeed an attempt to turn teaching tools inward in order to unpack the
foundations of university teaching, learning and research environments.
A full online lab was subsequently offered at Princess Sumaya University for
Technology in the summer semester of the academic year 2015/2016 and partially in,
Yarmouk University, University of Jordan and Hashemite University. The total number
of students in all universities was 61. Strikingly similar evaluation results to the pre-
vious semester were observed with 80% of students expressing satisfaction with the
use, implementation, performance and preference of the remote lab. Despite some
initial technical difficulties, it is pretty safe to assume that students indeed find the
implementation of the remote lab in university education practical, interesting and
challenging. Subsequently, the same lab is being offered at the current fall semester of
the academic year 2016/2017 in five different universities in Jordan: Princess Sumaya
University for Technology, Yarmouk University, University of Jordan, Hashemite
University and Al-Hussein Bin Talal University, covering a wide geographical area of
the country.
zamfira@unitbv.ro
436 A. Al-Zoubi
6 Conclusions
Acknowledgment. The author would like to acknowledge the generosity of the European
Education, Audiovisual and Culture Executive Agency (EACEA) for the support provided to the
TEMPUS project entitled: “Modernizing Undergraduate Renewable Energy Education: EU
Experience for Jordan”, number 530332-TEMPUS-1-2012-1-JO-TEMPUS-JPCR and the
TEMPUS project entitled: “Enhancing Quality of Technology-Enhanced Learning at Jordanian
Universities”, number 544491-TEMPUS-1-2013-1-ES-TEMPUS-SMGR. Additionally, the
author are especially grateful to Professor Manuel Castro and his team at the Electrical and
Computer Engineering Department (DIEEC) of the Universidad Nacional de Educación a Dis-
tancia (UNED), Madrid, Spain, for their encouragement, support and corroboration throughout
the life time of the project.
References
1. Salah, R.M., Alves, G.R., Guerreiro, P.: IT based education with online labs in the MENA
region: profiling the research community. Int. J. Hum. Capital Inf. Technol. 6(4), 1–21
(2015)
2. Salah, R.M., Alves, G.R., Guerreiro, P.: A federation of online labs for assisting engineering
and science education in the MENA region. In: 3rd International Conference Technological
Ecosystems for Enhancing Multiculturality, Porto, Portugal, 7–9 October 2015
3. Auer, M.E., Zutin, D.G., Al-Zoubi, A.Y.: Online laboratories for eLearning and eResearch.
In: The 5th Congress of Scientific Research Outlook and Technology Development in the
Arab World (SRO5), Fez, Morocco, 25–30 October 2008
4. Auer, M., Al-Zoubi, A.Y., Zutin, D.G., Bakhiet, H.: Design of application-specific integrated
circuits for implementation in a network of remote labs. In: 2008 ASEE Annual Conference
and Exposition, Pittsburgh, Pennsylvania, USA, 22–25 June 2008
5. Danilo, G.Z., Auer, M.E., Al-Zoubi, A.Y.: Design and verification of application-specific
integrated circuits in a network of remote labs. Int. J. Online Eng. (iJOE) 5(3), 25–29 (2009)
6. Auer, M., Al-Zoubi, A.Y., Zutin, D.G.: Design and implementation of mobile clients for
remote labs. In: 5th International Symposium on Remote Engineering and Virtual
Instrumentation, Düsseldorf, Germany, 22–23 June 2008
zamfira@unitbv.ro
Empowerment of University Education Through Internet Laboratories 437
7. Auer, M., Zutin, D.G., Al-Zoubi, A.Y.: Implementation of a mobile accessible remote lab.
Int. J. Interact. Mobile Technol. (iJIM) 2(3) (2008)
8. Jeschke, S., Al-Zoubi, A.Y., Pfeiffer, O., Natho, N., Nsour, J.: Integration of an online digital
logic design lab for IT education. In: The SIGITE 2008 Annual Conference, Cincinnati, OH,
USA, 16–18 October 2008
9. Al-Zoubi, A.Y., Nsour, J., Jeschke, S., Pfeiffer, O., Richter, T.: An electronic engineering
mobile remote laboratory. In: 8th World Conference on Mobile and Contextual Learning,
Orlando, Florida, USA, 26–30 October 2009
10. Al-Jufout, S., Al-Zoubi, A.Y., Jeschke, S., Nsour, J., Pfeiffer, O.: Application of remote
technology to electrical power system laboratory. In: The International Conference on
Education and New Learning Technologies, EDULEARN 2009, Barcelona, Spain, 6–8 July
2009
11. Jeschke, S., Pfeiffer, O., Al-Zoubi, A.Y., Nsour, J., Al-Jufout, S.: Web-based laboratories for
distance engineering education in Jordan. In: Sloan-C: International Symposium-Emerging
Technology Applications for Online Learning, San Francisco, USA, 17–19 June 2009
12. Jeschke, S., Al-Zoubi, A.Y., Pfeiffer, O., Natho, N., Nsour, J.: Classroom-laboratory
interaction in an electronic engineering course. In: International Conference on Innovations
in Information Technology, IIT, Al Ain, United Arab Emirates, 16–18 December 2008,
pp. 337–341 (2008). ISBN 978-1-4244-3396-4
13. Abu-Aisheh, A.A., Eppes, T.E., Otoum, O.M., AlZoubi, A.Y.: Remote laboratory
collaboration plan in communications engineering. In: IEEE Second Global Engineering
Education Conference (EDUCON), Learning Environments and Ecosystems in Engineering
Education, Amman, Jordan, pp. 837–840, 4–6 April 2011
14. Abu-Aisheh, A.A., Eppes, T., Al-Zoubi, A.Y.: Implementation of a remote analog and
digital communications laboratory for eLearning. Int. J. Online Eng. (iJOE), 6(2) (2010)
15. Alkouz, A., Al-Zoubi, A.Y., Otair, M.: J2ME-based mobile virtual laboratory for
engineering education. Int. J. Interact. Mobile Technol. (iJIM) 2(2) (2008)
16. Al-Zoubi, A., Hammad, B., Ros, S., Tobarra, L., Hernández, R., Rafael, P., Castro, M.:
Remote laboratories for renewable energy courses at Jordan Universities. In: IEEE Frontiers
in Education Conference Proceedings, Madrid, Spain, 22–25 October 2014
17. Tobarra, L., Ros, S., Hernández, R., Pastor, R., Castro, M., Al-Zoubi, A.Y., Hammad, B.,
Dmour, M., Robles-Gómez, A., Caminero, A.C.: Analysis of integration of remote
laboratories for renewable energy courses at Jordan Universities. In: IEEE Frontiers in
Education Conference Proceedings, El Paso, Texas, USA, 21–24 October 2015
18. Tobarra, L., Ros, S., Pastor, R., Hernandez, R., Castro, M., Al-Zoubi, A., Dmour, M.,
Robles-Gomez, A., Caminero, A., Cano, J.: Laboratories as a service integrated into learning
management systems. Int. J. Online Eng. (iJOE) 12(9), 32–39 (2016)
zamfira@unitbv.ro
Expert Competence in Remote Diagnostics - Industrial
Interests, Educational Goals, Flipped Classroom
& Laboratory Settings
1 Introduction
Engineering skills and competences that are desired from industry are important to inte‐
grate in the education system. This is to increase the employability of engineering grad‐
uates as well as to provide an education platform further education adequate also for
engineers active in industry. In the end, this is expected to result in increased competi‐
tiveness, productivity and innovation rate within the industrial system. The importance
of recruiting staff with suitable competences has been examined in several studies, for
example in an EU study from 2010 [1].
In engineering, there is a gap between the skills developed at universities and the
skills required by the industry [2]. Academia has an important role in bridging this gap.
zamfira@unitbv.ro
Expert Competence in Remote Diagnostics 439
zamfira@unitbv.ro
440 L. Claesson et al.
for universities with the task of strengthening Sweden’s competitiveness and ability to
create value [12].
The project is divided into three categories (a, b and c) based on the following areas:
(a) Electric power transmission, (b) Mechanical power transmission and (c) Smart home
and health applications. Furthermore, the design of three courses within in the area of
smart home and health applications will be described in Sect. 2.
Online Engineering can be described as methods to control and monitor physical
equipment, but also smart home and people, remotely over the Internet. In Sect. 3 the
experimental setups in the remotely controlled laboratories and the design of the instruc‐
tions for the experimental work in the remote laboratory will be described.
Furthermore, the pedagogical design, Flipped Classroom, implemented in the
courses will be discussed in Sect. 4. Flipped Classroom is a form of blended learning
which encompasses any use of Internet technology to leverage the learning in a class‐
room, providing teachers with more time to spend on interacting with students instead
of lecturing. This is commonly implemented with e.g. the aid of teacher-created videos
that students may study outside their scheduled time for lectures [13, 14].
Finally, in Sect. 5 the result of a questionnaire answered by students enrolled in the
courses will be presented.
This study aims to answer two questions;
• How can academia make engineering courses of relevance to industry?
• How can we make use of resources in academia such as laboratory resources, and
flipped classroom as pedagogical approach to reach the educational goals?
The courses are divided into two blocks, one block that covers methods for diagnosis
and one block concerning technology for remote diagnosis. The courses in the block
concerning methods for diagnosis are based on each other while the courses in the block
which covers technology can be studied independent of each other. Each course with
green frame in the block diagram shown in Fig. 1 results in 3 European Credit Transfer
and Accumulation System credits (ECTS-credits), except for the Ethics and Architec‐
ture’s lecture series. One academic year corresponds to 60 ECTS-credits.
The course block that concerns methods for diagnosis is divided into three special‐
izations; Mechanical systems, electrical systems and Smart homes and health applica‐
tions. Each course consists of six lecture sessions (2 × 45 min), six exercises (1 × 45 min)
and one laboratory session (3 × 45 min). Each course is given during a period of eight
weeks. Each course is given as a distance course and includes experimental work on
remotely controlled experimental setups. Thus, the lab assignments in the courses are
carried out remotely. Adjustments to the course schedule were made based on interest
from the companies involved. The courses include application modules and these can
be adjusted according to the participants’ interests.
Furthermore, the first time a course was given within the project only employees
form the companies participating in the project attended the course. The second time a
course was given both students from academia and industry may attend the course.
zamfira@unitbv.ro
Expert Competence in Remote Diagnostics 441
Fig. 1. Block diagram of the courses developed in the project and how different courses and areas
of expertise support a remote diagnostic system. Real equipment, with control, calculations and
automation inside blue frames and courses inside green frames.
The pedagogical approach utilized was Flipped classroom, in essence, the students
were provided with course material before each lecture, and the lecture focused on
deeper discussions and analysis concerning the course subject. In some of the courses
an inquiry was compulsory for the students to fill in before each lecture. The enquiry
was used as a tool to identify if the students had read and understood the lecture material.
In essence the considered courses within Smart home and health applications were
as follows.
The technical metrology course was held in cooperation with the SP Technical Research
Institute of Sweden. SP works close in cooperation with industry and academia, and
evaluate technologies, material, products, and process and provide an effective link
between research and commercialization [15]. Metrology is a special niche for SP, it is
Sweden’s national metrology institute.
In the course 11 students participated, company representatives. The course was
divided into three courses corresponding to the subject division in the project with elec‐
trical, mechanical and smart home & health applications.
The whole series of lectures in this course followed the well-known quality-assur‐
ance loop (originally formulated by Deming) [16]. Lecture 1 - A general lecture with
common concepts were held, furthermore specifics related to the different focus areas
were presented. Lecture 2 - Metrology with human, Lecture 3 - Temperature, fall and
zamfira@unitbv.ro
442 L. Claesson et al.
zamfira@unitbv.ro
Expert Competence in Remote Diagnostics 443
At the end of 2006, the Department of Applied Signal Processing at BTH started a
project known as Virtual Instrument Systems in Reality (VISIR) together with National
Instruments in USA and Axiom EduTech in Sweden to disseminate the online work‐
bench concept created at BTH using open source technologies in collaboration with
other universities and organizations [18, 19]. Apart from BTH, five universities in
Europe have set up VISIR online laboratories for electrical experiments, (1) University
of Deusto, Bilbao, Spain, (2) The National University of Distance Education, Madrid,
Spain, (3) Carinthia University of Applied Sciences, Austria, (4) FH Campus WIEN,
Wien, Austria and (5) Instituto Politécnico do Porto, Portugal [20–22].
Lecture 1 Introduction
The course was introduced and the curriculum and course layout were presented. Flipped
classroom as pedagogical approach were presented, as were the course structure with
lectures, experimental work, literature, schedule, grading and expectations on the
students. In all the lectures the course leader holding Ph.D. in computer science were
present, thus holding an engineering perspective throughout the course.
zamfira@unitbv.ro
444 L. Claesson et al.
Fig. 2. Sensor based bio-rhythms of individual anonymized personal data with time on the x axis
and number of events on y axis.
zamfira@unitbv.ro
Expert Competence in Remote Diagnostics 445
Fig. 3. The architecture of the smart home lab. X11 – sensor Fibaro: light, temperature and motion,
10 – sensor VISION: toilet door, open – close, 9 – sensor, Fibaro, living room, door, open – close, 7 –
motion detector, SP814-1
zamfira@unitbv.ro
446 L. Claesson et al.
engineered to blend in. Thus a person may in principle pass through the whole apart‐
ment without observing the sensors/actuators.
The lab setting has been engineered in cooperation with industry, thus it is an indus‐
trial platform with on the market sensors and software solution in use in everyday smart
home applications more specifically in care homes or in elderly and/or disabled persons
home/s. The system includes remote control and functions enabled via a smart phone.
zamfira@unitbv.ro
Expert Competence in Remote Diagnostics 447
of the smartphone fall detection application and geographical position to inform where
the fall has taken place are visualized.
As a proof of concept for simulating an elderly person falling down in their home,
a set of automation rules were written, to keep track of where in the apartment the person
was and to automatically notify when and where a person had fallen.
As MQTT is used by many different companies in IoT, it was interesting to see if it
could be integrated with other solutions. The basic protocol is always the same, but the
structure and naming of the events differ slightly between operators. The IBM IoT plat‐
form was one case study where it was shown that commands on their MQTT broker
could be translated and sent to and from the smart home gateway, making it possible for
solutions running in the IBM cloud to interoperate with the smart home system.
The reason for selecting z-wave are from an academic - industrial perspective, and
in one of the lectures in the smart home & health applications course a detailed presen‐
tation and discussion about other industrial standards, advantages and disadvantages
were presented. A bridge to IBM open system on GitHub has been engineered.
Learning Outcomes: To understand and manage large amounts of data to manage data
files and visualization.
zamfira@unitbv.ro
448 L. Claesson et al.
There are large amounts of data that can be used in the home appliance and health
applications [24]. Some are open, others are locked in the example. Various companies
and public bodies. To access, manage, visualize and understand this type of data is a
desirable qualification in the home appliance and health applications. Use the resources
available within the application, Diabetes, to develop a research question and visualize
data and services related to this application.
In the smartphone application in the Fig. 5 the accelerometer was presented. The
students downloaded the app themselves and then used it remotely and got access to the
log and could follow the lecture on threshold values.
Furthermore, we used two large companies’ standard solutions on IoT and meas‐
urements and metrology for the lab.
Initially, it is important to prepare students what the pedagogical method Flipped Class‐
room is. The teacher in a course starts with an introduction and explain what education
method flipped classroom implies to the students. This part is very important to get active
students in the online lectures.
zamfira@unitbv.ro
Expert Competence in Remote Diagnostics 449
In the course Analysis and modeling of smart home and health applications, in a
learning management system Its Learning, a student finds materials to read at least one
week before each session. The materials contain compulsory quiz to be answered.
The teacher can document which material the students had shown interest in before
lesson, but it was not possible to see if it really was read by students or just downloaded.
To find out if the documents was read, not just downloaded, a quiz were sent out to the
students. The quiz to be completed by the students contained information that must be
solved by the students. The number of correct answers could then give an indication to
if the students had read the article or the material before the lesson.
Furthermore, the students can provide feedback on whether the material needed to
be updated or adapted to their situation or wishes of content.
5 Results
After the sessions a questionnaire was passed to the students to acquire their opinion
about the courses. The questionnaire had 5 open ended questions. In total 17 students
responded to the questionnaire. The questionnaire in its full form and the answers to it
are presented below.
Questionnaire.
1. Did you find the course had relevance to the industry your working in?
2. If, what kind of job do you have, for example, business, academia, self-employed?
3. Has this course given you the opportunity to combine work with studies, if so, how?
4. What did you think about getting articles and study questions before you entered the
lecture? (Pedagogical approach Flipped classroom)
5. What is your opinion about the lab exercises 1 and 2?
The open ended questions answer from students were;
1. Yes (8), No (8), I do not work (1).
Comments from 7 students: The course has clearly been relevant! The course was
built on the “flipped classroom” and been so open has suited me very well. Person‐
ally, I appreciated the openness and formlessness (that’s the best word I am going
on right now) the issues very much, because by its nature given me a “nudge in the
right direction,” while they were big enough (not to be “lookup issues “), and made
room for me and my way of working. I definitely think that through my work with
the course got a good overview of smart home and its applications for health and
feel far better equipped to possibly one day working in the fields. IoT. I can develop
in my work, but most interested in self-interest. Not linking right now but work with
similar tools and I will hopefully get use of the education in future.
2. Automotive Industry (5), Calibration Technician (1), IT industry (3), Electrical
Engineer (3), Health Centre (1), Science Center (1) Manufacturing (1) and
Production (1).
3. Very good because assignments deadline was at the end of the course and it was up
to each individual to plan their work. Did thesis at the same time and was thus outside
and worked during daytime. I spent 1-2 h a week on the studies associated with the
zamfira@unitbv.ro
450 L. Claesson et al.
course. Had no time for the course. Did not read the course at all because I did not
have time.
4. The pedagogical design Flipped Classroom suits me very well.
5. Very good! I liked that the first lab was early and with the layout. It gave me the
inspiration to think about smart home opportunities and span, present and future
situation, etc. Thought about other lab because I had to go a little deeper in a quite
specific application, and work on the sensor level. The laboratory then supplemented
each other very well (a general laboratory system-level, application-specific labo‐
ratory component level) and inspired at least me (again, thanks to data transparency
and formlessness) to think freely and to take it beyond the assignments.
6 Conclusion
Academia usually educates students, in this approach a novel way to educate regular
students and industry staff has been performed. Furthermore, for increased competence
as a level of expertise for innovation for the already employed in engineering industry.
Conclusively, the answers to the two questions are as follows;
How can academia make engineering courses of relevance to industry?
• Hard to find students holding the requirements
• Hard to find students having the time being a student and performing work in parallel
• The production line of the companies needs to be taken into consideration (that is
holding a product deadline are main priority for the companies)
• Trust
• Business awareness
• Measuring innovation - over time - how
• Education platform - how to reach students - how to measure their process
The questionnaire reveals that equal number of persons had found the course of
interest to the industry they worked in as not of interest. In the second question of how
we can make use of academic resources of flipped classroom as pedagogical approach
to reach the education goals, one challenge has been to encourage the students to go
through the preparatory material in advance. The tools used by the teachers for meas‐
urement in the learning platform show if the student downloaded the literature, but not
if the student has read and understood the literature, it has been complemented with a
survey. A survey holding questions that potentially can be answered correct if the given
literature has been read.
Meanwhile, flipped classroom technology lead to greater effectiveness through a
more active learning. Some specific goals, the most active learning way, was that:
• Encourage students to more rules-bound prepare for each lesson
• Helping teachers to better identify students’ difficulties in good time to adjust the
learning
• Helping students to develop a stronger “need to know”
• Establish an interactive environment in the ‘classroom’
zamfira@unitbv.ro
Expert Competence in Remote Diagnostics 451
References
zamfira@unitbv.ro
Parallel Use of Remote Labs and Pocket Labs
in Engineering Education
Abstract. This paper shows, how Pocket Labs, being the latest trend
in engineering education, can be used together with already established
Remote or Online Labs. Not only technical aspects, but also didactical
methods and student’s motivation have to be considered.
1 Introduction
Remote Labs are already existing for several years. In this concept, laboratory
hardware is located at a place, where storage and maintenance of the exper-
iments is possible, which mostly is a University or School campus. Students
access the lab exercises via the Internet using a service broker; an example is the
iLab Shared Architecture (ISA) [1], a Web services based distributed software
framework to manage heterogeneous remote labs. The federation model of ISA
allows for an easier sharing of remote labs across different institutions assuming
the remote labs implement the ISA Web services API.
Pocket Labs became considerable when the prices of hardware reached such a
low level that it was possible to provide each student with his or her own piece of
laboratory equipment. This paper discusses the combination of both principles,
which was first time carried out in Fall 2016 at CUAS [2,3].
2 Lab Concepts
Online or Remote Labs provide a solution for students to perform laboratory
exercises at a self-chosen time and also from remote places, mostly at home.
As personal interaction with the measurement and experimentation object is a
critical issue for engineering students, especially Remote Labs contain real hard-
ware. Access to this hardware is provided via web interfaces and very often also
via video showing the exercise itself. Nevertheless, students are not in physical
contact with the experiment.
As Pocket Labs provide actual and physical contact of students with real
hardware, they can be used especially for basic exercises and for students without
much experience. If electric or electronic components are used, a limitation will
be for standard and cheap components, as any other solution will not be feasible
due to financial reasons.
c Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6 42
zamfira@unitbv.ro
Parallel Use of Remote Labs and Pocket Labs in Engineering Education 453
Figure 1 shows different lab infrastructure concepts, starting with the classic
lab, where students have to come to the University campus and perform their
exercises in a predefined time. Remote Labs have the advantage that exclusive
and expensive hardware has to be installed and maintained only once and can be
shared among students. Finally, Pocket Labs bring students again together with
lab hardware, combined with the advantage of more or less free-chosen location
and time.
3 Didactic Considerations
As the concept of Pocket Labs requires different teaching methods, some didactic
aspects have to be considered. Especially the difference between full-time and
part-time or evening students, who have to accomplish their studies parallel to
a regular job, has to be taken into account.
3.1 Motivation
As part-time students always try to optimize, they might skip the exercises at
all or part of them, thus taking into account lower grades. To prevent this, the
laboratory exercises have to be integrated into the respective courses in such a
way, that they are not only essential for passing the course, but also serve as an
obvious tool for the understanding of the course topics.
This situation is slightly different for full-time students. Usually, they need
more instructions, guidelines and “helping hands”, which makes it more difficult
to leave them alone with laboratory exercises. In that case, it is even more
important to define the exercises as integrated part of lectures.
zamfira@unitbv.ro
454 T. Klinger et al.
It is obvious, that all laboratory exercises for Pocket Labs have to be thoroughly
prepared and tested, before they are submitted to students. An important issue
is, whether the laboratory exercises with Pocket Labs are integrated into existing
courses or are stand-alone. Another possibility is, to define Pocket Lab exercises
as optional so that they may provide further and deeper understanding of the
topics of a corresponding lecture or in general.
The following examples show, how parallel and yet supporting use of lab concepts
and infrastructure can be performed. They are taken from different lectures, are
integrated in the respective curricula at different levels, and use even different
lab hard- and software, so that the general application of the concept is shown.
The course Electrical Engineering is part of the curriculum for first semester
students. In this course, students learn the first concepts necessary to analyze
electrical circuits, such as Ohm’s law, methods of network analysis (KVL/KCL),
operational amplifiers and the basics of RC and RL circuits switching. Labora-
tory work plays a major role for students to learn these concepts, therefore this
course was chosen as one of the pilot courses for a parallel use of Pocket and
Online Labs.
With this example we aim to show how Online and Pocket Labs can be
used in a complementary basis. Each one of them is more adequate for some
particular situation. For example, Online labs are very well suited for performing
measurements with circuits for which the internals should be kept hidden from
the students for didactic purposes. These circuits can be represented by black
boxes. Of course, implementing a black box for a Pocket Lab is possible, but it is
also a highly inefficient approach, since this very specific hardware setup would
need to be replicated every semester, for each student of the course. With an
zamfira@unitbv.ro
Parallel Use of Remote Labs and Pocket Labs in Engineering Education 455
online lab, since the same experiment setup is shared among all students, the
creation of the experiment and its maintenance are much simpler from the point
of view of the lecturer.
An example of a lab exercise delivered to the students of this course consists
in finding the Thévenin and Norton equivalents at terminals A and B of an
unknown network represented by a black box (Fig. 2), which, in our case, contains
a simple voltage divider. Since students do not know the internals of the black
box, the Thévenin and Norton equivalents can only be obtained experimentally
by measuring the open-circuit voltage on terminals A and B and the short-circuit
current that flows through A and B.
zamfira@unitbv.ro
456 T. Klinger et al.
Students can edit their program source code in the browser and they can run
their programs interactively. The teacher has the ability to search for similarities
between the files and can finally grade the results.
Analog computing is a technology that was widely used in the second half of
the last century to perform complex calculations which were not able to solve
with digital computer of that age in a reasonable time and with reasonable
effort. Although the idea of analog computing is itself very old (think about the
Antikythera mechanisms dated about 100 B.C.), it is nowadays outperformed
by digital computers; nevertheless it is a good basis for students to understand
the principles of calculation circuits [5].
It is not possible to provide every student or a number of laboratory places
with even a small analog computer; the development effort as well as the costs
would be too high. Therefore, the topic is ideal for a Remote Lab. Additionally,
students can learn the basics of calculation circuits, such as summing and differ-
ence amplifier, and others. With this understanding they are able to understand
also the function of simple analog computers, which may be provided as Remote
Lab by the University.
zamfira@unitbv.ro
Parallel Use of Remote Labs and Pocket Labs in Engineering Education 457
Fig. 4. User interface of an auto-focus remote lab in image processing lectures [6]
5 Conclusion
It could be shown, that the parallel use of Remote and Pocket Labs adds value
to engineering education, if some didactic aspects are considered. As students
are more or less left alone by accomplishing them, especially guiding mechanisms
and supporting materials have to be thoroughly considered.
References
1. Harward, V.J., et al.: The iLab shared architecture: a web services infrastructure
to build communities of internet accessible laboratories. Proc. IEEE 96, 931–950
(2008). doi:10.1109/JPROC.2008.921607
zamfira@unitbv.ro
458 T. Klinger et al.
2. Klinger, T., Madritsch, C.: Use of virtual and pocket labs in education. In: REV
2016, Madrid, Spain (2016)
3. Klinger, T., Madritsch, C.: Collaborative learning using pocket labs. In: IMCL 2015,
Thessaloniki, Greece (2015)
4. Gustavsson, I., et al.: A flexible electronics laboratory with local and remote work-
benches in a grid. Int. J. Online Eng. (iJOE) (2008)
5. Ulmann, B.: Analog Computing. Oldenbourg, München (2013)
6. Klinger, T.: Image Processing with LabVIEW and IMAQ Vision. Prentice Hall
PTR, Upper Saddle River (2003)
zamfira@unitbv.ro
The Effectiveness of Online-Laboratories
for Understanding Physics
Abstract. For the class “Experimental Physics for Engineers” (1500 students
each winter term) at the University of Stuttgart online laboratories (80% “virtual
labs”/simulations and 20% remote experiments) are optional learning resources.
In a new long-term investigative set-up the learning effects of online laboratories
as well as the other learning resources are to be detected.
1 Initial Situation
zamfira@unitbv.ro
460 D. Boehringer and J. Vanvinkenroye
goals are described in this phase and a small pre-test evaluates the knowledge of the
students before they run the exercises with online-experiments. Since 2010/11 the
orientation phase includes a short interesting and entertaining movie the aim of
which is to prepare students for the upcoming exercise and to pique their curiosity.
2. The execution phase: This is the main phase of the exercise. Here, the given task
should be mastered by the students using the online-experiment. In 2010/11 some‐
times more than one experiment per phase was included to offer more variety and
make the exercise more interesting.
3. The review phase: In this phase the progress of the students will be checked. This
phase is also implemented as a small test. From 2010/11 onward the questions of
the review phase are similar to questions of the exam and one of the questions of
one of the online-experiment’s review phase is actually included in the exam.
The preparation of the learning resources takes quite some time and effort. Hence
the lecturer asked himself “is it worth the effort?”, and contacted the eLearning depart‐
ment of the computer center not only for technical support (as in the beginning), but also
for the evaluation of the effectiveness of the respective learning resources for under‐
standing physics.
Online experiments were introduced in this class in the course of the LiLa (Library of
Labs) project in 2009 and closely accompanied and analyzed by the project team [4].
In 2009/10 26.7% of the students who took the exam performed at least one of the
three online-experiments (the fourth experiment was an exception for being a rather
difficult open ended question posed as a competition). In 2010/11 47.3% of the students
who took the exam participated. The total numbers of students (not only those who took
the exam) are shown in Tables 1 and 2.
In both years it could be observed that the more online-experiments the students
performed the more likely it was they would pass the exam (see Figs. 1 and 2). Most
students used the online-experiments in the course of their exam preparation during the
weeks before the exam and not as regular learning resource during the lecture period.
zamfira@unitbv.ro
Effectiveness of Online-Laboratories for Understanding Physics 461
Fig. 1. Percentages of students who passed the exam in 2009/10 respective the number of online-
experiments they performed; overall 68.5% of the students passed the exam.
In 2010/11 a much lower percentage of students passed exam compared to the year
before (47.8% compared to 68.5%). The main reason for this is that the kind of questions
asked and the structure of the exam had changed profoundly. Nevertheless, when asked
about the value of the online-experiments as preparation for the exam, 70.3% considered
them as helpful in 2010/11 compared to only 31.8% in 2009/10. Apparently the changes
made in each of the three phases of an online-experiment were successful.
Figures 1 and 2 cannot tell us about the effectiveness of online laboratories for the
learning outcome. The data seem to suggest that the more interested the students are,
the more they learn and the better their exam performance is.
As additional indicators students of the 2009/10 class were asked about their perceived
learning success because of online experiments (Fig. 4) and the effect online experiments
have on their motivation to deal with the topics of the lecture more intensively (Fig. 3).
zamfira@unitbv.ro
462 D. Boehringer and J. Vanvinkenroye
Fig. 2. Percentages of students who passed the exam in 2010/11 respective the number of online-
experiments they performed; overall 47.8% of the students passed the exam.
Fig. 3. Percentages of student answers on a bipolar scale with six units concerning the effect of
online experiments on the students’ motivation
zamfira@unitbv.ro
Effectiveness of Online-Laboratories for Understanding Physics 463
Online experiments seem to have a positive effect on the students’ motivation. And
according to most students they also have a positive effect on the learning success.
Fig. 4. Percentages of student answers on a bipolar scale with six units concerning the perceived
learning success because of online experiments
We made no further investigations about this topic after 2011. It was only in late
summer 2016 that we resumed our research. This time we want to install a generic set-
up for our investigations that is supposed to be in place for several years. It will be
discussed in the next section.
Studies that examine the effects of media-usage [for online experiments most important 6;
also see 5] for learning and student performance typically apply one of the following
investigative set-ups:
• Study of one group only: studies of the usage of media, their acceptance and the
influence they have on the learners’ motivation often concentrate on one large group
of students. Sometimes this large group is separated into smaller groups of students
with different media usage preference and the performance of the respective students
in exams.
• Study of two and more groups: if effects of certain media are to be detected, especially
concerning student performance in exams, often one group for which these media
are offered is compared with a control group for which they aren’t. In some cases
zamfira@unitbv.ro
464 D. Boehringer and J. Vanvinkenroye
different media are offered to different groups or the same media are offered and it is
examined whether the effects are the same. In rare cases these studies are performed
in some combination in consecutive years.
More sophisticated studies take into account other factors such as gender, GPA, class
attendance etc. and try to extract the actual effects of the different media. These studies
are rather rare since the necessary data are not easy or impossible to collect [e.g. 7].
zamfira@unitbv.ro
Effectiveness of Online-Laboratories for Understanding Physics 465
paint a more colorful picture of the students’ learning process, their decisions which
learning resources to use, and their respective learning outcome. First results will be
presented at the conference.
References
1. Tetour, Y., Boehringer, D., Richter, T.: Integration of virtual and remote experiments into
undergraduate engineering courses. In: 2011 First Global Online Laboratory Consortium
Remote Laboratories Workshop (GOLC), Rapid City, SD, October 2011, pp. 1–6. IEEE (2011)
2. Richter, T., Tetour, Y., Boehringer, D.: Simulations in undergraduate electrodynamics: virtual
laboratory experiments on the wave equation and their deployment. In: 2010 IEEE, Education
Engineering (EDUCON), Madrid, Spain, April 2010. IEEE, pp. 1091–1097 (2010)
3. Richter, T., Tetour, Y., Boehringer, D.: Library of labs. A European project on the
dissemination of remote experiments and virtual laboratories. In: SEFI Annual Conference
2011, Lisbon, Portugal, 27–30 September 2011
4. Richter, T., Tetour, Y., Boehringer, D.: Library of labs: a European project on the dissemination
of remote experiments and virtual laboratories. In: Werner, B. (ed.) International Symposium
on Multimedia (ISM 2011), Dana Point, California, USA, December 2011, pp. 543–548. IEEE
(2011)
5. Lindsay, E.D., Good, M.C.: Effects of laboratory access modes upon learning outcomes. IEEE
Trans. Educ. 48(4), 619–631 (2005)
6. Lindsay, E.D.: The impact of remote and virtual access to hardware upon the learning outcomes
of undergraduate engineering laboratory classes. Ph.D. University of Melbourne (2005)
7. Traphagan, T., Kucsera, J.V., Kishi, K.: Impact of class lecture webcasting on attendance and
learning. Educ. Tech. Res. Dev. 58, 19–37 (2010). doi:10.1007/s11423-009-9128-7
zamfira@unitbv.ro
Remote Control and Measurement
Technologies
zamfira@unitbv.ro
On the Fully Automation of the Vibrating
String Experiment
1 Introduction
– Virtual Labs provide computer based simulations which offer similar views
and ways of work to their traditional counterparts. Nowadays, simulations
have evolved into interactive graphical user interfaces where students can
manipulate the experiment parameters and explore its evolution.
c Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6 44
zamfira@unitbv.ro
470 J. Tajuelo et al.
– Remote Labs use real plants and physical devices which are teleoperated in
real time. Remote experimentation through the Internet has been available
for more than a decade and its interest and use has been growing over the
years [10,16,18].
Past studies have shown that online and hands-on labs are equally effec-
tive in terms of learning outcomes [2]. Moreover, online labs provide additional
advantages [15], such as that lab sessions can be watched by many people and
recorded or that online labs can be used in 24/7 from anywhere and can be
accessed by handicapped people.
Given the complementary uses of the previous experimentation approaches,
it is probably best if an experiment itself is offered in several ways. Here, we
present a lab implementation of a vibrating strings system that consist in both
forms: the virtual or simulated one and the real, remote one. For those readers
that might be interested in these resources, the virtual and remote laboratories
can be found in UNILabs, a network of interactive online laboratories. For those
readers interested in replicating the system or learning how to build a similar
one, the main instructions and tips to do so are given in this paper.
This work uses a free and open source software called Easy java/javascript
Simulations (EjsS) that eases the creation of Javascript applications to build
online lab interfaces. Since its appearance, more than a decade ago, EjsS has been
growing and nowadays it can also be used to easily create remote laboratories. It
has been massively used to create physics simulations: there are more than three
hundred at the ComPADRE-OSP digital library [7], as well as many virtual and
remote labs in the automatic control field (for example, those at the UNILabs
network [5,8]). While all these applications were based on Java and deployed as
Java applets, EjsS now offers the possibility to build Javascript simulations. In
this regard, there are now plenty of Javascript simulations created with EjsS:
again, ComPADRE-OSP offers a couple of hundred of them. However, to the
best of our knowledge, EjsS has only been used for building a Javascript remote
experiment in the present work and in [1].
All applications created with EjsS can be embedded into Moodle, the most
widely used free and open source Learning Management System (LMS), with
just a few clicks. For this, a plugin called EJSApp [9] is used, which allows the
one-click deployment of VRLs into Moodle. Once installed in a Moodle server,
EJSApp allows teacher users to add a new kind of activity called “EJSApp”
which, in turn, allows uploading .jar and .zip files previously generated with EjsS.
When the file selected to create the activity is a .jar file, then a VRL deployed as
a Java applet is added to Moodle. When the file selected is a .zip file generated
by EjsS, the VRL deployed into Moodle is in Javascript format. With EJSApp
not only the applications get embedded in the LMS but they also gain some
additional features automatically, such as: connection with a booking system that
may be used for controlling the access to the remote experiment, multilanguage
support, saving data and image files from the virtual or remote experiment
application to the users’ files repository in the LMS, grading, monitoring the
time spent by users working with the experiment and backup and restore options.
zamfira@unitbv.ro
On the Fully Automation of the Vibrating String Experiment 471
The virtual and remote labs developed in this work have been integrated in
UNILabs, a portal based in Moodle, using this solution.
Fig. 1. In an infinitesimal portion of the string dy appear two tensions, one at each
end of the portion, so that under a small displacement assumption the horizontal net
tension is null.
Fy = 0, (3)
Fz = T dα. (4)
T dα = dma (5)
2
∂ z
= (μdy) . (6)
∂t2
zamfira@unitbv.ro
472 J. Tajuelo et al.
∂2z
dα = dy, (7)
∂y 2
so as the equation that describes the wave motion is
∂2z μ ∂2z
= , (8)
∂y 2 T ∂t2
where
nπ T
ωn = . (10)
L μ
zamfira@unitbv.ro
On the Fully Automation of the Vibrating String Experiment 473
3
n
Fig. 2. First four normal modes of vibration for a fixed-fixed string. Each n mode has
(n + 1) points where the oscillation is zero. The wave length of mode n is λn = 2L/n.
where r is the radius of the string and ρ its volumetric density. If we select a
control parameter (as could it be the tension, for example), by measuring the
frequency f of the string as a function of this parameter we may establish a
relationship of the kind
f = αT β , (14)
so as for different tensions we may perform a least squares method to obtain the
constants α and β and, therefore, find the density of a string just knowing the
length of the string and its radius, i.e.,
1
ρ= . (15)
4π(Lrα)2
The key concept of this experimental set-up is the fact that there exists a depen-
dence between many parameters: a linear dependence between the period and
the length of the string, an inverse dependence between the frequency and the
radius, a square root dependence between the tension and the frequency, and so
on. With these many different cases, a student may explore with many parame-
ters, using linear-linear fits, linear-log fits, etc. so as to obtain physical quantities
measuring how the frequency depends on them.
3 Experimental Device
A schematics of the device is shown in Fig. 3. There are five different strings
(2) made of different materials (copper, kanthal, constantan, and nickel) and
with diameters ranging from 0.3 mm to 0.5 mm. One of the ends of the strings
(with the exception of the central string) is fixed on the aluminum structure of
zamfira@unitbv.ro
474 J. Tajuelo et al.
the device, while the other end of the strings is connected to a dynamometer
(10) that measures the tension along the string. In the case of the central string,
the fixed end is connected to the axis of a stepper motor (3), so that the tension
can be controlled. A LEGO carrier (7) is used to displace an aluminum rod in
close contact with the strings (8) along the y axis, in such a way that the length
of the vibrating part of the strings can be changed from 380 mm to 550 mm. This
length is measured by means of a rule and an indicator attached on the mobile
rod (9). A DC-LED (Galaxy 1000) light source (1) illuminates the system from
above, and a linear stage (RS 340-3749) (6) is setup below the strings along the x
axis. Two elements are attached on the top of this linear stage: (i) a light sensor
(Phywe 08734-00) covered by an opaque cap with a 0.3 mm slit oriented along
the y axis (5), and (ii) a LEGO gear connected to a LEGO servo motor (4).
As can be seen in the close view of Fig. 3, the rotation axis of the LEGO gear
(11) does not coincide with its center, so that the perimeter of the gear roughly
describes an ellipse when the LEGO servo motor rotates (12). The position of
the opaque cap and the gear along the vertical direction has been fine-tuned
in such a way that, first, the cap of the light sensor is placed less than two
zamfira@unitbv.ro
On the Fully Automation of the Vibrating String Experiment 475
millimeters below the horizontal plane formed by the strings, and second, the
apex of the gear perimeter trajectory coincides with the horizontal plane formed
by the strings (13). The stepper motor and the linear stage are controlled by
two identical drivers (EasyDriver), and An Arduino I/O boardcard is used to
send the convenient digital signals. A power supply (Lendher 3003D) provides
the current required by both drivers, and a second identical power supply is used
for the DC-LED light source. An oscilloscope (PicoScope 2203) is used to read
the measurement from the light sensor. A LabVIEW code has been developed
to control all of the above mentioned elements.
zamfira@unitbv.ro
476 J. Tajuelo et al.
package it in order to run it in a standalone mode (in the case of Java) or inside
a web page to be run in a web browser (Java and Javascript), [1,6,8,11,14,17,19].
Figure 5 shows the Javascript editor. The top part in the editor contains
description, model and view tabs. The editor allows to build a simulation or
remote laboratory by adding the mathematical behavior and a graphical inter-
face. Then, the main application is divided in two parts:
– The model. Using this tab in the editor, a developer can define differential
equations, write some custom code and/or make connections to other software
or hardware. The complexity of the simulation and model depends only on
the implemented system, the requirements and the knowledge about it.
– The view provides to the users a GUI that determines the interaction and
visualization capabilities of the application. This view can be built using the
editor by adding single view elements from the right panel of the EjsS editor
(right side of Fig. 5).
zamfira@unitbv.ro
On the Fully Automation of the Vibrating String Experiment 477
The vibrating string experiment described in this work can be simulated using
EjsS, introducing Eq. 12 to generate the data and by creating an interactive GUI.
This GUI consists of buttons, sliders, numerical fields, check boxes, graphs, and
two and three dimensional graphical elements that allow one to change and
visualize parameters of the lab.
Figure 6 shows the basic structure of the virtual laboratory when the .xhtml
is served to the client. As it was said in previous sections, the virtual laboratory
is based on a simulation of the system behavior and the GUI. The virtual lab also
allows students to get familiar with the available interaction and the protocol, in
order to be prepared to the remote version of the lab. The application window is
divided into three sections: a 2D and 3D graphical visual representation of the
system, the controls panel and the plots/graph panel.
zamfira@unitbv.ro
478 J. Tajuelo et al.
– Plots/Graphs panel: This part of the interface, at the bottom part of Fig. 6,
allows the user to see data in different plots and graphs. The vibrating string
laboratory plots the light intensity versus, first, the position of the linear
stage (see Sect. 5 for details in this calibration procedure), and second, time
(in order to obtain the frequency of the fundamental normal mode).
5 Experimental Protocol
Three CCD cameras allow the student for the visualization of a general view
of the experimental setup as well as a close view of the measurement elements
(dynamometers and length indicator). Once the student connects to the remote
controller, an automated initialization procedure is executed by the device: The
DC-LED is turned on, and the linear stage and the LEGO carrier are displaced
to their initial positions, determined by means of two LEGO limit switches. After
this initialization, the student has to proceed as follows:
1. Location of the strings’ positions: A complete sweep is performed by
the linear stage along its whole range of displacement, while the light sensor
attached on its top is continuously measuring the light intensity. Therefore,
the student can plot the function light intensity versus position along the x
axis. As can be seen in Fig. 8a, the light intensity is roughly symmetric with
a local maximum at the center. This is because we use a single light source
placed at the center of the device in order to avoid multiple shadows produced
by multiple light sources. Thus, the student can determine the position of each
string from the five local minima in the light intensity.
zamfira@unitbv.ro
On the Fully Automation of the Vibrating String Experiment 479
Fig. 8. (a) Light intensity versus position of the light sensor along the x axis. (b) Light
intensity versus time after the gear hits one of the strings. The inset graph represents
a close view of the results from t = 3.5 s to t = 3.55 s.
2. Selection of the string and the length to explore: After this calibration,
the linear stage is displaced to the string selected by the student, and the
LEGO carrier moves backward or forward to reach the desired length. Then,
the linear stage performs an automated fine-tuning to ensure that the thin
slit of the light sensor is placed exactly below the string shadow.
3. Selection of the tension (if needed): If the string selected by the student
is the central one, the tension can be varied by the stepper motor. In that
purpose, by clicking an increase tension (or decrease) control, the stepper
motor rotates a fixed number of steps in the clockwise (or counterclockwise)
direction, so that the tension is increased (or decreased) in steps of approx-
imately 0.05 N. The student observes the measurement of the dynamometer
by means of one of the CCD cameras, and is able to change the tension as
long as it is maintained below 10 N to avoid breakage of the string.
4. Execution of the experiment and data acquisition: When the previous
steps have been completed, the device is ready to execute the experiment in
those conditions selected by the student. Then, by clicking the corresponding
control, the light sensor starts to measure the light intensity, and the LEGO
gear attached on the top of the linear stage performs a 360◦ rotation, plucking
the string when it reaches its highest position. The Fig. 8b shows the results
of an actual experiment as an example. The instant in which the gear hits
the string and the relaxation dynamics are clearly observed. The student has
to analyze these data, calculating the frequency of the fundamental normal
mode, f1 .
5. Analysis of the results and comparison with theory: Once the student
has completed experiments under different physical conditions, the depen-
dence relation between f1 and the physical parameters of the string (T , L, ρ)
can be established. Then, the student should be able to discuss the experi-
mental errors and the validity of the theoretical model.
zamfira@unitbv.ro
480 J. Tajuelo et al.
6 Conclusion
Traditional hands-on laboratories are useful to achieve experimental skills like:
– Knowing the principles, techniques and instrumental measure devices to study
physical phenomena.
– Evaluate limitations of the measure process.
– Explain the effects of interference in measures, their consequences and how
to minimize the associated errors.
– Be able to calibrate measure instruments, take useful data collections and
perform a statistical analysis.
– Report correctly taken measures, and obtain relationships between physical
variables.
In distance learning education context, the implementation of new techniques
and methodologies that adapt these skills to the student is of the utmost impor-
tance. These techniques should allow both the comprehension of the physical
experiment and the evaluation of the expected abilities. Therefore, alternatives
to hands-on laboratories should be studied in detail.
In this work we propose an effective alternative that has been demonstrated
to be useful [2]. On one hand, we develop a virtual laboratory that allows the
students to practice, to be familiar with the techniques of measure, and to intro-
duce them to the concept of data adquisition. The virtual laboratory should
be built in such a way that it is close related to a real laboratory. In spite of
being in a virtual mode, it is important for the user to pay attention to how to
effectively measure physical quantities and how he or she has to collect correct
data. As a virtual laboratory, it lacks of proper physical (realistic) conditions,
and thereafter we need to offer a proper alternative to the hands-on lab.
For this purpose we also explain in this work how to build, in a easily afford-
able manner, a remote laboratory. This remote lab ensures that the data adqui-
sition is exactly the same a scientist would take if he or she was actually in a lab.
Through camera views, the experimentalist can control all the physical parame-
ters and can command the measure process. Raw data, extracted directly from
the experiment done in real time, will be used to analize results and to obtain
conclusions. We focus in this part on how to adapt a classical experiment to
be controlled with a computer. We use LEGO MindStormsTM construction kits,
stepper motors and an Arduino controller that are all connected in LabVIEW.
The control of LabVIEW is delegated in a JIL Server, which uses the XML-RPC
protocol to connect with a Javascript code written with Easy Java/Javascript
Simulations. This code is deployed in UNILabs, the webserver that allows the
user to connect to the lab with a web browser.
We choose the vibrating string as a prototypical example in the study of
vibrations and waves, which is one of the subjects that appears in the first
courses on Physics degree. The construction of an accurate hands-on laboratory
can be sensitive and, as such, we provide an alternative solution for those learning
center that cannot afford the physical lab. The selection of the vibrating string
was made with two ideas in mind. The first one is the physical representation
zamfira@unitbv.ro
On the Fully Automation of the Vibrating String Experiment 481
of the experiment. This experiment allows the student to know the dependence
between many parameters. We know that the fundamental frequency of the
standing wave depends on the density of the string, its length, and the tension.
The relationship between this parameters goes, respectively, as the inverse of the
square root, the inverse, and the square root. This fact allows the student to learn
about different representations (linear, inverse, logarithmic) in order to fit data
collection to a curve. The second idea is to present how cost-effective devices can
be used to build complex laboratories. With LEGOTM kits it is remarkably easy
to introduce experimentalists into robotic designs. The development of Arduino
boards also allows one to control stepper motors, light diodes, and so, in an
easy manner. These tools can be widely used to construct future laboratories,
compatible with the requirements of a given experimental setup and allowing the
students to measure, analyze and extract conclusion with laboratories developed
in the cloud.
Acknowledgments. Financial support from the Vice Chancellor for Academic Affairs
and Quality at UNED under grants GID2016-9-1 and GID2016-25-1 is acknowledged.
References
1. Bermudez-Ortega, J., Besada-Portas, E., Lopez-Orozco, J.A., Bonache-Seco, J.A.,
de la Cruz, J.M.: Remote web-based control laboratory for mobile devices based on
EJsS, Raspberry Pi and Node.js. In: 3rd IFAC Workshop on Internet Based Control
Education, Brescia, Italy, vol. 48, pp. 158–163. IFAC-PapersOnLine, November
2015
2. Brinson, J.R.: Learning outcome achievement in non-traditional (virtual and
remote) versus traditional (hands-on) laboratories: a review of the empirical
research. Comput. Educ. 87, 218–237 (2015)
3. Chacón, J., Vargas, H., Farias, G., Sánchez, J., Dormido, S.: EJS, JIL Server, and
LabVIEW: an architecture for rapid development of remote labs. IEEE Trans.
Learn. Technol. 8(4), 393–401 (2015). doi:10.1109/TLT.2015.2389245
4. Chang, G.-W., Yeh, Z.-M., Chang, H.-M., Pan, S.-Y.: Teaching photonics labo-
ratory using remote-control web technologies. IEEE Trans. Educ. 48(4), 642–651
(2005)
5. Chaos, D., Chacon, J., Lopez-Orozco, J.A., Dormido, S.: Virtual and remote
robotic laboratory using EJS, MATLAB and LabVIEW. Sensors 13(2), 2595–2612
(2013)
6. Christian, W., Esquembre, F.: Modeling physics with easy java simulations. Phys.
Teach. 45, 475–480 (2007)
7. Christian, W., Esquembre, F., Barbato, L.: Open source physics. Science
334(6059), 1077–1078 (2011)
8. de la Torre, L., Guinaldo, M., Heradio, R., Dormido, S.: The ball and beam system:
a case study of virtual and remote lab enhancement with Moodle. IEEE Trans.
Ind. Inform. 11(4), 934–945 (2015)
9. de la Torre, L., Heradio, R., Jara, C., Sanchez Moreno, J., Dormido, S., Torres, F.,
Candelas, F.: Providing collaborative support to virtual and remote laboratories.
IEEE Trans. Learn. Technol. 6, 312–323 (2013)
zamfira@unitbv.ro
482 J. Tajuelo et al.
10. de la Torre, L., Sanchez, J.P., Dormido, S.: What remote labs can do for you. Phys.
Today 69, 48–53 (2016)
11. de la Torre, L., Sanchez, J.P., Heradio, R., Carreras, C., Yuste, M., Sanchez, J.,
Dormido, S.: UNEDLabs - an example of EJS labs integration into Moodle. In:
World Conference on Physics Education (2012)
12. Farias, G., Keyser, R.D., Dormido, S., Esquembre, F.: Developing networked con-
trol labs a MATLAB and easy java simulations approach. IEEE Trans. Industr.
Electron. 57, 3266–3275 (2010)
13. French, A.P.: Vibrations and Waves. CRC Press, Boca Raton (1971)
14. Galan, D., Heradio, R., de la Torre, L., Dormido, S., Esquembre, F.: Automated
experiments on EjsS laboratories. In: International Conference on Remote Engi-
neering and Virtual Instrumentation, Madrid, Spain, pp. 78–85, February 2016
15. Gravier, C., Fayolle, J., Bayard, B., Ates, M., Lardon, J.: State of the art about
remote laboratories paradigms - foundations of ongoing mutations. Int. J. Online
Eng. 4(1), 19–25 (2008)
16. Heradio, R., de la Torre, L., Galan, D., Cabrerizo, F.J., Herrera-Viedma, E.,
Dormido, S.: Virtual and remote labs in education: a bibliometric analysis. Com-
put. Educ. 98, 14–38 (2016)
17. Heradio, R., de la Torre, L., Sanchez, J., Dormido, S.: Making EJS applications
at the OSP digital library available from Moodle. In: International Conference on
Remote Engineering and Virtual Instrumentation, Porto, Portugal, pp. 112–116,
February 2014
18. Heradio, R., de la Torre, L., Dormido, S.: Virtual and remote labs in control edu-
cation: a survey. Annu. Rev. Control 42, 1–10 (2016)
19. Pastor, R., Sanchez, J., Dormido, S.: Web-based virtual lab and remote experi-
mentation using easy java simulations. In: Proceedings of the 16th IFAC World
Congress (2005)
20. Wannous, M., Nakano, H.: NVLab, a networking virtual web-based laboratory that
implements virtualization and virtual network computing tech. IEEE Trans. Learn.
Technol. 3(2), 129–138 (2010)
zamfira@unitbv.ro
Identifying Partial Subroutines for Instrument
Control Based on Regular Expressions
1 Introduction
Networked Control Systems (NCS) are systems that are operated over a network (Jinhui
et al. 2013). Increasing use of smart objects in Internet of Things (IoT) (Whitmore et al.
2015) applications and their complex architectures have necessitated semi-autonomous
and autonomous capabilities in these applications. IoT applications are based on the
internet which requires exchanging commands in discrete packets or frames over the
Internet. As such, these nodes operate with a finite set of basic commands. There can be
many complex commands based on these basic commands that may not be explicitly
described in the system, but which are specific to a given application. Such an IoT system
contain two types of nodes: masters and slaves. Master nodes control slave nodes. They
also have the responsibility of collecting data and have higher decision making capa‐
bilities. There can be many such master-slave combinations in the IoT system. A super‐
visory system is required to monitor and validate the progress of the IoT system and
steer the overall IoT system.
zamfira@unitbv.ro
484 A. Maiti et al.
Remote Access Laboratories (RALs) can be seen as IoT systems, which allow
students to access and control scientific experimental setups educational purposes
(Harward et al. 2008; Mejías et al. 2017). The Experiment sites are the slaves containing
the experimental rig and the students’ site contain the Controller Interface (CI) is the
master which sends commands to the experiment and collects data from it. RAL is
somewhat different from regular IoT as it is more dependent on human inputs, but it still
implements all the characteristics of an IoT system. A Peer-to-Peer (P2P) RAL system
extends the operational aspects of a traditional laboratories to allow individuals from
their homes and school to create and share experiment as part of their curriculum. This
means that the creators or the users of the experiments can only have a very basic fixed
set of commands. But more complex commands composed of these basic commands
can be used in the experiment to improve performance, but the experiment creators may
not be able to identify or construct/implement them. A P2P RAL is shown in Fig. 1
where it, if the human users are removed, becomes a general IoT system with multiple
sets of master and slave nodes.
This paper presents a method to create a list of regular expressions from the inter‐
action between the master and slave nodes. Such regular expressions (based on automata
theory) represent the way the experiments are used and can be used to create subroutines
and save them automatically without anyone specifically creating them. The regular
expressions can be used for other applications as well. Although the focus here is on
RAL, the proposed methodology can be adapted to any IoT system that operates with
known set of finite commands. It provides a computational solution to a control problem.
The remainder of the paper is organized as follows: Sect. 2 discusses the corre‐
sponding related work in algorithmic information theory, Automata theory and RAL.
Section 3 and 4 shows the problem addressed here and method to obtain the regular
expressions. Section 5 and 6 presents the method to convert the regular expressions to
the subroutines (or algorithms/programs) and an example.
2 Related Works
This section discusses the related work in remote laboratory and algorithmic information
theory fields.
zamfira@unitbv.ro
Identifying Partial Subroutines for Instrument Control 485
RAL are a class of NCS that is used to control equipment over the internet. The particular
type of RAL concerned in the current context is a Peer-to-Peer setup (Maiti et al. 2015a,
b). Such a system has a Controller Interface (CI) which is the master node that takes inputs
from the human operators, processes them according to a program and sends corre‐
sponding to the controller unit of the experimental setup in remote location and receives
any feedback. The Controller Unit (CU) is the slave nodes or devices in the RAL system
and it can be seen as a DFA as it changes states depending upon the command executed on
it. Each state is discrete and based on a particular event. The language between CI and CU
is the communication protocol for the instrument control in the P2P RAL. This language
consists of the very basic (or atomic) components of instrumentation-
• read (r) - reading the value of a port (sensors and actuators)
• write (w) - writing a value to a port (for actuators)
• wait (a) - pause the CU for maintain synchronization
There can be other composite commands based on these. The experiment creators
in these RAL systems are not capable of creating such advanced programs for their rigs.
They are capable of creating very basic commands and the learning experience to create
higher level programs is part of the educational aim of the RAL system.
zamfira@unitbv.ro
486 A. Maiti et al.
The aim of this paper is to establish a method based on control theory and to aid the
makers of the experiments in creating complex programs. The main contributions
include a method to create regular expressions that can describe the language between
a particular CI-CU pair. These regular expressions are then used to create complex
programs.
In terms of a general IoT, this problem can be applicable where a slave device inter‐
acts with a master device for the first time. It may have interacted with other master
devices previously and thus is capable of “knowing” what to expect from the new inter‐
action. These master nodes may work without human interventions and generate random
commands themselves. Thus, the slave node can keep track of what the master node is
requesting for and determine whether it is suitable for the overall goal of the IoT and
the node itself. This can be done if the commands exchanged previously are used to
determine a set of regular expressions that can define the language between the nodes.
The new interaction can be based on this set of predefined regular expressions. This
approach of collecting information from regular use of the devices automatically is part
of reinforcement learning approaches to implement intelligent devices. With time
devices that are capable of learning by analyzing past data (interactions or commands
in this case), can gain high reliability in their operations (Garcia and Fernande 2015;
Schmidhuber 2015).
This RAL application is similar to the problem of simultaneous localization and
mapping (SLAM) (Jaulin 2011) used in autonomous cars in that are large number of
sensors and actuators are involved. However, the experiments are not same are
unmanned vehicles or robots. They are run by human and are dependent on the human
inputs throughout the entire operation.
This section describes the RAL control strategies in terms of algorithmic information
theory.
The states in the CU are changed according to the commands from the CI. Each experi‐
ment has a finite set of actuators and sensors connected to the CU ports (R). At any given
time, the rig can have a discrete value on the ports. Thus, any command (C) executed
results in the change of state space of the rig Y i.e. changes the rig from one discrete
state to another i.e.
where A and B are constant matrices for an experiment rig. In terms of simple control,
only the actuators have any impact on state transition as these are the only components
that can directly change the configuration of the experimental rig. From a decision and
control point of view, the state space contains the values of the all active ports on the
MCU for both sensors and actuators. Technically, the state space of the rig may be
zamfira@unitbv.ro
Identifying Partial Subroutines for Instrument Control 487
where, N is a set of composite commands which may be already known for the experi‐
{ }
ment, i.e. N = (𝕣 | 𝕨 | 𝕒)∗ and |N| ≥ 0.
The aim is to create a ‘program’ or set of instructions that can generate this sequence
of symbols. If only repeating sequences of commands are used, then the corresponding
programs can be represented as a static sequence of statements generating the
commands/symbols. While it is easy to create a program that only generates static repe‐
tition of symbols a more complex technology is required to create functions or subrou‐
tines that involves variable inputs i.e. the values of v in Eqs. 2 and 3.
The method to obtain the algorithms is depicted in Fig. 2. It is based on a Deter‐
ministic Finite state Automata (DFA) and its corresponding regular language/regular
expression in this work that is able to identify conditional statements and iterations.
More powerful automata may be used to implement more complex functions. A parti‐
tioning algorithm is used to obtain clusters of commands that are issued within definite
time periods from training data sets. These clusters are then converted to its minimal
regular expression. Once the DFA is created, it is minimized to remove any redundancy.
Then the DFA is converted to a ‘program’ or ‘algorithm’ which is capable of processing
input parameters to generate variable output sequences, but according to the constraints
defined within itself.
zamfira@unitbv.ro
488 A. Maiti et al.
This section presents the method to create the regular expression list and their potential
applications.
A symbol in the regular expressions is considered as a command along with its port(s).
For, e.g. 𝕨p1 represents a write command on port p1. For parallel executions, multiple
ports can be specified e.g. 𝕨p1p2 represents a write command on p1 and p2. The symbols
( )
𝕨pv can be parameter restrictive where a symbol is unique for a port (p) and value (v).
For sake of simplicity, the following sections make mention of port restrictions only,
but the procedures can be done with parameter restrictions as well.
For finding closely related command sequences, the parameters that are passed in
the write commands or whether it is a read or write command are not relevant. The only
things that matter are the repeating sequence of commands. A subsequence of commands
can be regarded as an element of a regular language that is accepted by the CU automaton
(Maiti 2015a, b) ensuring that the rig (or CU) is always in a stable state. For every such
subsequence, there can be a regular expression obtained from it. This can be done by
using the Hopcroft algorithm for minimization (Garcia 2013) and Kleene’s algorithm
(Gross and Yellen 2004) for creating regular expression from a DFA. The DFA may be
constructed (as shown in Fig. 3) with respect to a particular subsequence(s) to a particular
CU (e) as
{ }
Yse = Q, 𝛴, 𝛿, q1 , F
where,
• Q contains (r + 2) states. There is r non-final states (q1 … qr) corresponding to every
read/write command in the sequence in order, a single non final fail state qf, and a
single final state that is appended at the end (qr+1).
• Σ = set of unique commands from input sequence
• q1 is the first state corresponding to the first command from Eq. 2.
{ ( ) ( ) }
• 𝛿 = δ qi , c → qi + 1 , δ qi , c′ → qf for i ≥ 0, c is the command that is executed
between qi and qi+1 and c′ represents any command symbol except c.
• F contains only one state which is qr+1.
zamfira@unitbv.ro
Identifying Partial Subroutines for Instrument Control 489
For creating regular expressions from command sequences the notations of the
regular expressions must be improved to include the time information. The time infor‐
mation is neglected while creating the minimal regular expression for a sequence. Once
a regular expression is obtained, the time information is embedded. To achieve this a
time value ρi is associated with every command in the command sequence as discussed
in Eq. 2 where ρi = ti+1 − ti. Time gaps can be decisive when it is known to be constant
value and the regular expression contains the constant value in place or they can be
indecisive when it is represented only as a variable. Thus, a regular expression
( )+
𝕣p1 7𝕨p2 5𝕣p3 𝜌3 𝕨p3 𝜌4 𝕨p4 3𝕨p5 𝜌5
means that the time gap between 𝕣p1 and 𝕨p2 is represented as constant 7, between 𝕣p3 and
𝕨p2 is always a constant 5 and similarly the number of times 𝕨p4 and 𝕨p5 is repeated is
always with a time gap of 3. However, the time gap found between 𝕨p3 and 𝕨p4 is variable
and thus marked with symbol ρ4. For any practical application, there can be a tolerance
value for which the time between 𝕨p4 and 𝕨p5 can vary but still be regarded as constant
3 in the original subsequence.
zamfira@unitbv.ro
490 A. Maiti et al.
<Subsequence><regular expression>
where the regular expression is obtained as discussed previously. In the second phase,
a new list LX is created as a copy of LG along with a new attribute of number of appear‐
ances that represents the number of times the regular expression has appeared in LG. An
element in LX is as
where initially the number of appearances for each regular expression is set to 1.
There can be two broad types of similarity that could be found within LG:
1. Static time-bound: Any two regular expressions can be completely identical. These
are basic composite commands that maintain the time difference between commands
sequences that never change i.e. for two sequences s1 and s2
𝜌s1
i
= 𝜌s2
i
for i > 0 (4)
Ideally, these types of commands will be found at a very low level containing only
a few atomic commands. They do not involve iterations or conditional checking. These
commands are easy to identify and can be automatically stored without further
processing with human interventions.
2. Dynamic time-bound: The regular expressions are identical in this case if the
commands are the same and in the same order, but the time information is different.
If the time gaps are perfectly divisible or multiples of the corresponding subse‐
quence. The time gaps can be regarded as variable wait (𝕒) commands being executed
between the subsequences. This repetition represents a variable that is collected as
an input to the corresponding command. The relation between the corresponding
time gap in the sequence can be linear or non-linear. This type of subsequence
requires further processing that can be solved with further analysis of the data or
involve human interventions depending upon the context of usage. These types of
commands contain conditional checks and iterations. Such statements can be
zamfira@unitbv.ro
Identifying Partial Subroutines for Instrument Control 491
contrived based on the operators in the regular expression as described in the next
section.
If any two elements (e1, e2) ∈ LX are found to be identical with static time-bound
then e2 is removed from the list and the number of appearance for e1 is increased by 1.
Since e1, and e2 are exactly the same in this case, there is no need to update time infor‐
mation in e1’s regular expression.
If any two elements (e1, e2) ∈ LX are found to be identical with dynamic time-bounds
i.e. they differ only with regards to time information, then also e2 is removed and e1’s
number of appearance is increased by 1. But in this case the time information needs to
be updated and any previous constant time information needs to be replaced with a
variable symbol. For example, if there are two elements with regular expressions
( )∗
𝕣p1 5𝕨p2 6𝕣p3 7 𝕨p4 9wp5 and
( )∗
𝕣p1 5𝕨p2 6𝕣p3 7 𝕨p4 15𝕨p5
with number of appearances of 1 each, then they are identical, except the time infor‐
mation between 𝕨p4 and 𝕨p5 is different. Thus, the new regular expression representing
both would be
( )∗
𝕣p1 5𝕨p2 6𝕣p3 7 𝕨p4 𝜌1 wp5
This method to create the regular expressions can be used in many applications in an
IoT system such as:
• Validations: With increasing reliance on devices to deliver critical services, it is very
important to ensure safety and integrity of intelligent devices in an IoT system. This
can be done through validation of what is executed on the device. Validation aims
to verify whether executing a command will lead to an unstable state. This can be
zamfira@unitbv.ro
492 A. Maiti et al.
done by strictly enforcing the regular expressions and not allowing anything that does
not match known regular expressions.
• Evaluation: In some applications such as in RAL, these regular expressions can be
used to evaluate the control of the devices. This can be done by following a relaxed
matching of any two sequences where, the difference between them can determine
how much the student has deviated from an ideal sequence of control.
• Variable control interface: The control interface and the available controls may be
altered in real time depending upon the current network conditions. Such a scheme
would choose different command sizes to ensure that a desired level of interactivity
is maintained while ensuring that proper time difference between successive
commands is maintained.
Apart from these two there can be other applications such as finding patterns in the
control and suggesting the next set of commands. The application addressed in this paper
is with regard to automatically identifying potential functions, at least partially, that can
be stored corresponding to a particular experiment in a RAL system.
This section describes the method to convert the regular expressions into a subroutine
of a program. It requires analyzing some advanced aspects of the regular expressions.
5.1 Iterations
zamfira@unitbv.ro
Identifying Partial Subroutines for Instrument Control 493
In this example a 𝕣p1 5𝕨p2 6𝕣p3 is followed by either a 𝕨p4 9wp5 or a 𝕣p4 9wp5. This indi‐
cates a condition check must be done to decide which command is to be executed. Thus
the | operator in regular expressions can be converted to if .. else statements.
However, creating such regular expression will require further processing on LX.
Moreover, as the if statements are complicated and require a large amount of data to
determine the conditions, the method to create potential subroutines in this paper ignores
if statements or any | operator. This is because the data set from makers’ interaction may
not be sufficiently large to establish a conclusive condition to check. Also, unlike loops,
there cannot be any default condition that can be placed either.
The method to find the real while loop conditions and if statement conditions can be
done with better computational intelligence tools that establishes the relationships of the
changing sensor values and estimates which sensors values are relevant to the conditions.
These methods have been designed for a specific purpose of allowing the makers to
construct the subroutines here. But these descriptions may be used for various other
applications as well. For example, in a general IoT system perspective, having a | oper‐
ator indicates multiple options in the interaction between the devices. Also, having a *
operator indicates multiple occurrence of commands, but the number of such occurrence
may need to be optimized based on the proper conditions.
In a RAL environment, once a maker creates an experimental rig, they run the rig locally
over a LAN to record the interactions. These interactions are then used to create the list
LX i.e. regular expressions. The procedure to convert the LX list of regular expression
to subroutines is as follows:
Step 1. Remove any element from LX that contains the | operator.
Step 2. The list LX is sorted descending order of the number of appearances. This
will put the regular expressions with the most frequency of occurrence at the front of
the list.
zamfira@unitbv.ro
494 A. Maiti et al.
Step 3. If there is any regular expression expr in LX that is part of another regular
expression expr’ and both have an equal number of appearances, then expr can be
removed from LX. This is because expr is always a part of a larger expression.
Step 4. For every regular expression expr in LX, process the with Algorithm Convert()
and as described earlier. Once the output program is ready, the name of the program is
generated randomly and all the variables used in them are put in as parameters to the
subroutine/program.
Finally, when all the potential algorithms are ready, the human maker of the experi‐
ment can then choose to save any algorithm as a subroutine or reject them. They may
alter some of the algorithms by placing the conditions for a while and for loops and if
segments.
This procedure can generate a rough sketch of the algorithm as a subroutine/program
that can be used to reproduce what the maker wants the users of the experiment to do
with it. The maker need not be presented with a potential program if the number of
appearances for the corresponding regular expression is very low, only the most
frequently used regular expressions are of interest. The maker may have control of how
many they want to see and save. The actual programing statements will depend upon
the programming language used for the devices.
zamfira@unitbv.ro
Identifying Partial Subroutines for Instrument Control 495
unison for moving forwards and backwards. Two sensors (D and E) were mounted atop
another actuator (C). The sensors (D and E) did not stream any value but the user had
to request the value through a CI when desired. A, B and C were also controlled through
a CI. In total there were 3 actuators and 2 sensors. This example was parameter restric‐
tive, i.e. a symbol write symbol is represented by 𝕨pv. Also in this example, the ports A
and B operate in parallel, thus they may be considered as a single port AB.
The maker of this experimental setup was a novice and thus the only commands the
maker could create was 𝕨AB, 𝕨C, 𝕣D and 𝕣E. There were no explicit wait commands used.
The inputs passed from the CI to the LEGO Mindstorms brick. The CI consists of 7
buttons associated with 4 ways the 𝕨AB (forward v1 backwards v2, left v3 and right v4)
may be executed, 𝕨C, 𝕣D and 𝕣E respectively. The experiment was designed to move the
robot around and collect data with the sensors at certain positions. A session of 145 s
was recorded and used as a training data set. The network latency was considered negli‐
gible for training.
5.5 Results
The histogram (partial) of the regular expressions is shown in Fig. 5. The regular
v1
expression (𝕨AB )* representing a move forward command has the highest number of
v2 v3 v4
appearances as this command is executed many times. Similarly, (𝕨AB )*, (𝕨AB )*, (𝕨AB )*
have appeared multiple times as well, as these are basic commands of the interface. But
there is no need to save these commands in while loops. The next set of commands that
( v1 v2 v3 v4
)∗
appears most is 𝕨AB 𝜌1 𝕨AB 𝜌2 𝕨AB 𝜌3 𝕨AB 𝜌4 . This indicates that the car was moved
multiple times with the set values v1, v2, v3 and v4. The corresponding program can be
saved as a particular function. The maker can replace the ρ1, ρ2, ρ3 and ρ4 wait variables
with constant values. This application allows the experiment maker to identify the
correct composite functions with multiple read/write commands. This way the maker
can learn about programming devices in new ways, which is an educational goal of the
RAL system. Also, after saving this composite function, it allows the maker to create a
better interface with specific commands reducing the effects of Internet latency on inter‐
activity.
zamfira@unitbv.ro
496 A. Maiti et al.
v1 v2 v3 v4
Fig. 5. The histogram of LX where s1 = (𝕨AB )*, s2 = (𝕨AB )*, s3 = (𝕨AB )*, s4 = (𝕨AB )*,
( v1 v2 v3 v4
)∗
s5 = (𝕨cv5)*, s6 = (𝕣D)*, s7 = (𝕣E)*, s8 = (𝕣D 𝕣E)*, s9 = 𝕨AB 𝜌1 𝕨AB 𝜌2 𝕨AB 𝜌3 𝕨AB 𝜌4 .
The histogram may be used for evaluation as well. This particular experiment (or
device) learns the sequence from the training data which is then uniquely attached to it.
Thus, it can be expected for this sequence to occur in any future interaction with other
nodes. For e.g. if the other users are given the same interactive CI and they do not perform
in an accurate manner, for e.g. skip some commands, then the corresponding histogram
will be different. The deviation in the histogram can be measured to determine how the
user has performed compared to an ideal histogram specific to this experiment.
The method to convert the expressions to the output programs is very basic and
heuristic in nature. However, it illustrates the usefulness of a regular expressions list.
Also, it is perfectly applicable in RAL as the system is expected to support the makers
as much as possible but not necessarily present accurate results.
The proposed method to create regular expressions from the interactions between
devices in an IoT system is robust. It can be used to define various interactions that
would otherwise be unknown at any given time. This creates a generalized model of the
interaction between devices autonomously. There can be several applications for these
properties as described earlier.
However, the process is very time consuming. For any given sequence with n
components the number of subsequences initially is:
|L | = n(n + 1)
| X| 2
The algorithm has a time-complexity of O(n2), and n could be very large. Thus, it
may be difficult to use this method in real time applications. However, it is well suited
for applications where the requirement is for post and pre-processing of the interactions.
Also, the regular expressions can be created and stored offline and once the interaction
starts any incoming sequence can be matched with the list of regular expression in real
time thus allowing for real-time validation and for adaptive controller-interfaces as
mentioned in Sect. 4.3.
zamfira@unitbv.ro
Identifying Partial Subroutines for Instrument Control 497
Another limitation is that the procedure described in this paper addresses only
command and port combinations. e.g. wp4 and rp3. Only the parameters passed with the
write commands are considered as variable. This is useful in context of RAL where
commands are in general associated with a specific port. However, further work can
examine ways to create subroutines that can account for variable ports and parameters.
Further work also needs to formulate and optimize the conditions that can be checked
in the if statements, and within conditional checks.
The main concern of this approach is that it limited to applications like RAL which
does not demand a very accurate reduction and detection of the subroutines. These can
rely on human to manually rectify any possible errors. If the number of sensors or
actuators increases or there are very large variations of inputs, then it becomes more
difficult to determine the regular expressions accurately.
6 Conclusions
A method to define a set of regular expression that covers the language between to nodes
in a NCS were discussed. This allows to the nodes to maintain validity of commands
and integrity of the nodes operation. It can also allow for altering interactions as required
according to the network latency conditions which are especially important in the context
of the Internet. Regular expressions can also be used for determining the deviation of
the usage of a node from a perceived ideal. These regular expressions have been applied
in case of a P2P RAL where they are converted to corresponding programs that can be
stored for a particular device. The regular expressions can enable the nodes to determine
the cause of any fault and trace it back to a particular set of commands that may be part
of a regular expression, but may not have been executed before. This can help in rein‐
forcement of learning approach of the devices.
References
Jinhui, Z., Yuanqing, X., Peng, S.: Design and stability analysis of networked predictive control
systems. IEEE Trans. Control Syst. Technol. 21, 1495–1501 (2013)
Whitmore, A., Agarwal, A., Xu, L.: The Internet of Things–a survey of topics and trends. Inf.
Syst. Front. 17, 261–274 (2015)
Harward, V.J., Del Alamo, J.A., Lerman, S.R., et al.: The iLab shared architecture a web services
infrastructure to build communities of Internet accessible laboratories. Proc. IEEE 96, 931–
950 (2008)
Markowsky, G.: An introduction to algorithmic information theory. Complexity 2(4), 14–22
(1997)
Grunwald, P.D., Vitanyi P.M.B.: Algorithmic information theory. arXiv:0809.2754v2 [cs.IT]
(2008)
Maiti, A., Kist, A.A., Maxwell, A.D.: Real-time remote access laboratory with distributed and
modular design. IEEE Trans. Ind. Electron. 62 (2015a)
Garcia, J., Fernande, F.: A comprehensive survey on safe reinforcement learning. J. Mach. Learn.
Res. 16, 1437–1480 (2015)
zamfira@unitbv.ro
498 A. Maiti et al.
Schmidhuber, J.: On learning to think: algorithmic information theory for novel combinations of
reinforcement learning controllers and recurrent neural world models (2015). http://arxiv.org/
abs/1511.09249
Maiti, A., Kist, A.A., Maxwell, A.D.: Components relationship analysis in distributed remote
laboratory apparatus with data clustering. In: IEEE International Symposium on Industrial
Electronics, pp. 861–866 (2015b)
Garcia, P., Lopez, D., de Parga, M.V.: DFA minimization: from Brzozowski to Hopcroft (2013)
Gross, J.L., Yellen, J.: Handbook of Graph Theory. Discrete Mathematics and it Applications
(2004). ISBN 1-58488-090-2
Li, M., Vitányi, P.M.B.: An Introduction to Kolmogorov Complexity and Its Applications, pp.
105–106. Springer (2008)
Mejías, A., Herrera, R.S., Márquez, M.A., et al.: Easy handling of sensors and actuators over TCP/
IP networks by open source hardware/software. Sensors 17, 94 (2017)
Jaulin, L.: Range-only slam with occupancy maps: a set-membership approach. IEEE Trans. Rob.
27(5), 1004–1010 (2011)
zamfira@unitbv.ro
Internet of Things Applied to Precision
Agriculture
Abstract. Nowadays, the number of small family farms has grown con-
siderably and it represents the main type of agricultural enterprise in
the world. The family activity in agriculture is considered significant in
terms of production of strategic food for the population, mainly in devel-
oping countries. Small family farmers, in general, are always on influence
of weather, and as a consequence, usually they do not maximize the har-
vest, reducing then the incoming. Analyzing the small family farming
current context, this paper proposes the development of a low-cost solu-
tion for control, monitoring and automation of agricultural greenhouse.
The proposed solution was designed using prototyping as Raspberry Pi
(RPi) and Arduino in conjunction with sensors (temperature, humid-
ity, and light, among others) and few actuators (drip system, fans and
incandescent lamps). For interaction between the farmer and the sys-
tems, it was developed a web human machine interface. Currently there
is a prototype of the proposed system running in the campus facilities
of Universidade Federal de Santa Catarina (UFSC) – Araranguá. Perfor-
mance and stability tests were made in the system in order to validate
the effectiveness of the proposed architecture. As a conclusion, with this
study developed using sensors and actuators in a controlled environment
prototype, it is possible to conclude that low cost solutions for family
farms are extremely necessary and feasible.
1 Context
Nowadays, industrial farms contribute positively in the trade balance of several
countries in the world, such as Brazil. As a result, the real importance of family
farms is overshadowed, thus, a segment within agriculture that is essential for
the countries for several issues, such as: economic, socio-cultural, environmental,
food security, among others, is no longer prioritized.
According to [5], there is no an universal definition of family farms, although
some are more widely accepted. However, as reported by [9] in 36 definitions of
family farms, nearly all definitions specify that to be considered in this category,
a member of the household owns, operates and/or manages the farm either in
part or fully.
c Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6 46
zamfira@unitbv.ro
500 R. Marcelino et al.
zamfira@unitbv.ro
Internet of Things Applied to Precision Agriculture 501
2 Purpose or Goal
Considering the importance of family farmers and the need to provide the oppor-
tunity to use low-cost technology that could contribute in the incoming of these
farmers, this work proposes the development of an efficient and innovative low-
cost solutions to control, monitor and automate small greenhouses.
The proposed idea is based on Internet of Things (IoT) concepts, where
the system was developed focusing the interconnection of all agents involved.
The projected system should be capable of acquiring data from various sensors
(temperature, humidity, solar irradiation, among others) in real time and display
them in a human machine interface (HMI). In addition, these data need to be
stored in a database in such a way that could be consulted at any time, either for
simple visualization or even for a data processing based on the data acquisition
history of the sensors.
Besides to the data acquisition, the system should also have a user-friendly
interface, where the user has control over all actuators that would be used in the
automated structure. Considering that this activation would be done remotely
using the internet as a mean of communication, this feature would bring conve-
nience for the farmer.
3 Approach
The embedded system proposed was focused on speed of development and
results. It was chosen to use processing solutions, such as Arduino and Raspberry
Pi platforms, because they represent solutions for fast and efficient projects.
The proposed system can be divided into four large sections, according to
Fig. 1:
1. Sensors and Actuators;
2. Acquisition/Processing Central;
3. Web application;
4. Human machine interface.
The Acquisition/Processing Central has two data processing systems, being
them: Arduino and Raspberry Pi. Arduino is responsible for performing low-level
and real-time data acquisition. After this process, these data is transported from
Arduino to Raspberry Pi via UART serial communication. On the other side,
Raspberry Pi has the responsibility to realize a higher-level data processing and
to communicate with the web server, aiming to store the data and to provide
information to the user through a friendly interface.
zamfira@unitbv.ro
502 R. Marcelino et al.
The monitoring and control system developed for data acquisition and actuation
in the crop is composed of some type of sensors and actuators. According to
Wendling (2010), sensor is a term used to designate devices sensitives to some
form of energy in the environment, relating information about a physical quantity
that needs to be measured. There are two main types of sensors, being: analog
or digital. Analog sensors can measure any type of value over the time within its
operating range. In the other side, digital sensors can assume only two values in
their output signal, being zero or one. Actuators are devices that will somehow
act on system variables.
As shown in Fig. 1, the block (1) represents the sensors and actuators used in
the system. Among the sensors used, the DHT11 is responsible to capture tem-
perature and ambient humidity. This sensor can measure temperatures between
0 and 50 ◦ C and humidity between 10% and 95%. It uses an NTC thermistor and
for temperature and HR202 for humidity. The internal circuit reads the sensors
and communicates with Arduino via the One Wire communication protocol.
Another sensor used in the system is DS18B20, which is responsible to mea-
sure soil temperature. It works in the range of −55 to 125 ◦ C with the error of
±0.5%. This sensor has an A/D converter circuit and a memory, in which the
data is saved. After the conversion, the Arduino reads the data through One
Wire digital communication.
Aiming to measure the soil moisture, it was used the sensor Hygrometer. This
is an analog sensor that is composed by a comment that, through two probes,
zamfira@unitbv.ro
Internet of Things Applied to Precision Agriculture 503
measures the soil moisture by measuring the current between the probes and by
a trimmer potentiometer circuit, in which the sensitivity is adjusted.
Among the actuators used, fans were used for air circulation and, as a conse-
quence, to change the temperature of the environment. As the proposed system
is a prototype, it was used two small fans 12 V 1 A. These fans can be started in
the HMI through digital ports from Arduino.
Another actuator used is incandescent bulb. Its function is to warm the
environment, according to the pre-defined values. The user can activate this
actuator in the HMI, where the Raspberry Pi receives this requisition and sends
it to the Arduino, which starts the system sending a digital signal for a relay.
Finally, an irrigation system was proposed to adjust soil moisture and nutri-
tion of the crops. This actuator is constituted by an electromechanical valve that
when triggered via relay, allows the flow of water. Again, the user can control
this system through the HMI.
As shown in Fig. 1, the block (2) represents Arduino and the microcomputer
Raspberry Pi. In general, Arduino is the component responsible to read the data
from the sensors and to trigger the actuation systems proposed for this system.
Another important responsibility is to send this data acquired to Raspberry Pi.
The communication is bidirectional, so the Arduino will receive commands from
the Raspberry Pi to control the actuators.
In the other side, microcomputer Raspberry Pi is responsible for the high-
level data processing and the connection between the prototype and the web
application. Its roles are to receive data from the sensors captured by the
Arduino, process the received data, send it to the user application, receive com-
mands to trigger the actuators via web application and send these commands to
the Arduino. It was used serial connection to connect the Arduino and Raspberry
Pi. In the other side, the connection between Raspberry Pi and web application
was defined through a wireless network (WiFi IEEE 802.11).
The description of the web application development can be divided into two
subsections: the first one detailing the subsystem responsible to receive data
from the physical system and the second one describing the human machine
interface (HMI) created to interact with the users.
zamfira@unitbv.ro
504 R. Marcelino et al.
he affirms that IoT is a huge mashup of Web services, browser technology, and
embedded technology. JavaScript is pretty useful in all those places [6].
MongoDB technology was chosen to storage the data acquired from the sen-
sors and the data generated from the HMI. This is a high performance non-
relational database that is ideal for storing large volumes of information. This
technology has sufficient maturity for applications of this size and this type.
Another reason to use MongoDB in this project is the fact that this database
can manage data of any structure and it is possible to add new functionalities
without redesigning the entire database [14].
Basically the web server developed exposes clients to a REST API where,
through HTTP calls, the microcomputer Raspberry Pi sends the data collected
through the Arduino to be stored in the cloud. For real time communication
between the web application and Raspberry Pi, web sockets were used. This
technology creates a persistent communication channel between remote objects
and transfer the data through the network asynchronously. In the other way of
the communication channel, where the user can send commands for the con-
trol system, it was defined that would be used the same online communication
channel (web sockets).
zamfira@unitbv.ro
Internet of Things Applied to Precision Agriculture 505
that one of the main objective was to provide the possibility to monitor and
control the system from anywhere, at any time, through a computer or even a
mobile device.
The web application described in the previous item, besides being responsible
for the acquisition, processing and storage of the data, also provides the user the
opportunity to interact with the system, as is described below:
Figure 2 demonstrate the home page of the web application for interaction
with the users. In it, it is possible to see the sensors that are inside the automated
greenhouse arranged in an intuitive way. The updating of these values on this
page occurs automatically as soon as the sensors change their state.
Figure 3 shows another system module, where the user can visualize through
graphs or tables the data history of a specific sensor. This query can be done
using time period filters, that can be defined by the user.
Figure 4 shows the part of the web system where the user can send commands
directly to the automated system. Actuators that are physically arranged in the
greenhouse structure are connected to the acquisition/processing central that
zamfira@unitbv.ro
506 R. Marcelino et al.
receives the commands from the cloud through web socket. This part of the
whole system allows control of, for example, the irrigation valve, the fans, and
the bulb.
The main outcome of the developed prototype was to prove that the utilization
of new and more economics embedded systems can act effectively in control
of familiar greenhouses, being one of the solutions of IoT applied in precision
agriculture.
The system shows its viability of working for about a month without any
problem. The signals were collected, processed, decisions were taken, and the
data were stored for future analyses.
All of the sensors used are low cost sensors and highly available as well. The
application did not require big precision, therefore, devices like the DHT11 and
the DS18B20 met the application without any signal conditioning.
For the Acquisition/Processing Central, the use of two data processing was
salutary, being that the Arduino was responsible for the low level and the
Raspberry PI was responsible for the integration with the web. Both systems
responded adequately in the means of speed, sending and receiving data in cor-
rect time.
In general, the propose system could be considered an appropriate solution
to low economic family farms.
The web application used the not-relational database MongoDB, and it
worked very well considering the instability, data security, and fast answering
to the queries. After a month using the prototype, the database was using just
zamfira@unitbv.ro
Internet of Things Applied to Precision Agriculture 507
100 MB considering that the system was storing data each 10 min. The web
server, developed in Node, works stably, giving the impression that it was the
correct solution for this problem.
The website layout, developed to support responsiveness, is accessible by
mobile devices. The HTML webpage showed responsiveness overall with the
smartphones tested.
In addition to the HTTP calls that the RPi made to the server, it also existed
a persistent communication using the websocket. This communication did not
have any problem; sending messages between the site to RPi was performed
how it was supposed to. An important part to achieve this result was the use of
Socket.IO, that is an abstraction of the websocket.
It is also important to point out that use of javascript is facilitated, mainly
Node.js, because it has a large library that stand for many necessities and it is
supported by international players like Google, Facebook, Intel among others.
This IoT Technologies are tools integrated in the model of acquisition and
monitoring system, which goal is to become the best system to receive and share
the information about the monitored system.
5 Conclusion/Recommendations/Summary
The Internet of Things is being used in many fields today. The agriculture
is just another field of application. The cost reduction of sensors and elec-
tronic devices, miniaturization of these devices, use of global positioning system
(GPS), geographic information system (GIS), are all elements that boost this
perspective [17].
With a stronger use of sensing in the crops, soil analysis, humidity, temper-
ature, weather, among others, it will build many crop data to make the best
decision at the right time and the precise way of improving the agriculture pro-
duction [8].
In this study, the focus was centered in the computer technologies applied in
greenhouses. The actual controlled environments do not have enough computer
technology applied and this study proved the viability to apply this technology
in family farmer.
The family farmers employ many people around the world, but they do not
have enough investment capacity, however, they build a strong network of local
agricultural production and regional consumption. This paper showed a real pos-
sibility of applying new and economical computational technologies to support
family’s crops. With the application of modern technologies, the family farmers
will not lose space for the big producers, where the capacity of automation is
larger.
The developed system responded as expected, acting upon request and stor-
ing the data for future analyzes. The system answered in milliseconds despite
the problem not requiring fast answers.
Aiming to guarantee the operation of the system, it is suggested the instal-
lation of energy maintenance system for the moment when there are shortages.
zamfira@unitbv.ro
508 R. Marcelino et al.
The use of the system such as Arduino and Raspberry PI, represented high
speed of development with low cost. There was not necessity to develop a dedi-
cated system because the family farmers do not have enough money to support
a specific and dedicated system.
For new plants, we hope to apply the system in a real greenhouse to evaluate
the results of productivity and quality of the crops. At this moment, it was used
only a prototype of a real greenhouse. This work, at this moment is only focused
on the possibility of using this system, and their application. Now the goal is to
apply it in the real field.
The use of two processing systems is justified by the lack of analog inputs
in Raspberry, being exclusive for Arduino the dedication to processing sensor
readings and the activation of the actuator, and finally, by the speed of the
development of web applications that the Raspberry enables, since it has an
embedded operating system.
It can be also defined that some economic feasibility studies still need to
become a target among the main concerns of small-scale agricultural technicians,
engineers and managers. It will be important to relate the costs and benefits of
implementing these systems, as well as the rates of return on investments. It is
necessary to translate into money what the cost reduction means and increases
in productivity, as well as gains in food/flower production in these greenhouses.
The integration of applied IoT in precision agriculture will make the agricul-
tural sector more productive. Planting at the right time, right place and with
the right resources will give birth to a new way of planting, focused on product
quality and high productivity.
References
1. Ai, W., Chen, C.: Green house environment monitor technology implementation
based on android mobile platform. In: 2011 2nd International Conference on Artifi-
cial Intelligence, Management Science and Electronic Commerce (AIMSEC). Insti-
tute of Electrical and Electronics Engineers (IEEE), August 2011. http://dx.doi.
org/10.1109/AIMSEC.2011.6010025
2. Bittencourt, A., Müller, R.S.W.: Avaliação dos Princı́pios de Usabilidade. N/A,
São Paulo (2014)
3. Eredics, P.: Measurement for intelligent control in greenhouses. In: Proceedings of
the 7th International Conference on Measurement, pp. 440–447 (2009)
4. Guanziroli: Agricultura familiar e reforma agrculo xxi. Editora Garamond, June
2001
5. Heberlê, A.L.O.: A agricultura familiar brasileira no contexto mundial,
January 2014. https://www.embrapa.br/busca-de-noticias/-/noticia/1871776/
artigo-a-agricultura-familiar-brasileira-no-contexto-mundial
6. Hunter, L.: The smartest way to program smart things: Node.js: The reasons to
use node.js for hardware are simple: it’s standardized, event driven, and has very
high productivity, February 2015
7. Moreno, J.C., Berenguel, M., Rodrı̀guez, F., Baños, A.: Robust control of green-
house climate exploiting measurable disturbances. In: 15th Triennial World
Congress of the International Federation of Automatic Control (2002)
zamfira@unitbv.ro
Internet of Things Applied to Precision Agriculture 509
8. Lee, W.S., Ehsani, R.: Sensing system for precision agriculture in Florida. Comput.
Electron. Agric. 112, 2–9 (2015)
9. Lowder, S.K., Skoet, J., Singh, S.: What do we really know about the number and
distribution of farms and family farms worldwide? Background paper for the state
of food and agriculture (2014)
10. Moga, D., Petreus, D., Stroia, N.: A low cost architecture for remote control and
monitoring of greenhouse fields. In: 2012 7th IEEE Conference on Industrial Elec-
tronics and Applications (ICIEA), pp. 1940–1944, July 2012
11. Rangan, K., Vigneswaran, T.: An embedded systems approach to monitor green
house. In: Recent Advances in Space Technology Services and Climate Change
2010 (RSTS and CC-2010), November 2010. Institute of Electrical and Electronics
Engineers (IEEE). http://dx.doi.org/10.1109/RSTSCC.2010.5712800
12. Shin, C.S., Lee, Y.W., Lee, M.H., Park, J.W., Yoe, H.: Design of ubiquitous glass
green houses. In: 2009 Software Technologies for Future Dependable Distributed
Systems, March 2009. Institute of Electrical and Electronics Engineers (IEEE).
http://dx.doi.org/10.1109/STFSSD.2009.48
13. Rong-Gao, S., Zhong, W., De-Chao, S.: Greenhouse temperature and humidity
intelligent control system. In: Proceedings of the 3rd WSEAS International Con-
ference on Circuits, Systems, Signal and Telecommunications, pp. 120–125 (2009)
14. MongoDB: Internet of Things (2016). https://www.mongodb.com/use-cases/
internet-of-things
15. Teemu Ahonen, R.V., Elmusrati, M.: Greenhouse monitoring with wireless sensor
network. In: Proceedings of the IEEE/ASME International Conference on Mecha-
tronic and Embedded Systems and Applications, pp. 403–408 (2008)
16. Thiemo Krink, R.K.U., Filipic, B.: Evolutionary algorithms in control optimiza-
tion: the greenhouse problem. In: Proceedings of the Genetic and Evolutionary
Computation Conference, pp. 440–447 (2001)
17. Zhang, N., Wang, M., Wang, N.: Precision agriculture-a worldwide overview. Com-
put. Electron. Agric. 36, 113–132 (2002)
zamfira@unitbv.ro
Computer Vision Application
for Environmentally Conscious
Smart Painting Truck
1 Introduction
Growing population and increasing number of megacities result in high traffic density
contributing to the need for new rules and regulations to address environmental health
and safety concerns. These concerns necessitate several regulations to maintain and
upgrade road infrastructure for passenger safety and to achieve carbon emission targets
2 Literature Review
Previous applications of image processing for road marking recognition and detection
for environmental benefit are quite limited in the literature. Related studies usually
focus on advanced driver assistance applications and intelligent transportation systems
(Charbonnier et al. 1997; Cruz et al. 2016; Kheyrollahi and Breckon 2012; Lin et al.
2016; Mathibela et al. 2015). One of the relevant studies includes three-level road
marking algorithm for automatic extraction of the repainting process parameters
Mathibela et al. (2015). In addition, Woo et al. (2008) presented a novel robotic system
for damaged lane mark detection, while Kheyrollahi and Breckon (2012) proposed a
multi-step processing pipeline for an automatic real-time road markings and text
recognition under a variety of driving, lighting and road conditions. Lin et al. (2016)
developed a road marking quality assessment mechanism where the proposed system
receives digital images of the road marking surface, processes images, and analyzes
them to capture the geometric characteristics of the marking. The geometric charac-
teristics are then assessed to determine the quality level of the finished marking.
Cruz et al. (2016) studied the environmental impacts of road markings considering the
entire life-cycle of each material used in the process. To the best of our knowledge,
zamfira@unitbv.ro
512 A. ElSayed et al.
3 Methodology
In this paper, a real-time tank level monitoring has been designed where the painting
vehicles are directed to the nearest refill station when the inventory level in the tanks is
low. The system allows data collection and analysis for online real time quality and
performance measurement. Resulting information is shared with multiple parties to
improve the decision making at the operational and managerial levels. The block
diagram of the proposed system is depicted in Fig. 1.
Proposed system is based on a number of assumptions explained below.
• A homogeneous fleet of single tank painting vehicles is used for road markings.
• A single material, namely paint, is used in the road marking process.
• A single refill station serves each vehicle.
• All model parameters such as the distances, tank capacity, transportation cost per
unit of traveled distance between each region are known and deterministic.
zamfira@unitbv.ro
Computer Vision Application for Environmentally Conscious 513
The sets, parameters and variables of the model are explained below.
Sets
i: index for regions, i 2 I ¼ f1; 2; 3; . . .; ng,
v: index for vehicle, v 2 V ¼ f1; 2; 3; . . .; mg,
where,
n: is the number of regions the vehicles can pass through,
m: is the number of working vehicles.
Parameters
T: transportation cost per unit distance,
arcði; jÞ: arc length from region i to region j ðj 2 IÞ,
Cvmin : minimum paint level at vehicle v.
Variables
xv : amount of paint in vehicle v,
1; if arcði; jÞ visited by truck v
yijv :
0; otherwise:
Objective Function
8v 2 V; if xv Cvmin in region s find the optimum value of z to move car v from s to
refill station t,
where:
!
X
m X
n X
n
min z ¼ min arcði; jÞyijv T : ð1Þ
v¼1 j¼1 i¼1
Equations (2) and (3) represent the shortest path flow constraints. Equation (4)
represents the binary variable.
zamfira@unitbv.ro
514 A. ElSayed et al.
Painting Tank
Camera
Spray Valve
Painting Tank
Camera
Spray Valve
After the recording of the captured input stream, each captured frame is processed
using the following steps:
Step 1. Adjust the frame size for proper processing.
Step 2. Apply de-noising filter to reduce the noise in the captured image.
Step 3. Separate the captured road marking from the background and produce a
binary image using thresholding.
Step 4. Use morphological operations to fix the problems in the thresholding
process.
Step 5. Calculate the painted area based on camera calibration information.
The system allows two laser range finders to be attached to the tank module for
accurate calculation of the thickness of the painted area which then can be used to
calculate the paint volume. Using this information, the remaining amount of paint and
the time to refill the tank can also be calculated. Furthermore, a scale-invariant feature
transform (SIFT) Lowe (1999) tracking technique and data from the accelerometer are
utilized to detect if the vehicle is moving or not. This technique temporarily pauses the
calculations if the car is stationary until the car starts moving or until the task is
completed.
zamfira@unitbv.ro
Computer Vision Application for Environmentally Conscious 515
4 Numerical Experiments
The MILP model aims at minimizing the transportation cost. In the numerical exper-
iment, the unit transportation cost is assumed to be $0.40/mile, where the maximum
paint level for each vehicle is 100 gallons, and the minimum paint level is 20 gallons.
There are total of three painting vehicles on service and five regions. The shortest path
algorithm is solved using GAMS 24.7.4. According to the results, the total trans-
portation cost for three vehicles is found to be $18.
The proposed computer vision technique has been implemented using OpenCV
library with Python language. A live feed from the 5 MP camera that captures 30
frames per second has been used. Figure 4 shows an example of the frame processing
steps. For this example, the consumption volume of paint for marking an area is
calculated by the computer vision system. The painted area calculated from the cap-
tured videos is 16,179.5 cm2 where the thickness of the paint is 1 mm, and the amount
of paint used for this area is 0.43 galloon.
(a) (b)
(c) (d)
5 Conclusion
This research introduced a novel technique which estimates the paint consumption used
for highway road marking systems and determines the minimum transportation cost.
For this purpose, a shortest path problem is defined and formulated as a mixed integer
linear program. Moreover, a computer vision-based system equipped with multiple
zamfira@unitbv.ro
516 A. ElSayed et al.
cameras and laser sensors is proposed in order to calculate the paint consumption
during the road marking process. Using this data the remaining time to reach the
minimum paint level in the tank is determined. The proposed system can be used for
more advanced road marking signs such as words and/or directional arrows, which are
currently handled manually.
As a further study, the smart painting system can altered to accommodate multiple
road marking materials. Additionally, a global positioning system can be included to
determine the exact position of the painting vehicle(s) for a superior functioning
shortest path algorithm.
References
Charbonnier, P., Diebolt, F., Guillard, Y., Peyret, F.: Road markings recognition using image
processing. In: IEEE Conference on Intelligent Transportation System, ITSC 1997, pp. 912–
917, 9–12 November 1997. doi:10.1109/ITSC.1997.660595
Cruz, M., Klein, A., Steiner, V.: Sustainability assessment of road marking systems. Transp. Res.
Procedia 14, 869–875 (2016). doi:10.1016/j.trpro.2016.05.035
Kheyrollahi, A., Breckon, T.P.: Automatic real-time road marking recognition using a feature
driven approach. Mach. Vis. Appl. 23, 123–133 (2012). doi:10.1007/s00138-010-0289-5
Lin, K.-L., Wu, T.-C., Wang, Y.-R.: An innovative road marking quality assessment mechanism
using computer vision. Adv. Mech. Eng. 8, 6. doi:10.1177/1687814016654043
Lowe, D.G.: Object recognition from local scale-invariant features. In: Proceedings of the
Seventh IEEE International Conference on Computer Vision, vol. 1152, pp. 1150–1157
(1999). doi:10.1109/ICCV.1999.790410
Mathibela, B., Newman, P., Posner, I.: Reading the road: road marking classification and
interpretation. IEEE Trans. Intell. Transp. Syst. 16, 2072–2081 (2015). doi:10.1109/TITS.
2015.2393715
Veit, T., Tarel, J.P., Nicolle, P., Charbonnier, P.: Evaluation of road marking feature extraction.
In: 2008 11th International IEEE Conference on Intelligent Transportation Systems, pp. 174–
181, 12–15 October 2008. doi:10.1109/ITSC.2008.4732564
Woo, S., Hong, D., Lee, W.-C., Chung, J.-H., Kim, T.-H.: A robotic system for road lane
painting. Autom. Constr. 17, 122–129 (2008). doi:10.1016/j.autcon.2006.12.003
zamfira@unitbv.ro
Remote Monitoring and Detection of Rail Track
Obstructions
Abstract. Railway snow and sand monitoring has become an important safety
issue, as they pose a serious threat to lives, property, and security. Snow disasters
every year bring immeasurable losses to society, and sand obstructions can cause
the shutdown of railway lines for weeks. This paper presents a detailed design
and implementation of a non-contact railway monitoring system using a camera
mounted on a mobile platform. The captured image is processed to identify the
level of obstruction on a rail track. The goal is to transmit the obstruction data to
the cloud in real time and enable monitoring using a web based graphical user
interface. Once the obstruction crosses a predetermined threshold, the system will
alert the officials.
1 Introduction
Over 140,000 miles of railroad track comprise the United States (US) railway network.
About a thousand of rail road incidents have been reported by the Federal Railroad
Administration in the last decade [1, 2]. Weather conditions have a major influence on
rail tracks and affect the operating efficiency, physical infrastructure, and the safe
passage of freight and people.
Despite the availability and sophistication of advanced weather information systems,
adverse snow and sand deposits continue to cause problems for railway operators. Snow
and sand accumulations can occur on rail switches, brake riggings, track flange ways,
and grade crossings, thereby reducing control over the train and increasing the risk of
derailments and other types of incidents. The wheels of locomotives and rail cars are at
risk for slipping and sliding when tracks are coated with snow and sand. Excessive sand
deposits cause track blockages, ballast ingress, and jamming of switches and gear boxes.
Figure 1 shows rail track obstructions in Cape Town and Morocco [3, 4].
A rail track consists of rails and ballasts. The tracks are higher than the ballasts. A
train can run without much danger even if the ballasts are submerged by an obstruction;
however, if snow or sand is covering the rails, then it possesses danger to the train’s
operation [2].
This paper introduces a non-contact remote monitoring system for rail tracks that
detects the amount of obstruction on the rail track utilizing a camera on a mobile
zamfira@unitbv.ro
518 M.M. Uddin et al.
(a) Cape Town, South Africa [3] (b) Oriental Express, Morocco [4]
platform. The images collected by the camera will be continuously monitored and the
amount of obstruction is quantified using the proposed image processing algorithm. The
obstruction data obtained from the image processing unit are sent to the cloud server for
remote monitoring. Any time the obstruction level crosses a threshold, the system will
provide an alarm for remote observers.
2 Existing Systems
The traditional method of measuring snow or sand depth involves using a scaled rod,
which is inserted into the deposited snow or sand on the ballast of the rail tracks. This
is a labor intensive and time consuming operation. With the development of new tech‐
nologies, automated systems have been introduced for rail track monitoring for obstacle
detection. Currently, available obstruction depth measuring systems use ultrasonic
waves and lasers.
zamfira@unitbv.ro
Remote Monitoring and Detection of Rail Track 519
This method involves sending a short burst of laser pulse to the ground and measuring
reflected light from the snow/sand surface. It utilizes the pulse and phase measurement
of the laser signals. Figure 3 demonstrates the principle of operation of a typical laser
measurement system. L1 − L2 is the difference of the pulse traveled (before and after
the snow/sand present on the tracks). By using angle a (from Fig. 3), the snow depth can
be obtained:
where, hs is the snow depth. This system has a long range for measurement and better
accuracy. It is not affected by weather conditions either. However, the cost of such a
system is very high. The same as the previous system, this one also has the drawback
of stationary/fixed installation, thereby limiting its use to only a small area. Additionally,
it cannot monitor the whole length of the rail track.
zamfira@unitbv.ro
520 M.M. Uddin et al.
The system proposed in this paper utilizes a video camera to monitor and detect obstruc‐
tions on rail tracks. A camera system is mounted on a moving platform (train or quad
copter) pointing toward the track. The camera feed is captured periodically and is
processed using a proposed image processing algorithm to identify the rail track and, in
turn, to determine whether the track has any obstruction that could pose a danger for a
passing train. The system reads an image from the video camera and recognizes the rail
tracks from the background, which is in turn processed to measure the amount of rail
track coverage by snow/sand or any other object. The coverage on the ballast and rail
areas are independently processed and computed. As rails are higher than the ballast,
the rails are used to set threshold levels to compare the amount of snow/sand present on
the tracks. Even if the ballast is completely submerged by the snow/sand, it still does
not pose serious threat to the rail track operation. However, if the rails are submerged,
then it will be identified as an immediate danger.
Figure 4 illustrates the processes involved in the proposed monitoring system and
can be divided into three major components: image processing, obstruction monitoring,
and graphical user interface (GUI).
zamfira@unitbv.ro
Remote Monitoring and Detection of Rail Track 521
The system uses a video camera focused on the rail tracks. The images from the camera
are processed and filtered to separate the rail track from the background. The steps in
the image processing part are shown in Fig. 5.
zamfira@unitbv.ro
522 M.M. Uddin et al.
Also the camera is mounted on a fixed pole directed toward the rail track. The camera
is focused to acquire the views of the rail and ballast, which are the key components in
the inspection. Figure 6 shows an image of a captured frame. In real-life operation, a
camera with higher resolution will be mounted on a mobile platform such as a train, rail
car, or quad-copter.
zamfira@unitbv.ro
Remote Monitoring and Detection of Rail Track 523
The binary image is filtered through a structural element of sizes varying from 3 × 3
to 8 × 8. The erosion of an image f (x, y) by a structural element s is denoted by f 𝜃 s [6],
where 𝜃 denotes the erosion between f and s. The output image g(x, y) is processed based
on the following rule, repeating for all x and y:
{
1 if s fits f
g(x, y) =
0 if otherwise
Thus, erosion creates a new image that marks all of the locations of a structuring
element’s origin at which it fits the input image. Figure 8 shows the eroded images
(structural element s of size 3 × 3). The operation has stripped away a layer of pixels
from the object, shrinking it in the process. Pixels are eroded from both the inner and
outer boundaries of regions, so erosion enlarges the holes enclosed by a single region
as well as makes the gap between different regions larger. Erosion also eliminates small
extrusions on a region’s boundaries. For better results the erosion process is applied
twice, both with structural element s of size (3 × 3). Much of the granular noise is
removed in the process. There is still some left that will be removed in the later steps.
Erosion removes small-scale details from the binary image and simultaneously also
reduces the size of regions of interest. By subtracting the eroded image of the rail track
from the original image, the boundaries of each region are found as b = f − g where, b
is an image of the region boundaries. Another process called dilation is used to smooth
out the eroded image. The dilation of the eroded image g by a structuring element ś
(denoted as g ⊕ ś ) produces a new binary image h = g ⊕ ś with ones in all locations
(x, y) of a structuring element ś origin at which that structuring element ś fits the input
image g, repeating for all x and y, i.e.
{
1 if s fits g
h(x, y) =
0 if otherwise
Dilation basically has the opposite effect of erosion [6], it adds a layer of pixels to
both the inner and outer boundaries of regions for smoothing out the regions of interest
zamfira@unitbv.ro
524 M.M. Uddin et al.
from the rails (as illustrated in Fig. 9). Erosion removes the unwanted noise, and the
dilation process enhances and amplifies the remaining regions for better processing. For
better results, three iterations of dilation process are applied with s(4 × 4).
A Gaussian low pass filter of suitable standard deviation (𝜎 = 4.0) is used for
smoothing out the image. Since edge detection is susceptible to noise in the image, this
step is essential to remove the noise and also smooth out the weak edges.
The smoothed out image is then filtered with a Sobel kernel in both horizontal and
vertical directions to find the first derivative in the horizontal direction and vertical
directions. These two images identify the edge gradient and direction for each pixel as
follows [6]:
zamfira@unitbv.ro
Remote Monitoring and Detection of Rail Track 525
Hough Transform
Hough transform is used for feature extraction to isolate the edges that correspond
to the rail tracks from the others [10]. Due to imperfections in either the image data or
the morphological filter, there may be missing points or pixels on the desired curves as
well as spatial deviations between the ideal line/circle/ellipse and the noisy edge points
as they are obtained from the edge detector. Figure 11 shows how the four points can
be interpreted by the edge detector.
The image obtained from the morphological filter is passed through a low pass filter
to locate regions that are either 1 or 0. This yields a hard filtered image with edges (not
yet useful for rail detection). Basically the concept is to accumulate all the points in the
low-pass filtered image and calculate the space of each point and how many lines inter‐
sect the point. A point (x, y) will have a line
y = ax + b,
y′ = a′ x′ + b′ ,
zamfira@unitbv.ro
526 M.M. Uddin et al.
Figure 13 shows the final image output from the processing unit. The edges from the
canny edge detector and Hough transform are superimposed in this final image. This
shows the rail area and ballast area separately. Different colors are used to highlight
different regions of the rail tracks. As snow/sand starts to accumulate on the tracks, the
amount of rail track area visible in the image starts to decrease. This information is
extracted in the form of percentage of area covered by the obstruction.
zamfira@unitbv.ro
Remote Monitoring and Detection of Rail Track 527
using blob detection [9–12]. Blobs are considered as homogenous areas that have the
similar properties (like color and shape). It detects and extracts the sand/snow/rock
particles present on the ROIs extracted in the previous step.
We added sand, pebbles, and stones to the 2D rail track for testing purposes. The
amount of obstruction added was recorded, and the image processing unit (IPU) detec‐
tion and estimation of these obstructions were performed. Figure 15 shows the compar‐
ison of obstruction levels between the ones estimated by the IPU unit and manual calcu‐
lation. The percentage of area covered by the obstruction seems to closely match the
values from the manual estimation.
zamfira@unitbv.ro
528 M.M. Uddin et al.
Fig. 15. Manual and IPU approach for estimating obstruction levels.
The system is designed for remote monitoring of obstructions present on rail tracks. The
IPU processes and detects the amount of obstruction present on the tracks, which is then
sent to an online server in real time. Simultaneously, GPS coordinates from an on board
chip provides the location information of the obstruction. Both the obstruction values
from the IPU and the GPS location along with a time stamp are sent to the server in the
form of a.CSV file. Threshold levels can be adjusted by a user depending upon an esti‐
mated critical value for a train line. The critical value is set based on the amount of
allowable coverage of the tracks. If the rails are found to be covered by the IPU, then a
flag is raised and sent to the server via the data file. The server reads the flag field and
raises an alert on the graph with the corresponding GPS location.
zamfira@unitbv.ro
Remote Monitoring and Detection of Rail Track 529
Visual Studio 2013 is used to develop the ASP.NET application. Camera feed of the
rail track is projected on the page using JavaScript code. The graphical layout is
displayed using CanvasJS charts, which allows us to make interactive graphs with
animated projections (Fig. 18).
zamfira@unitbv.ro
530 M.M. Uddin et al.
Figure 18 shows how the different JavaScript codes are used to project the raw image
from the camera and the obstruction levels with GPS locations on the ASP.NET GUI.
4 Conclusion
The paper describes the development of a remote railroad obstruction detection system
utilizing image processing techniques. A real time camera output is processed and passed
to the cloud and presented to the remote users via a GUI. The project involves detecting
the individual components of a rail track, detection and calculation of obstruction levels,
and cloud based remote monitoring user interface. The system has been designed,
developed, and tested with a 2D rail track image. The system can be extended for a
physical 3D track using a mobile camera mounted on a drone. Alerts can be set to trigger
emails or text messages to concerned officials.
References
1. Phillips, D.A.: RWDI Consulting Engineers & Scientists, Analysis of Potential Sand Dune
Impacts on the Railway Tracks and Methods of Mitigation. www.iktissadevents.com/files/
events/gtrc/1/presentations/d2-s8-duncan-phillips.pdf
zamfira@unitbv.ro
Remote Monitoring and Detection of Rail Track 531
zamfira@unitbv.ro
Improving Communication Between Unmanned Aerial
Vehicles and Ground Control Station Using Antenna
Tracking Systems
1 Introduction
zamfira@unitbv.ro
Improving Communication Between Unmanned Aerial Vehicles 533
2 Problem Formulation
3 Problem Solution
For monitoring the crop vegetation status [5] a special designed data acquisition system
which has different type of sensors was used. The most important sensor is composed
by four STS VIS-NIR Ocean Optics spectrometers [6]. They are mounted on an UAV
capable to carry the necessary equipment. Hyperspectral data obtained from the spec‐
trometers are recorded through a software developed by our research team. A Raspberry
Pi 3 (RPi3) development board was used so that to benefit of its computing power,
graphic facilities and included access point characteristics. The access point is a wireless
communication (Wifi) component, which help us to transfer information between the
GCS and the data acquisition system placed on the UAV. To have real time data we are
limited by the wireless communication distance between the GCS and the UAV.
The GCS controls a standard Antenna Tracking System (ATS) so that to follow the
position of the UAV. Being a commercial and open source solution currently used for
First Person View (FPV) flight, it is not the paper task to describe it. A similar ATS is
also mounted on the UAV.
zamfira@unitbv.ro
534 S. Pop et al.
In Fig. 1 the GCS antenna tracking system is presented. It consists of a control device
connected with a microcontroller development board, an Inertial Measurement Unit
(IMU) sensor, a GPS module, and as actuators we used two servos, one for pitch correc‐
tion and the other to correct azimuth.
In Fig. 2 is presented the structure of the UAV antenna tracking system together with
the data acquisition system used for crops vegetation status monitoring. UAV is
equipped with the same type of pan-tilt ATS module as the GCS.
Fig. 2. Structure of the UAV antenna tracking system and data acquisition system
zamfira@unitbv.ro
Improving Communication Between Unmanned Aerial Vehicles 535
UAV. The magnetometer and the barometer sensors offer data that are used for
keeping UAV in LOS.
This antenna tracking system can be mounted on different vehicles (Rover or UAV)
and what the most important is that it do not depend on a flight controller.
IMU–10DOF Breakout sensors, have the following functions:
• The LSM303 module has a 3D digital linear acceleration sensor and a 3D digital
magnetic sensor. LSM303 provides the information about GCS compass head on the
three axis of the magnetometer. Before starting it is necessary to calibrate the sensor
in order to know the GCS azimuth. Data obtained from the three-axis accelerometer
were not used in our application;
• L3GD20 module is a three-axis angular rate sensor, used to obtain roll, pitch and yaw
data. The pitch angle of the UAV antenna tracker help us for keeping constant steering
towards GCS;
zamfira@unitbv.ro
536 S. Pop et al.
4 Results
In Fig. 5 the main data acquisition window is presented. It includes values related to the
acquisition process: spectrometers identifiers, number of spectrum reading iterations,
integration time, reference and zero point levels, data acquisition control (start and stop),
and reflection coefficients charts.
Hyperspectral data acquired from the spectrometers are stored on a SD card by the
RPi3 developed software or can be sent in real time to the GCS. Collected data are used
to compute the Normalized Difference Vegetation Index (NDVI) to estimate crop yields.
The accurate starting point of the measurement has to be known so that the information
to be correlated with the map configuration.
Communication delays can occur in transmission between the UAV and the GCS
preventing to get real time data from the spectrometers.
Each flight mission is programmed in the GCS. An example of a flight mission
prepared in the Mission Planner software [9], is presented in Fig. 6.
zamfira@unitbv.ro
Improving Communication Between Unmanned Aerial Vehicles 537
Fig. 6. Example of a flight mission prepared in Fig. 7. Relative position of the UAV ATS and
the GCS software GCS antenna
4.2 The Calculation of UAV Antenna Position Related to the Pitch Angle
In Fig. 7 we present the necessary elements to determine the relative position of the
UAV antenna tracker and GCS antenna. The UAV ATS controller records the takeoff
position and continuously the data from the GPS.
Permanently it calculates the distance between the projection on the ground of the
UAV GPS position and the fixed GCS position. Using the GPS data, the UAV ATS
controller computes the d distance between the two antennas.
We had to choose one of the methods used in navigation, to calculate the d distance
between the two antennas: a rhumb line (or loxodrome) or to apply the haversine
formula.
A rhumb line, or loxodrome [10] is an arc crossing all meridians of longitude at the
same angle, i.e. a path with constant bearing as measured relative to true or magnetic
north. This method is recommended to calculate the distance d over long range
(>1000 km).
The haversine formula is an equation used to calculate distances between two points
on a sphere from their longitudes and latitudes co-ordinates and we implemented it in
our software, having the advantage of the obtained GPS data (longitudes and latitudes)
from the UAV ATS controller. Usually, GPS co-ordinates, longitude and latitude, are
in deg-min-sec suffixed with N/S/E/W (e.g. 45°37′60″N, 25°34′60″E) format. To apply
the haversine formula, this co-ordinates must be converted in signed decimal degrees
without compass direction, where negative indicates west/south (e.g. 45.6333, 25.45 or
40.798302, −73.985006).
The haversine formula is:
d = R∗ c (3)
zamfira@unitbv.ro
538 S. Pop et al.
where:
Δϕ = (lat 2 – lat1); difference of latitude, Δλ = (lon 2 – lon1); difference of longitude,
R = 6371 km; radius of the Earth, d is the distance computed between Ground Control
Station (GCS) and the UAV.
The BMP180 module [11] offers data regarding the atmospheric pressure, used to
compute the H altitude at which UAV antenna is positioned during the flight. The basic
formula for absolute altitude is:
⎛ ( ) 1 ⎞
⎜ p 5.255 ⎟
H altitude = 44330 ∗ ⎜1 − ⎟ (4)
⎜ p0 ⎟
⎝ ⎠
where p is the measured pressure by the BMP180 sensor and p0 is the pressure at sea
level e.g. 1013.25 hPa.
Based on d and H values, the controller determines pitch angle for having the LOS
between UAV and GCS. Pitch angle is converted in Pulse-Width Modulation (PWM)
to command the standard servo. At the same time with pitch angle command, values
from the L3GD20 module are used to compute the correction of the angular position of
the UAV ATS, but also related to the aircraft pitch.
4.3 The Calculation of the Position of the Antenna Azimuth Angle Versus UAV
ATS
To compute the azimuthal orientation toward magnetic north, between the GCS and
UAV ATS, it is necessary to know the azimuth angle which is given by the LSM303
module and is recorded by the UAV ATS controller. An initial calibration has to be
made in order to read the takeoff azimuth angle. The LSM303 module must be level to
the earth’s surface. Tilt compensation circuits and techniques can be used to normalize
the magnetometer reading to correct all the influences especially magnetic interferences
and the effect of the earth’s field [12].
The x and y component of the earth’s magnetic field, that is, the directions planar
with the earth’s surface, are used for the computation of α, the magnetic compass heading
angle. The magnetometers x and y readings are used in the set of equations:
zamfira@unitbv.ro
Improving Communication Between Unmanned Aerial Vehicles 539
angle and antenna azimuth angle, enter as data in an average subroutine, in order to
minimize tracking errors. As a result, averaged values are converted as PWM signals to
command CRS.
5 Conclusions
The paper presents a new approach to improve the communication between the pan-
tilt UAV ATS module and the pan-tilt ATS module from GCS, using only commer‐
cial antennas, to improve the quality and the distance they can transmit safety the
information.
The ATS is controlled by a Companion-microcontroller development board, being
responsible to maintain the LOS between GCS and UAV.
In the next future, different types of communication interfaces and antennas will be
tested.
Acknowledgment. This paper was realized within the Partnerships Programme in priority
domains-PN-II, which runs with the financial support of MEN-UEFISCDI, Project no. PN-II-PT-
PCCA-2013-4-1629 and it was also supported by a grant of the Romanian National Authority for
Scientific Research and Innovation, CNCS/CCCDI – UEFISCDI, project number PN-III-P2-2.1-
BG-2016-0132, within PNCDI III.
References
1. Zeng, Y., Zhang, R., Lim, T.J.: Wireless communications with unmanned aerial vehicles:
opportunities and challenges. In: IEEE Communications Magazine, vol. 54, Issue no. 5, pp.
36–42 (2016)
2. Çuhadar, I., Dursun, M.: Unmanned air vehicle system’s data links. J. Autom. Control Eng.
4(3), 189–193 (2016)
3. Li, B., Jiang, Y., Sun, J., Cai, L., Wen, C.-Y.: Development and testing of a two-UAV
communication relay system. Sensors 16(10), 1696 (2016). doi:10.3390/s16101696
4. Barton, J.D.: Fundamentals of small unmanned aircraft flight. Johns Hopkins Apl. Tech.
Digest 31(2), 132–149 (2012)
5. Thenkabail, P.S., Lyon, J.G., Huete, A.: Hyperspectral Remote Sensing of Vegetation. CRC
Press, New York (2016). ISBN 1439845387, 9781439845387
6. Ocean Optics Spectrometers, STS Series. http://oceanoptics.com/product-category/sts-
series/. Accessed 20 September 2016
7. https://www.adafruit.com/product/1604. Accessed 4 Apr 2016
8. https://www.u-blox.com/en/product/neo-m8-series. Accessed 9 Apr 2016
9. http://ardupilot.org/planner/docs/mission-planner-overview.html. Accessed 4 May 2015
10. Alexander, J.: A rhumb way to go. Math. Mag. 77(5), 349–356 (2004)
11. https://www.bosch-sensortec.com/bst/products/all_products/bmp180. BST-BMP180-
DS000-09. Pdf. Accessed 4 May 2016
12. Bingaman, A.N.: Tilt-Compensated Magnetic Field Sensor, Master of Science in Electrical
Engineering Thesis, Blacksburg, Virginia, 13 May 2010
zamfira@unitbv.ro
Remote RF Testing Using Software
Defined Radio
1 Introduction
The wireless revolution has paralleled the Internet revolution and continues to
act as an economic driver. One of the economic characteristics of the Internet
economy is the scalability of software start ups. Venture capital has gravitated
towards start up organizations that have low cost of entry and fast time to
market. In contrast, hardware start ups face big capital investments and long
time to market horizons. This inhibits the flows of capital to these start up, and
ultimately places obstacles in innovation. While there are many costs associated
with bringing hardware to market, a significant one is test and evaluation. Radio
frequency (RF) test equipment is expensive and special purpose. Software defined
radio (SDR) [8] moves the implementation of radio functions from the analog,
component element domain to the software domain running on general purpose
processors. This same technology is being used in test equipment and has the
potential to virtualize RF testing.
The University of Utah runs Emulab, a large scale testbed including low cost
PCs, WiFi access points, and importantly, USRPs, the Universal Software Radio
Peripheral [7]. Similarly, Rutgers University runs the ORBIT test bed using
software defined radios for evaluating wireless networks [9]. These systems allow
remote users to configure and run tests on a large number of network nodes. The
focus is on the network protocol stack and examining different routing and media
The authors would like to thank the Manhattan College Summer Research Scholars
program for supporting this work.
c Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6 50
zamfira@unitbv.ro
Remote RF Testing Using Software Defined Radio 541
access control (MAC) layers. Because they are based on software defined radio,
there is some possibility to run experiments at the physical (PHY) layer. The
modulation format can be implemented in GNU Radio running on commodity
PCs [1].
The intention of this research is to push the virtualization line down to the
transmitter and receiver line up. The line up generally refers to digital up con-
version, crest factor reduction, digital predistortion, digital to analog conversion,
analog up conversion, and amplification for the transmitter. On the receiver side,
the line up starts with a preselector filter, low noise amplifier, analog down con-
version, analog to digital conversion, and digital down conversion. Receivers also
include channel estimation and equalization and the demodulation and decision
process. RF testing is expensive due to the specialized equipment and compo-
nents required. Organizations from start ups to university laboratories may not
have the capital budget to afford this equipment. An RF laboratory as a service
model could change the economic model of the hardware startup.
Ultimately, it is important to evaluate wireless systems in realistic channel
conditions. Propagation emulators are also very expensive and only provide an
estimate of real conditions. Being able to host a wireless prototype system in an
actual field scenario could be very effective. One particular relevant use case is
the Industrial Internet of Things (IIoT). This field is characterized by severe mul-
tipath distortion and interference environment and has a significant impact on
the feasibility of any given solution. Indoor wireless applications suffer from the
same issues as IIoT and are difficult to model accurately. Labor intensive walk
tests are required to accurately map coverage patterns. These use a backpack
full of consumer phones in a logging mode. This only produces high level data.
Gathering more fundamental coverage data today requires bulky test equipment.
They often are conducted by an engineer that maps the signal strength versus
location and interferers. Other environments, especially urban developments,
require a three dimensional coverage map. Remotely accessed surveying equip-
ment could provide this useful feature, perhaps even a drone mounted system.
All of these applications can be satisfied by a software defined radio platform.
In particular, we propose a system on chip that mates to an integrated chip
with data conversion and RF upconversion. An example is the Xilinx Zynq
processor with dual-core Cortex A9 ARM processors and Virtex 7 FPGA fabric
as hosted on the ZEDBoard and the Analog Devices AD9361 Integrated RF Agile
Transceiver. This system allows for broad frequency coverage, diverse filtering
options, FPGA options, real-time SDR implementation on one core, and test
executive software running in Linux on the other core. This paper outlines the
construction of the software defined radio signal chain, software control program,
and depicts a sample of its potential usage. Furthermore, this paper discusses
the future potential usage of this system to address pressing issues regarding
high frequency communications.
zamfira@unitbv.ro
542 S. Miller and B. Horine
zamfira@unitbv.ro
Remote RF Testing Using Software Defined Radio 543
filtering and the power amplifier chain. In particular, we are investigating the lin-
earization of RF power amplifiers. We implement the digital portions of this line-
up in the ZedBoard. To demonstrate the feasibility and explore the performance
limits, we tested a Mini-Circuits GALI-19+ power amplifier [3] (7 GHz–10 mW)
on a TB-409-19+ evaluation board [4] and a ZHL-5W-2G-S+ [5] (800–2000 MHz
5 W). The entire test setup can be seen below labeled Fig. 2. The signal that was
fed back into the SDRs receiver was also tapped off to a spectrum analyzer to pro-
vide a real-time reference signal power measurement. The GALI-19+ is used as a
driver amplifier. Its characteristics are illustrated in Fig. 3.
zamfira@unitbv.ro
544 S. Miller and B. Horine
More output power is required to drive PAs into saturation. This is the role
of the high gain ZHL-5W-2G-S+ amplifier. The IO characteristics of this driver
amplifier can be seen in Fig. 4. Furthermore, the gain characteristics can be seen
in Fig. 5.
4 Results
The FMCOMMS2 is considered to be linear from 800 MHz 6 GHz within its
datasheet [6]. A 20 MHz wide LTE waveform is sent through the system without
an amplifier in the loop (Fig. 6) and with the GALI 2 GHz PA driven into satura-
tion Fig. 7. The loopback mode shows a clean rolloff from the signal to the out of
band region. The saturated amplifier version exhibits a shoulder indicative of third
order intermodulation of the various subcarriers in the LTE waveform. While the
in-band distortion cannot be so easily observed, a receiver algorithm can be incor-
porated into the ZedBoard processor and the constellation error easily calculated.
Unfiltered 20MHz LTE Signal Amplified and Attenuated 20 MHz LTE Signal
160 150
140
140
130
120 120
Response (dB)
Response (dB)
100 110
100
80
90
60 80
70
40
60
20 50
−20 −15 −10 −5 0 5 10 15 20 −20 −15 −10 −5 0 5 10 15 20
Frequency (MHz) Frequency (MHz)
zamfira@unitbv.ro
Remote RF Testing Using Software Defined Radio 545
References
1. http://gnuradio.org/
2. https://github.com/analogdevicesinc/libiio/blob/master/examples/ad9361-
iiostream.c
3. http://www.minicircuits.com/pdfs/GALI-19+.pdf
4. http://www.minicircuits.com/pcb/WTB-409-19+ P02.pdf
5. http://www.minicircuits.com/pdfs/ZHL-5W-2G+.pdf
6. http://www.analog.com/en/design-center/evaluation-hardware-and-software/
evaluation-boards-kits/eval-ad-fmcomms2.html#eb-overview
7. Hibler, M., Ricci, R., Stroller, L., Duerig, J., Guruprasad, S., Stack, T., Webb, K.,
Lepreau, J.: Large-scale virtualization in the Emulab network testbed. In: Proceed-
ings of the 2008 USENIX Annual Technical Conference, pp. 113–128 (2008)
8. Mitola, J.: The software radio architecture. IEEE Commun. Mag. 33(5), 26–38
(1995)
9. Ott, M., Seskar, I., Siraccusa, R., Singh, M.: Orbit test software architecture: sup-
porting experiments as a service. In: IEEE Tridentcom (2005)
zamfira@unitbv.ro
Remote Control of Large Manufacturing Plants Using
Core Elements of Industry 4.0
1
University of Technology, Arts and Sciences, Cologne, Germany
hasan.smajic@th-koeln.de
2
Schneider Electric GmbH, Ratingen, Germany
Abstract. The most big manufacturing plants such as large transfer line, large
packaging machines, steel production line or plants for food and beverage are
built for a long-live deployment and are usually energy-intensive processes. A
redesign of mechanical parts is for a long usage term (20 till 30 years) not required.
But upgrading of automation parts with more performance is a needful continu‐
ously process. An efficiently implementation of modern automation and IT tech‐
nologies by currently engineering tolls is not the state of the art. On the beginning
of this paper a model concept will be introduced, showing how the partition of
the large plants can be performed in small collaborating parts with less
complexity. The next model describes the distributed and allocated hardware and
software components to determined plant parts. The developed distributed control
system based on smart Ethernet nodes, which allows remote control and mainte‐
nance according elements of industry 4.0.
1 Introduction
The very fast advances in digital hardware and communication technology have led to
the development of a new generation of microcontroller-based control systems, known
as distributed control systems [1]. These are characterized by a digital fieldbus commu‐
nication, small decentralized control systems instead of one centralized system and a
microcontroller-based distributed intelligence, which uses more and more onboard
computing power in smart sensors and actuators.
The concepts of different bus-layers for fieldbus, as they are realized at the current
development status of complex production systems, have reached their limit of capability
and efficient usability in regard to engineering, maintainability and extensibility [2]. By
using existing and implementing new functions and interfaces in intelligent field devices,
a control system for production systems is to be developed. Unlike most previous
approaches, it must based on continuous communication and functions for remote program‐
ming and remote management of plug&play-able field devices.
The first goal of this project is to find out, what impact standard implementation
methods for automation devices (interconnectible, interoperable, interchangeable and
zamfira@unitbv.ro
Remote Control of Large Manufacturing Plants 547
The building of automation architecture is a complex issue. For four different technology
levels (mechanical parts of plant, hardware, communication system and software) the
physical and logical models have been described with object oriented methods according
UML Notation (Fig. 1). Based on developed models the distributed control system
consider decentralization of tree dimension. Not only the hardware and data processing
is distributed, but also the PLC control algorithm and data base. Each smart device on
the lowest level has its own PLC task and his own part of data base, connected with
industrial Ethernet.
The first step for plant modeling is identification of mechanical parts and their clas‐
sification to function groups and function units. After hierarchically decomposition of
plant objects in sublevels follows the assignment of functionality and description of
relationships and interfaces.
zamfira@unitbv.ro
548 H. Smajic and N. Wessel
The physical architecture describes the mapping of the hardware elements to esti‐
mated plant objects. That means the distribution of the technical objects to the Control-
Nodes, which are connected with the industrial Ethernet. These technical objects, the
sensors and actuators as well as markers as virtual objects, are in an object oriented
viewpoint instances of the appropriate classes Sensor, Actuator and Marker. Also they
are assigned as embedded objects to an instance of a generalized class “Bus-Node”.
The logical architecture of the distributed automation system will be displayed in a
class-structure diagram. The model shows the static structures between classes and
instances as well as the appropriate relationships and serves as the basis for the gener‐
ation of node-specific software-modules. The classes Sensor, Actuator and Marker are
derived by inheritance from the two abstract classes Producer and Consumer. As derived
classes they are a specialization of this basis. A sensor takes over the role of a message
producer in this control concept with horizontal communication in the lowest automation
layer. The class Sensor is a specialization of Producer, that as basis class contains the
attributes “identifier”, “priority” and “blocking time” of the accordant communication
object and a method for message creation in the case of an event. On the other side an
actuator depends on information from specific sensors and Markers in the system.
For the control-concept a comprehensive model according UML-notation (Unified
Modeling Language) has been created, that describes the communication, the distributed
architecture and the PLC-application, which is to be distributed. The described tree
models are the basis for the later software development process and integration of the
smart devices.
The new architecture for remote automation engineering has been developed using the
intelligence of smart devises with the aim to substitute control task of a centralized PLC
(Fig. 2). Each smart device on the lowest level has its own PLC task and its own part of
the process data base. They process communication between devices is over industrial
Ethernet (Modbus TCP/IP). The following type of devices are used:
• Sensors, detectors, encoders and RFID switches for data detection
• Programmable logic control (PLC) for data processing
• Human Machine Interface
• Velocity speed driver and Motion driver
• Smart meter for energy efficiency
• Ethernet Network with Modbus TCP/IP
• Programming and SCADA software
The gateway server provides all central services including administration tasks,
routing to the appropriate automation devices and security mechanism. The server also
runs an SQL server with modules for authentication and logging. There are features for
global data access, data for energy efficiency and maintains of the whole process. For a
high scalability the Hyper-V services, which are integrated into Windows 2008 Data‐
center, are used on the gateway machine. The root machine consists of two VLAN
zamfira@unitbv.ro
Remote Control of Large Manufacturing Plants 549
switches for network data transfer between the virtual clients. Each VLAN is connected
with one physical network adapter. The first VLAN is connected with the internet and
is used by the thread management gateway server who links to routing and RAS
(including VPN connections). Another roll of this virtual server machine is to establish
a web server and secured internet access from inside of the laboratory. The second
VLAN is connected with the virtual machine pool. In order that the virtual machines
can connect to the PLCs and the HMIs, the VLAN is connected to the physical network
adapter, which is connected to the internal laboratory network.
For an integrated authentication service, the active directory services are used. In
this directory we integrated a container for external users. Only administrators and users
in this group are allowed to connect via VPN. External standard users have some more
restrictions which are realized due to adapted group policy objects.
For remote access with a Web Client the user needs just standard Web-Browser. The
connection to devices is established over external URL Gateway Server and the name
of the virtual machine, before username and password must be given. The connection
to the virtual working desktop can be realized in a simple way. In the first step a VPN
connection has to be done. For establishing the VPN connection the user only need the
VPN client supporting SSTL, L2TP, PPTP or IPsec. The easiest way is to use the already
integrated client of the most frequently used operating systems, such as Microsoft
Windows, Mac OS X, iOS and many Linux distributions. After connection the user is
able to start a remote access app which uses the RDP protocol. On actual Microsoft
zamfira@unitbv.ro
550 H. Smajic and N. Wessel
Windows machines (XP, or later) the tool is already integrated. OS X users can download
a small application from the Microsoft web site. If the application is running, users only
need to enter username, password and the name of the virtual machine.
After establishing the connection to the virtual machine the user can open applica‐
tions for writing, download and testing applications to the PLC, HMI and Drives.
For Design and implementation of developed models for distributed control system
(capture 2) the application software UAG from Schneider Electric is used. It allows the
design the complete large plants including mechanical parts, control objects, commu‐
nication and logic.
The Plant design for mechanical objects is compliant to VDI3260 standard. Object
oriented model is used for design of technological object like motors, valves, pumps.
Those object are implemented via FDT international standard as cyber physical systems
(CPS) into application. UAG provides first the single entry based on the Smart Control
object Devices (ScoD) and then generates the applications providing synchronized,
consistent databases. In other words the tools reflect the image of a single database
(Fig. 3).
The ScoDs as CPS are applied within the process context by parameterization. It
propagates the information incrementally, changing only the affected/modified parts.
Incremental generation means that UAG adds incrementally information to the PLC and
HMI applications. Manual modifications within the PLC/HMI are not changed by an
incremental generation.
In a case that an upgrading of automation parts occurs, the software controls for the
generated system the global resource mapping, the PLC application code with config‐
uration and variables and for the HMI application the variables, symbols, archive and
access information and more.
zamfira@unitbv.ro
Remote Control of Large Manufacturing Plants 551
On this way by all object changes the software can manage the actually physical and
logical model of the plant. This feature can significantly reduce the new project cycle
times and save life cycle cost.
The described automation architecture has been already tested in the model factory of
university Cologne. The remote access services are running on an application for high
rack warehouse very stable with a high reliability. The remote control system is available
from outside primarily for our project partner. The first experience has shown that
developed architecture with internet and web technologies can be used for automation
engineering in following topics: remote programming of PLC, HMI, Drives, Motion,
Networks and remote monitoring.
It has been proven that implemented Industry 4.0 elements give high advantages for
monitoring of overall production resources efficiency. With just a standard Web-
Browser big energy data can be shown vie PLC access. Typically those data can be used
for monitoring of energy consumption, process part performing, benchmark perform‐
ance and reporting. We are currently developing data interface to MES and ERP Systems
(SAP).
References
1. Meier, H., Smajic, H.: Distributed Control System Based on Smart Sensors and Smart Actuator,
Mechatronics & Robotics 2004, IEEE Industrial Electronics Society, APS - European Centre
for Mechatronics Aachen, Germany, pp. 103–105, 15 September 2004. ISBN 3-938153-30-X
2. Kühnle, H., Lorenz, K., Klostermeyer, A.: Neue Wege in der Fabrikautomatisierung. In:
Werkstatttechnik, Heft 3, pp. 138–141 (2010)
3. Meier, H., Smajic, H., Faller, C.: Webbased automation for complex manufacturing systems,
machine tools and factories of the knowledge. Mach. Eng. 4(1–2), 52–59. ISSN 1642-656
(2004)
4. Falkman, P., Helander, E., Andersson, M.: Automatic generation: a way of ensuring PLC and
HMI standards. In: IEEE 16th Conference on Emerging Technologies and Factory Automation,
ETFA 2011, Toulouse, 5–9 September 2011
5. Smajic, H., Faller, C.: Remote laboratory for education in automation engineering. In: IEEE,
CTI global Engineering Education Conference, Istanbul, Turkey (2014)
6. Unity Application Generator (UAG) 3.3 SP4, Extended User Manual (2016). http://www.schneider-
electric.com/en/download/document/33003669K01000/
7. Umsetzungsempfehlungen für das Zukunftsprojekt Industrie 4.0, Berlin, 2. Oktober 2012. https://
www.bmbf.de/files/Umsetzungsempfehlungen_Industrie4_0.pdf
zamfira@unitbv.ro
Games Engineering
zamfira@unitbv.ro
Dinner Talk: A Language Learning Game
Designed for the Interactive Table
1 Introduction
Technologically, the game Dinner Talk is based on meme media technology [1]. Now,
taking the perspective of games science, Dinner Talk is a dynamic placement game.
But practically, Dinner Talk is a Game-Based Learning approach to support learning of
a foreign language using the Interactive Table.
When people play (digital) games, they have an experience. It is this unique
experience that is in focus. But, as familiar as we are with experiences, they are very
hard to describe. You cannot really see them, touch them, or hold them - you cannot
even really share them. No two people can have identical experiences of the same thing
- each person’s experience of something is completely unique. The (digital) game itself
is not the experience, it is the enabler for the experience. What we can do, is to create
interesting artefacts (e.g. rule sets, computer programs) that are likely to create certain
kinds of experiences when a player interacts with them [10].
This is what we do while designing games for the Interactive Table. The digital
game Dinner Talk running on the Interactive Table, is simply a tool to generate unique
experiences in learning languages.
The potential of learning with digital games is far from being exhausted, and certainly
not at school and in further education. In a digital game, real people act in a virtual
world. Actions with real content, e.g. doing math, managing resources, solving com-
binatorial problems, understanding speech and acting accordingly form the basis for
real learning and training. Dinner Talk is the name of a de facto infinite family of
digital games that support learning German as a foreign language. Dinner Talk is a
so-called placement game [1] - in the category of combinatorial games one of the
easiest subcategories. It consists of a simple game board and playing pieces, which are
labelled with German text. If the pieces are placed in such a way that texts on
neighboring pieces match, the players will receive gratification in the form of points on
a High Score List.
zamfira@unitbv.ro
Dinner Talk: A Language Learning Game Designed 557
It is permitted for the dynamic variation of these games that set pieces can be
re-arranged to allow for exploration, especially competitive, and explorative learning.
Dinner Talk provokes the confrontation with the content of phrasings in the German
language and is almost infinitely scalable with respect of linguistic complexity. It has a
large number of parameters. Every instantiation of parameters leads to a particular
game of the Dinner Talk family [1].
zamfira@unitbv.ro
558 J. Schuldt et al.
Fig. 1. First version of the core variant of the game family Dinner Talk played on the Interactive
Table.
zamfira@unitbv.ro
Dinner Talk: A Language Learning Game Designed 559
Acknowledgement. Part of the authors work, has been supported by the German Federal
Ministry for Education and Research (BMBF) within the joint project Webble TAG under grant
no. 03WKP41D (Webble TaT).
References
1. Jantke, K.P., Arnold, O., Bosecker, T.: Exploratory game play to support language learning:
dinner talk. In: 8th International Conference on Computer Supported Education, pp. 161–
166 (2016)
2. Arnold, O., Bosecker, T., Hume, T., Jantke, K.P. Response to the challenging refugee influx:
a potentially infinite family of serious games for learning of foreign languages playfully. In:
Proceedings of the e-Society, Vilamoura, Portugal, 9–11 April 2016
zamfira@unitbv.ro
560 J. Schuldt et al.
3. Jantke, K.P., Bosecker, T.: Exploratives spielerisches Lernen von Fremdsprachen. magazin.
digitale.schule (2015)
4. Hammond, M.: What is an affordance and can it help us understand the use of ICT in
education? In: Education and Information Technologies, vol. 15, Issue no. 3, pp. 205–217
(2010)
5. https://github.com/cubbles. Accessed 17 Nov 2016
6. Kuwahara, M.N., Tanaka, Y.: Webbles: programmable and customizable meme media
objects in a knowledge federation framework environment on the web. In: Karabeg, D.,
Park, J. (eds.) Second International Workshop on Knowledge Federation, Dubrovnik,
Croatia, 3–6 October 2010
7. Tanaka, Y.: Meme Media and Meme Market Architectures: Knowledge Media for Editing,
Distributing and Managing Intellectual Resources. IEEE Press & Wiley-Interscience,
Hoboken (2003)
8. Dawkins, R.: The Selfish Gene. Oxford University Press, Oxford (1976)
9. Fujima, J., Jantke, K.P.: The potential of the direct execution paradigm: toward the
exploitation of media technologies for exploratory learning of abstract content. In: Urban, B.,
Müsebeck, P. (eds.) eLearning Baltics 2012: Proceedings of the 5th International eLBa
Science Conference, pp. 33–42. Fraunhofer Verlag (2012)
10. Fullerton, T.: Game Design Workshop: A Playcentric Approach to Creating Innovative
Games, 3rd edn. Taylor & Francis Ltd., Boca Raton (2013)
11. Buisine, S., Besacier, G., Aoussat, A., Vernier, F.: How do interactive tabletop systems
influence collaboration? In: Computers in Human Behavior, vol. 28, issue no. 1, pp 49–59
(2012)
12. Rogers, Y., Lindley, S.: Collaborating Around Large Interactive Displays: Which Way is
Best to Meet? Interact. Comput. 16(6), 1133–1152 (2004)
13. Shneiderman, B.: Touch screens now offer compelling uses. IEEE Softw. 8(2), 93–94, 107
(1991)
zamfira@unitbv.ro
The Experimento Game: Enhancing a Players’
Learning Experience by Embedding Moral
Dilemmas in Serious Gaming Modules
1 Introduction
Learning with digital games will necessarily constitute a basic form for teaching and
learning. It must be remembered that digital games conquered the leisure market and
the market for entertainment media. One cannot ignore Digital Game-Based Learning
(DGBL) [1]. Therefore it is necessary to understand how to implement it and for which
purposes it is appropriate. A Digital Game-Based Learning approach for Experimento
is going to support learners to improve their critical thinking skills and the ability to
carry out a change of perspective. Playful modules, within the Siemens Stiftung’s
OER-Platform, encourage the learner to critically reflect upon existing knowledge,
independently develop relevant new questions and seek for answers. Thus, the
Experimento Game boosts the learners’ self-confidence. In addition, it provides them
with the necessary methodologies to independently find answers for new questions and
deduct solutions to real-life problems.
It has been established that children typically focus on action, navigation and
inter-action in their use of digital media. It is important for children to be able to have a
recognizable influence on games when the completion of the game depends on their
actions, strategies, choices and decisions.
zamfira@unitbv.ro
The Experimento Game: Enhancing a Players’ Learning Experience 563
zamfira@unitbv.ro
564 J. Schuldt et al.
zamfira@unitbv.ro
The Experimento Game: Enhancing a Players’ Learning Experience 565
Fig. 1. Extract from the Storyboard Editor. On the left is the storyboard that the editor creates,
on the right is a transformed storyboard that the game engine can read.
zamfira@unitbv.ro
566 J. Schuldt et al.
Testing the game’s flow with potential users has not been executed as of yet, but a
paper prototype is in preparation. Once the first digital prototype has been developed,
user testing will expand: collecting data via questionnaires and guided testing. After
evaluating said data the game’s next iteration will be initialized.
The specifications for the Experimento Game state that the design should have simple
2D graphics and be an abstraction of the real world, as the environment should not
resemble a certain region.
In the beginning of the design process a survey with 204 participants showed that a
total abstraction of environment and character is more suitable for a younger audience
than the targeted 11–13 year olds. Out of a collection of 36 character-designs, those
with apparent features, such as eyes and mouths, had a more positive feedback;
showing the importance of being able to read a characters emotion to create empathy.
To compromise on an art style, the “Consensus mapping” process was modified.
With the general specifications in mind, each artist searches for environment and
character inspiration and select ten pictures to present to the team. After giving and
receiving feedback, the whole group decides on three pictures that are guidance for the
environment and character design. With this guidance the artists create own interpre-
tations, setting an art style for the game and the characters that serve as a template for
future designs.
As a result of the survey, the characters received facial features thus giving them
humanoid traits. In order to prevent association with a specific cultural area,
non-human colours and the omission of noses were chosen (Fig. 3). The players will be
able to choose between three characters – a female, male or a gender-neutral persona.
zamfira@unitbv.ro
The Experimento Game: Enhancing a Players’ Learning Experience 567
Game development is a prospering industry that is set to grow rapidly in the near
future. Therefore theoretical concepts and practical skills are required for designing
games and software engineering. The Storyboard Interpretation Technology (SIT) is a
feature to support the development of games, especially for educational contexts.
According to the above-cited sources, storyboards are digital objects that can be stored
in a database. Digital systems, intended to implement the anticipated processes rep-
resented in a storyboard, may read the digital document. When humans interact with
the system, the system interprets the storyboard. The Experimento Game is based on
the concepts of SIT. A particular advantage of this technology is the simplicity of
modifying a storyboard in use.
The gaming module aims at supporting the development of a child’s critical
reflection, taking into account that children must be encouraged to understand the deeper
meaning of a problem. Using that knowledge, they are incited to analyze and solve
problems in a scientific context. Furthermore, these experiences do not remain abstract
in that they are directly linked to their everyday lives and everyday problems. However,
the game benefits children most by initiating positive behavioral changes through the
teaching of critical thinking. This is achieved by inviting children to deal with and
analyze Moral Dilemma Situations, which they are confronted with in the game. The
virtual world of the Experimento Game is, metaphorically speaking, the wrapping of the
actual learning content. The more attractive the wrapping is, the more likely it is that a
task is accepted. An attractive wrapping can induce students to learn, not only voluntary,
but also with more intensity and more frequent repetition, while at the same time not
perceiving learning as a burden. The evaluation of the game’s acceptance exceeds this
papers scope. A separate report upon the systematic evaluation will follow.
zamfira@unitbv.ro
568 J. Schuldt et al.
Acknowledgement. The authors gratefully acknowledge the fruitful collaboration with their
customer, Siemens Stiftung. As a non-profit corporate foundation, Siemens Stiftung promotes
sustainable social development, which is crucially dependent on access to basic services,
high-quality education, and an understanding of culture. To this effect, the Foundation’s project
work supports people in taking the initiative to responsibly address current challenges. Together
with partners, Siemens Stiftung develops and implements solutions and programs to support this
effort, with technological and social innovation playing a central role. The actions of Siemens
Stiftung are impact-oriented and conducted in a transparent manner www.siemens-stiftung.org.
References
1. Prensky, M.: Digital Game-Based Learning. Paragon House, St. Paul (2007)
2. Jantke, K.P.: Digital games that teach: a critical analysis. TUI IfMK, Diskussionsbeiträge
22, August 2006
3. Gee, J.P.: What Video Fames Have to Teach us About Learning and Literacy. Palgrave
Macmillan, New York (2007)
4. Söbke, H., Reichelt, M.: “Rat(t)en in der Lehre” - Über die Spiel(un)lust unserer
Studierender am Beispiel digitaler Apps, Teaching Trends 2016 Digitalisierung in der
Hochschule: Mehr Vielfalt in der Lehre. Münster: Waxmann
5. Krebs, J.: Moral Dilemmas in Serious Games. In: 2013 International Conference on
Advanced ICT (2013)
6. Sicart, M.: Moral dilemmas in computer games. Des. Issues 29(3), 28–37 (2013)
7. Ritterfeld, U., Cody, M., Vorderer, P. (eds.): Serious Games: Mechanisms and Effects.
Routledge, New York (2009)
8. Krebs, J., Jantke, K.P.: Methods and technologies for wrapping - educational theory into
serious games. In: 6th International Conference on Computer Supported Education, pp. 497–
502 (2014)
9. Jantke, K.P., Knauf, R., Gonzalez, A.G.: Storyboarding for playful learning. In: World
Conference on e-Learning in Corporate, Government, Healthcare and Higher Education,
e-Learn 2006, AACE, pp. 3174–3182 (2006)
10. Jantke, K.P., Knauf, R.: Didactic design through storyboarding: standard concepts for
standard tools. In: Proceedings of the 4th International Symposium on Information and
Communication Technologies (ISICT) Workshop on Dissemination of e-Learning Tech-
nologies and Applications, Cape Town, South Africa, January 2005, pp. 20–25. ACM Press,
New York (2005)
11. Fernandez-Vara and, C., Osterwil, S.: The Key to Adventure Game Design: Insight and
Sense-making. http://meaningfulplay.msu.edu/2010/
12. Cornell University: Storyboarding in Game Design. http://www.cs.cornell.edu/courses/
cs3152/2013sp/labs/design1/. Accessed 16 Nov 2016
13. Jantke, K.P., Knauf, R.: Taxonomic concepts for storyboarding digital games for learning in
context. In: 4th International Conference on Computer Supported Education 2012, pp. 401–
409 (2012)
14. Arnold, S., Fujima, J., Jantke, K.P.: Storyboarding serious games for large-scale training
applications. In: 5th International Conference on Computer Supported Education 2013,
pp. 651–655 (2013)
zamfira@unitbv.ro
The Experimento Game: Enhancing a Players’ Learning Experience 569
15. Norman, D., Nielsen, J.: The Definition of User Experience. https://www.nngroup.com/
articles/definition-user-experience/. Accessed 21 Nov 2016
16. Bevan, N.: What is the difference between the purpose of usability and user experience
evaluation methods? In: UXEM 2009 Workshop, INTERACT 2009, Uppsala, Sweden
(2009)
17. Hodent, C.: Developing UX Practices at Epic Games, GDC EU (2014). http://www.
gdcvault.com/play/1020934/Developing-UX-Practices-at-Epic. Accessed 21 Nov 2016
zamfira@unitbv.ro
The Finite State Trading Game: Developing
a Serious Game to Teach the Application
of Finite State Machines in a Stock
Trading Scenario
Abstract. In this paper a new methodology to teach the topic Finite State
Machines to upper vocational school students is proposed. A Serious Game
solution was created consisting of nine learning objectives split into categories
about the basics of Finite State Machines, the parallels between Finite State
Machines and stock trading and the application of Finite State Machines in order
to construct Artificial Intelligence. This paper focuses on the existing parallels
between Finite State Machines and the concepts of automated stock trading. The
learning objectives were determined using Bloom’s Taxonomy and imple-
mented into the Serious Game “The Finite State Trading Game” (FSTG). In this
turn-based trading game, the user strives to beat a Non-Player Character by
skillfully trading shares at various difficulty levels. In order to evaluate the
Serious Game approach, a pre-test and post-test situation was performed with
students of a local upper vocational school class at the Technical University of
Munich. The analysis of the results showed major improvements of the students’
knowledge about Finite State Machines for every tested statement. Given the
success of this test setting, FSTG appears to be a promising solution to be used
to support or even substitute traditional ways of teaching.
1 Introduction
Nowadays computer games are played all over the world: For example, in the United
States 72% of the households play video games, with a staggering number of twelve
years adult gamer have been playing in average in their lives so far [1]. But: Do they
know, that all these games – as well as many other applications, like traffic lights – are
based on so-called Finite State Machines (FSMs)? FSMs represent the logic of these
actions. The National Institute for Standardization and Technology defines an FSM as a
© Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6_54
zamfira@unitbv.ro
The Finite State Trading Game 571
“model of computation consisting of a set of states, a start state, an input alphabet, and
a transition function that maps input symbols and current states to a next state” [2].
FSMs represent a core concept regarding automation. Automation is part of the
curriculum at upper vocational school (UVS) in Bavaria, Germany [3], as well as at
IT-related degree courses, e.g. in the Bachelor’s Program of the Technical University of
Munich (TUM) [4, 5].
The curriculum for 12th graders at UVS in Bavaria, Germany, [3] covers a special
class dedicated to engineering, called ‘Technology and Computer Science’. This class
comprises the modules basics of modern programming languages, programming styles
and data structures, and object-oriented programming. They are the basis for the next
modules capstone project and systems and processes. The capstone project aims at
understanding, analyzing and evaluating a complex technical system like a power sta-
tion or a car, as well as developing and implementing problem-solving strategies which
are especially suited for this kind of technical system. With systems and processes the
students are expected to gain a better understanding of technical systems and processes.
They should be able to identify the essential elements and processes of a technical
system, and describe their contribution to its overall functionality. The students learn to
understand and apply systems as well as to analyze and evaluate them [6].
FSMs represent a core concept regarding automation. Automation is also part of the
curriculum at UVS in Bavaria, Germany [3]. Furthermore, using the so-called State
Pattern, an FSM representation can be developed for almost every IT-based application
ranging from control systems to business simulations and even computer games [7].
These State Patterns are the optimum architecture for the correct implementation of an
FSM. Because of the ubiquity of FSMs in reality, teaching FSMs to UVS students
seems to be a promising and valuable approach to gain a better understanding of real
world scenarios. Yet – which application would be best suited to meet the learning of
UVS students? Traffic lights build on a small number of states and are present
everywhere to control traffic in our daily life. A more challenging FSM application
would be a stock trading system and its processes of buying and selling shares [6].
A share is traded if the offered share price is equal or lower than the limit price and a
potential buyer has the money to invest in the share.
Using educational software is a common trend [8]. Serious Games promise to
support self-regulated personalized learning (SRPL) [9]. Two examples shall clarify
how playing Serious Games may encourage SRPL. As a first example, [10] describes a
learning scenario for UVS students in the context of the Pupils´ Academy of Serious
Gaming. An important goal of the academy is to reduce the dropout rate of students by
strengthening their study skills even before they start studying at university. The
ERPsimTM Distribution business game is used to improve activities related to e.g. time
management or teamwork. Secondly [11] presents the learning platform Learn@WU
aiming at better ways for self-learning. Furthermore, the trend towards Serious Gaming
is supported by the increasing availability of computers in classrooms. Thus, sum-
marizing the previous considerations, a Serious Game approach based on an FSM
example seems to be well suited for learning about systems and processes in ‘Tech-
nology and Computer Sciences’ classes in Bavarian UVS.
The UVS students are young people who have realized at secondary school level: “I
learn best in an application-oriented way” [10]. Therefore, we developed a stock
zamfira@unitbv.ro
572 M. Utesch et al.
trading game named ‘The Finite State Trading Game’ (FSTG) [6]. With FSTG the
learners strive to skillfully trade shares in order to beat a Non-Player Character
(NPC) through higher profits. The scenario of automated stock trading was used for
two reasons: Firstly, stock trading is a challenging FSM example out of everyday life.
Secondly, automated stock trading is strongly related to the science of Information
Systems. As a precondition, the IT-related knowledge is very heterogeneous at UVS
due to a large variety of the students´ learning histories. Thus, the proposed Serious
Game solution should address FSMs from scratch. This leads to the following research
question [6]:
How to design, create, and present a Serious Game to teach the topic of Finite
State Machines to upper vocational school students in Bavaria using stock trading as
an example?
This research question was solved by four steps [6].
• In the first step we reviewed literature to identify the most important learning
objectives. Additionally, we classified and verified them using the Taxonomy of
Bloom [12] in the revision of Krathwohl [13] (Sect. 2).
• In a second step, we integrated these learning objectives into a Serious Game
solution (Sect. 3).
• In the third step, the software was tested with UVS students. The students’ per-
formance was measured by a pre-test and post-test scenario.
• In the fourth step we evaluated the software. The students’ learning progress was
quantified and analyzed in order to prove the suitability of the Serious Game
solution for the target group (Sect. 4).
For being able to teach students about FSMs, the first essential step is to describe the
components of an FSM. According to Yuan and Qu [14] an FSM “consists of a finite
set of states, a start state, an input alphabet, and a transition function that defines the
next state based on the current state and input symbols.” It is important to note, that the
denomination of the elements of an FSM is not standardized. While the term state is
used in practically every literature due to its occurrence in the term Finite State
Machine, the other elements appear in different names in the literature.
In order to teach the concept of FSMs, three steps (A), (B) and (C) have to be
conducted.
(A) The students have to remember and understand the basic concepts and attributes
of an FSM [6].
(B) By using stock trading as an example, the students should gain knowledge about a
real world FSM application and its parallels to automation.
(C) The students should analyze the application of FSMs to construct an Artificial
Intelligence as well as evaluate its benefits and limitations.
zamfira@unitbv.ro
The Finite State Trading Game 573
zamfira@unitbv.ro
574 M. Utesch et al.
zamfira@unitbv.ro
The Finite State Trading Game 575
Fig. 2. Third evolution stage of the non-player character finite state machine
In the following example we assume that the first strategy is applied to the dataset
of the three companies. Additionally, we imply that a striking limit of five percent or
less share price movement for buying as well as selling shares is valid. That means, that
shares are bought if the share price has increased five percent or more compared to the
last turn. Similarly, shares are sold, if the share price has decreased five percent or more
compared to the last turn.
In the presented dataset case, the trading system would invest in company A’s shares
and sell shares of company C. Again, it is important to note, that the system would not
take action unless the limit is met. A striking limit of ten percent for buying and selling
would mean that neither a share would be bought nor sold. Figure 3 shows the devel-
opment of an example share including absolute and relative striking limits. It proves that
it is sensible rather to use relative than absolute striking limits, and shows once again
how the conditions determined for the FSM can directly be used in a business related
progress chart.
zamfira@unitbv.ro
576 M. Utesch et al.
Fig. 3. Example progress chart with absolute and relative striking limits
LO_5_D2 The students shall be able to understand that buying and selling strategies
that are used for trading can be translated into conditions triggering
transitions in Finite State Machines.
The presented examples for the dataset in Table 1 show, that any automated trading
at least resembles the behavior of an FSM. In fact, some trading software are created by
using an FSM Model due to the easy implementation and the possible direct translation.
Based on the input data, the FSM is making decisions. The raw data may be classified
and translated into a machine-readable pattern to enable the FSM to process it. This
input is then compared to each available condition of the current state of the machine to
check for any transitions to be triggered. The FSM that has been constructed by now
takes a certain stock price movement for triggering the transition to the next state. By
adding more input and more conditions – like the legal form of the company, the
country of origin of the company, or long time share price analysis – this FSM would
approximate more and more to the complex logic used in real stock trading. The
presented FSM can thus be seen as a minimalistic and very simple depiction of the
trading logic a broker company uses in real life, although it is important to note, that it
does not represent a fully functional logic.
LO_6_C3 The students shall be able to apply that the data foundation used for stock
trading can easily be used as an input for a Finite State Machine.
The usual civic trying to invest their money in some stocks and highly-specialized
trading companies have one thing in common. They both try to gather as much data as
they can get from a company they plan to invest in. Independently from the very
different channels and possibilities each of them has to pursuit this goal, both
zamfira@unitbv.ro
The Finite State Trading Game 577
occasional customers and professional stock traders take the collected data as their
foundation to act (or not to act). On the one hand, the usual stock trader trades shares
every once in a while, but on the other hand modern stock trading is based on high-end
server systems which are buying and selling stocks in high frequency all over the day.
Nowadays, since in these days almost every share price – regardless of whether from
the past or the present – is available, the systems always have some data which are
ready to work with. Data such as share prices are ideally suitable for automated
decisions because of their availability and easy processing due to the machine-friendly
format. Consequently, the Input Alphabet of the logic constructed so far in Fig. 2 are
the share prices of the current round and the last round. The FSM can make decisions
by using only these data. But the trading systems that are being used in real life are
supplied with as many data as possible of every company offering shares, in order to
increase the chance of success for the decisions. These data are classified with complex
algorithms, for example by using automated news engines to classify news and their
impact on the future share price [16]. Based on these classified data the trading systems
decide to buy and sell shares automatically.
3 Application
All of the presented learning objectives were implemented into a Serious Game about
FSMs, which is called The Finite State Trading Game (FSTG). In this turn-based game
the user strives to successfully trade shares on the market to make profit in order to win
the game against an evolving FSM represented by the NPC [6]. Table 2 shows the
learning objectives and the corresponding NPC’s evolution stage side-by-side in order
to allow a quick overview.
zamfira@unitbv.ro
578 M. Utesch et al.
In order to verify the FSTG approach the software was tested as part of the Pupils’
Academy of Serious Gaming for 11th grade UVS students [17]. The test lesson itself is
fully described in [6].
According to [6] we created a questionnaire to quantify the test results. The
questionnaire relies on a five point Likert scale [18]: 1 – Strongly Disagree,
2 – Disagree, 3 – Neither Agree nor Disagree, 4 – Agree and 5 – Strongly Agree. The
participants self-assessed their knowledge in a pre-test before the start and a post-test
after the end of the teaching unit.
The classification of the test statements corresponds to the classification of the
learning objectives in Bloom’s Taxonomy (see Fig. 1). The statements are discussed
shortly for the category B – for A see [6].
TS_5_C2 I am familiar with the concept of automation using software and the most
relevant steps for the realization.
The statements TS_5_C2 to TS_7_C3 all test for general knowledge about
automation and Artificial Intelligence in real world scenarios, like stock trading. While
a foreknowledge in these areas helps the students to play the Serious Game success-
fully, the statements are primarily designed to show possible improvements of the
students’ knowledge in the post–test. The higher the scores in the post–test, the more
likely the students have understood the explained theory in the Serious Game and the
associations between FSMs, Artificial Intelligence and share trading. Statement
TS_5_C2 asks for a general understanding of the students for automation, what should
ease to understand the learning content regarding automation of the software. In the
Serious Game the automated stock trading scenario helps the students to understand the
learning content.
TS_6_D2 I know usage situations of Finite State Machines in daily life.
This statement focuses on daily life examples of FSMs like traffic lights or auto-
mated stock trading, which students may know without having any experience with
FSMs. After using the software, the students should be able to identify such automated
systems.
TS_7_C3 I am familiar with limit setting in stock trading and its possible automation
using software.
4 Results
The students filled the same questionnaire two times on separate documents. This
means that they assessed their knowledge twice: in a pre-test as well as a post-test of
FSTG [6]. This paper focuses on the test statements TS_5_C2 to TS_7_C3.
zamfira@unitbv.ro
The Finite State Trading Game 579
Fig. 4. Results of the answers to the test statements of category B in the pre-test and the post-test
zamfira@unitbv.ro
580 M. Utesch et al.
4.3 Findings
The results of the post–test seem to verify the success of our FSTG approach to learn
about FSMs by playing a Serious Game. The evaluation statements with less or no
progress show, which parts of the software need some rework.
As a qualitative result, FSTG succeeded to keep the students motivated during the
lesson without further incentives. On the quantitative side, major progress on every
statement can be shown [6]. Only 0.05% of all answers of the post-test were negative.
This is a remarkable fact, when compared to overall 73.50% negative answers in the
pre–test. Hence the number of positive answers shifted from 16.50% in the pre–test to
89.00% positive answers in the post–test. However, a shift of the answered options to
the positive side had to be expected due to the fact that most statements are directly
addressed in the software. Nevertheless, this test is a success, since big progress was
recorded in all test statements. The positive results in category B of the post-test prove
that the students understood the parallels between FSMs and stock trading. Addi-
tionally, the shift of the results shows that the setting of FSTG in a trading scenario
helps the students to grasp the learning content – the application and the concept of
automation in a real life environment, like the trading market.
5 Conclusion
Upper vocational schools (UVS) represent an essential part of the school system in
Bavaria, Germany. In this school form, the curriculum of grade 12 contains the
modules capstone project as well as systems and processes, among the complementary
modules basics of modern programming languages, programming styles and data
structures, and object-oriented programming. The use of the Finite State Machine
(FSM) concept allows to introduce these modules due to the easy scalability and the
widespread application fields of FSMs. In combination with the various advantages of
Self-Regulated Personalized Learning, a serious game in form of an IT-based learning
solution appears to be a promising way of teaching about FSMs.
As a consequence, the goal of this research was to design, create, and present a
Serious Game to teach the topic of FSMs to UVS students in Bavaria by using stock
trading as an example.
In order to do so, nine learning objectives regarding the topic FSMs have been
created, classified and verified by using the Taxonomy of Bloom [13]. The learning
objectives were split into three categories: The basics of FSMs, the parallels between
FSMs and stock trading and the application of FSMs in order to construct Artificial
Intelligence. This research focused on the most important parallels between FSMs and
the mechanisms of a stock trading system. In the next step, the learning objectives were
integrated into a Serious Game called “The Finite State Trading Game” (FSTG). In this
turn-based game the user strives to trade shares efficiently in order to beat the
Non-Player Character (NPC) opponent in several levels and with varying difficulties.
This computer-controlled opponent uses an FSM logic to make its decisions. During
the gameplay important facts about the attributes and the behavior of the NPC are
explained to the user by using a graphical representation of the NPC FSM.
zamfira@unitbv.ro
The Finite State Trading Game 581
The developed Serious Game was tested in a lesson in the context of the “Pupils’
Academy” [19] with students of the local UVS “Staatliche Fachober- und Beruf-
soberschule Technik München” at the Technical University of Munich. Before and
after using the Serious Game the students self-assessed their knowledge about FSMs in
order to record their managed progress on the topic. For the creation of the test
statements the five-point scale developed by Likert [18] and again the Taxonomy of
Bloom was used.
The evaluation of the test sheets proved the test on the target group as a success. In
the pre-test, 73.5% of the answers were rated negative. Nevertheless, only 0.05% of the
given answers in the post-test were negative answers. Comparing the two test rounds, a
significant shift to the positive side of the answers was recorded for each of the test
statements [6].
Since the first test setting of the software has been successful, the developed
Serious Game seems to be suitable for widespread usage in classrooms. The learning
objectives seem to be fitting to the requirements to teach this concept as well as to the
target group. Due to the similarity of the German vocational secondary school and the
German gymnasium in teaching style, the next logical step is to test in gymnasium
classes as well. While it is currently more a supplement to the present teaching way, the
software could work as a stand-alone teaching way in the future by adding more
teaching content and functionalities. Additionally, former research [17] has shown
positive impact of playing serious games on the results in the final exam, the German
Abitur. Thus, we want to focus with our research on this aspect.
The construction of an own FSM in the program illustrates a possible vertical
expansion of FSTG. After successfully playing the levels of the game the users could
get the option to create an own FSM themselves in order to play against the created
logic. The students could instantly apply and intensify their learned knowledge and
thus also provide another verification method for the efficacy of the software in this
way.
Based on the results of the lesson as part of the Pupils’ Academy FSTG can be rated
as a success and has shown once more that serious games are a valuable didactic
approach and can be used to modernize the present ways of teaching.
References
1. Entertainment Software Associate ESA. 2011 essential facts about the computer and video
game industry. http://www.isfe.eu/sites/isfe.eu/files/attachments/esa_ef_2011.pdf
2. Black, P.: Finite state machine. http://xlinux.nist.gov/dads/HTML/finiteStateMachine.html
3. Bayerisches Staatsministerium für Unterricht und Kultus, “Lehrpläne für die Fachoberschule
und Berufsoberschule Ausbildungsrichtungen Technik, Agrarwirtschaft, Gestaltung Unter-
richtsfach: Technologie/Informatik Ausbildungsrichtung Wirtschaft, Verwaltung und
Rechtspflege Unterrichtsfach: Technologie Jahrgangsstufen 11 bis 13”, ed (2006)
4. Technical University of Munich: Bachelor’s Program Informatics: Curriculum and Courses
(2012). http://www.in.tum.de/en/current-students/bachelors-programs/informatics/curriculum-
and-courses.html
zamfira@unitbv.ro
582 M. Utesch et al.
zamfira@unitbv.ro
A Serious Game for Learning Portuguese Sign
Language - “iLearnPSL”
1 Introduction
Sign language is a language which uses a system of articulations of the hand to mediate
a conversation between people with hearing or speech disabilities. Each country has its
own sign language and some have more than one so; there is no standard language in
this matter.
In Portugal, it is estimated that there are about 120,000 individuals with some
degree of hearing loss (included here are the elderly who are gradually losing their
hearing) and about 30,000 with severe and profound deaf problems [1, 2]. Therefore,
Portuguese Sign Language (PSL) was created and developed as a form of communi-
cation. This community is not only made up of people with hearing problems but also
by family members, professionals and friends who daily get along with them.
There are few resources (such as dedicated dictionaries, games and interactive
stories) available for this community, and there is still little support in this matter. Thus,
there is a need to respond to the special educational requirements of deaf children, who
present difficulties in communication, learning, and social interaction.
In 2013, the Leap Motion Controller was released. This devise facilitates computer
support hand recognition. Moreover, the data gathered by the device is relatively
accurate and can be used in several classification methods.
The main motivation of this project involves the lack of interactive applications for
learning PSL. Following this trend, it was designed a serious game that tries to
overcome this gap. The first goal of the game is to teach users to learn the numbers
from 0 to 9. Other activities related to the program followed in the first cycle education
will be developed.
This paper is organized in 5 sections. Section 2, “Related Work” describes some
examples of tools, applications and projects being developed, Sect. 3, “Developed
Game”, presents the interface, methodology and architecture of the game, Sect. 4,
“Preliminary Validation”, presents the first analysis of the game with therapists and
finally, Sect. 5, enunciates the final remarks.
2 Related Work
This section introduces Leap Motion Controller and a brief explanation of its char-
acteristics and Application Programming Interface (API). There are also described
some tools, applications and projects in development for auxiliary sign languages
serious games. Finally, some sign language recognition systems are described.
zamfira@unitbv.ro
A Serious Game for Learning Portuguese Sign Language - “iLearnPSL” 585
The Leap Motion API recognizes and tracks the hand, finger and finger-like tools.
The device operates in an intimate proximity with high precision and tracking frame
rate and reports discrete positions, gestures, and motion. Each frame of tracking data
contains the measured positions and other information about each entity detected in that
snapshot. The hand model provides information about the identity (left or right),
position (coordinates in the working area space), palm orientation (directional vector)
and other characteristics of a detected hand, the arm to which the hand is attached, and
lists of the fingers associated with the hand [6], Fig. 2.
Fig. 2. Leap motions hand model “PalmNormal” and direction vectors define the orientation of
the hand [6].
A finger object provides a bone object describing the position (coordinates in the
working area space) and orientation (directional vector) of each anatomical finger bone.
All fingers contain bones ordered base to tip [6]. In Fig. 3 there are presented the
fingers bone model of a hand.
The main weakness of the leap motion is its lack of accuracy when something
obstructs the device and the hand or if some fingers obstructs other fingers [4].
zamfira@unitbv.ro
586 M. Torres et al.
classification accuracy of 84,1% and 74,8%, respectively, however it has some limi-
tations, namely, the user must be standing and/or has to put on a glove.
Other conducted their research using a Leap Motion Controller with different
approaches. A research work was focused on the development of a decision tree
algorithm where the application decides if a certain hand gesture is correct through
elimination and a very small cross over probability with an accuracy of 82,7% [9]. And,
some used a Support Vector Machine (SVM) as a supervised learning model that uses
training observations to recognize patterns and perform classifications or regression
analyses to recognize the gesture. Each of these were developed in their own sign
language with an a result ranging between 85% and 99% [10–13]. Also, a project with
the usage of two Leap Motion for an Arabic sign language recognition study was also
observed. This had a classification accuracy of 97,1% [14].
zamfira@unitbv.ro
A Serious Game for Learning Portuguese Sign Language - “iLearnPSL” 587
3 Developed Game
This section described the hand gesture recognition model and the developed game
environment.
Fig. 4. Number gesture of PSL. On the left, the representation of the hand gesture [33]. On the
right, the same hand representation displyed in Unity interface through the leap motion.
zamfira@unitbv.ro
588 M. Torres et al.
In the game, when a hand gesture is accomplished in real-time, these variables will
be compared, with a studied margin of error, if it matches the respective hand gesture
model. This margin was determined by multiple testing and analyses of this variables
result; the margin limits were defined by what we considerate appropriate to the
respective hand gesture. The variables can contain the following information:
• Finger extension: The Unity. Leap Dynamic Link Library (DLL) has a function that
can determine if a finger is extended or not;
• Bone Orientation: This is determined by the angle of a certain bone vector of a
finger with the palm normal vector and palm direction vector. This is also a function
from the Unity DLL that determines angles between 3 dimensional vectors.
Example: if the game is waiting for a finger to be extended, the bone orientation
must be higher than 75º of angle to be considered approved;
• Finger Contact: It determines if there is contact between two particular bones of
different fingers. This is obtained true if the distance of these two bones is lower
than 20 mm (virtually).
The users’ hand can float around in the Leap motion detection space because all of
these variables have the palm of the hand as a reference. We describe an example of the
detection of the hand gesture associated to the number “2”: The hand model of number
“2” is obtained when the thumb and pinky fingers are extended and the other fingers are
not extended. So, in the game scenario, if it is asking to represent the number “2” hand
gesture, when the player is executing that same gesture, the game algorithm will detect
if there is a match between the fingers extended by the real-time players’ hand with the
respective hand gesture model, this case, number “2” gesture. Table 1 shows the
variables in consideration to detect the hand gesture.
Table 2 shows the functions that classifies the hand model variables as approved.
For the same case, classifier for the number “2” gesture. These results can be by margin
of values or a Boolean value.
zamfira@unitbv.ro
A Serious Game for Learning Portuguese Sign Language - “iLearnPSL” 589
Table 2. Hand model Variable classifier for gesture number “2” in PSL.
Hand model Function Margin/IO
variables
Thumb finger Finger extended True
extension
Index finger Finger not extended True
extension
Middle finger Finger not extended True
extension
Ring finger Finger not extended True
extension
Pinky finger Finger extended True
extension
Thumb finger Angle between thumbs Proximal bone vector and Palm <110º,
orientation normal vector >85º
Angle between thumbs Proximal bone vector and Palm <60º,
directional vector >40º
Angle between thumbs Intermediate bone vector and <110º,
Palm normal vector >85º
Angle between thumbs Intermediate bone vector and <60º,
Palm directional vector >30º
Pinky finger Angle between thumbs Proximal bone vector and palm >75º
orientation normal vector
Angle between thumbs Intermediate bone vector and >75º
Palm normal vector
All finger Orientation classifications are determined by the angle of the fingers
Proximal bone and Intermediate bone direction vector with the palms normal vector. In
the case of the ‘thumb’, it also uses the palms directional vector. This occurs because
the thumb can bend in front ways and sideways. This methodology was decided to
improve accuracy and to reduce the conflict between gestures with very similar
patterns.
As studied in the research papers mentioned above [10–13], despite of all the Hand
Gestures Models being capable to create, with some precision, some gestures, executed
by the player, they are difficult to be captured by the Leap Motion if they are in their
regular position in space. So, for some of these gestures the player needs to orientate
their hand towards the Leap Motion so it can capture all the fingers clearly. This occurs
because the Leap Motion capture cannot determine the fingers with precision if they are
not all clearly visible, giving in this case false data.
zamfira@unitbv.ro
590 M. Torres et al.
game development. It takes the hand tracking data from the Leap Motion device and
allows the users hand to interact with a virtual scene. Scripts in the Leap.Unity
namespace interacts directly with “GameObjects” and other UnityEngine components
where they take tracking data and put it to use in Unity.
In this case, it is a 3D game to teach users to learn numbers from 0 to 9 using PSL.
The player surpasses multiple levels by executing numeric gestures in PSL. The dif-
ferent challenges consider executing a certain number, giving the result to various type
of calculations such as adding, subtracting, among other challenges. Figure 5, presents
the overall game architecture.
zamfira@unitbv.ro
A Serious Game for Learning Portuguese Sign Language - “iLearnPSL” 591
As the game is aimed to children, the interface was developed using an environ-
ment suitable for them in a user-friendly way. It will be employed in a school envi-
ronment and the usage of bright colors and animated objects and images is adequate.
The playing mechanics is a first-person game where the player can walk around,
interact with objects and enter several rooms using the mouse and/or keyboard. These
classrooms are where the challenges take place.
The level management is based on unlocking the next classroom where the player
can find new challenges of different gestures to accomplish and different/harder math
problem.
In the challenge mechanism, the strategy is to ask one single gesture at a time, to
avoid game algorithm conflicts in the detection of multiple hand gestures, so there will
be no cross over, as well as to allow the player to be concentrated in only one gesture.
Figure 6 presents the scheme of a game challenge.
zamfira@unitbv.ro
592 M. Torres et al.
In Fig. 7, one can see the game interface where the player is trying to execute a
challenge where he/she must make the PSL number “2”, solution of summing one
apple with another apple.
Fig. 7. On the left: example of the game interface of a challenge asking the result of the
equation. On the right: game interface of the challenge succeeded.
As observed in Fig. 7, while the user is trying to execute the hand gesture, a virtual
simulated hand appears in the game environment. This feature is very helpful because it
has a feedback mechanism where the hand can be monitored and if a certain gesture is
being incorrectly performed, the virtual hand will show an indication of which finger is
not being executed correctly.
The game has a help system that includes a list of all gestures and so the player can
see and practice the same way as the game challenge does. Also, it has a score log
where the player can track his/her progress throughout the game. The score mechanism
is based on challenges accomplishment where points are earned by the player.
4 Preliminary Validation
This project was submitted to a preliminary validation of the user interface and
activities performed by consulting a Sign Language Interpreter and a Special and Deaf
Children Therapists and Assistants.
By the opinions collected, the feedback was very positive and very well-accepted.
In global, they described it as an interesting, autonomous and helpful tool to learn PSL.
Also, a good way for typically developing children who are interested in learning PSL.
It could have a good impact to allow deaf children to be more included in the com-
munity [34].
Some aspects were pointed out for future development such as the game challenge
strategy. The game must be very well explained so the children and therapists can have
a good understanding of what the challenges are about (which hand gesture must be
accomplished). Also, the challenges difficulty/content should be aware of the
age/school grade of the children.
Deaf children do not have much support at school or at home. They merely use
conversation and some simple images in their learning/education of PSL and do not use
any kind of auxiliary applications. Also, the majority comes from hearing parents and
their lack of knowledge of the PSL aggravates this situation where the children can find
only few or even none help at home. Moreover, they believe that this project could
improve the child’s communication and social interaction.
zamfira@unitbv.ro
A Serious Game for Learning Portuguese Sign Language - “iLearnPSL” 593
5 Final Remarks
The goal of this research is the development of a serious game (using Unity software)
for gesture recognition in Portuguese Sign Language focusing on assisting deaf dis-
abled children using the Leap Motion sensor. It is a multi-level 3D game (iLearnPSL)
where the player must surpass challenges within the game engine considering the
learning of numbers from 0 to 9.
The sign language recognition is a condition based system, where each hand
gesture has multiple relevant characteristics (variables) which match a certain value
with a studied error margin.
Despite the potential of the Leap Motion, making it an adequate device to help
solving this problem, it has some difficulties in detecting all fingers of a hand in some
hand gestures. To solve this, the player must withdraw his/her hand from the correct
position and face it to the device while executing the same gesture.
Nowadays, there are few interactive learning applications or games for Portuguese
Sign Language. So, this project will make a step forward on this matter. The advantage
of the proposed application is that the children will have higher integration, interac-
tivity and dynamism to the game allowing, through gestures, to learn Portuguese Sign
Language. Moreover, it includes a feedback system for error detection in the execution
of a certain gesture, allowing that the user will not be led into error by the same gesture,
if poorly executed. These reasons make this a more effective and automatic learning
game.
As future work, it is expected to develop more robust hand gesture models to
increase the game accuracy, as well as to extent the content of the game to other type of
challenges as the alphabet and colors. Also, the goal is to develop a hand recognition
system using machine learning, where Support Vector Machines (SVM) can be a
possibility.
To improve its validation and credibility, real world experiments with deaf children
will also be considered.
References
1. Carvalho, C.A.F.: A narrativa em crianças surdas: papel da Língua Gestual Portuguesa
(2013). (in Portuguese)
2. Informação - Comunidade. http://www.apsurdos.org.pt/index.php?option=com_content&
view=article&id=43&Itemid=57. Accessed 7 July 2015. (in Portuguese)
3. Weichert, F., Bachmann, D., Rudak, B., Fisseler, D.: Analysis of the accuracy and
robustness of the leap motion controller. Sensors (Basel) 13(5), 6380–6393 (2013)
4. Potter, L.E., Araullo, J., Carter, L.: The leap motion controller. In: Proceedings of the 25th
Australian Computer-Human Interaction Conference on Augmentation, Application, Inno-
vation, Collaboration, OzCHI 2013, pp. 175–178 (2013)
zamfira@unitbv.ro
594 M. Torres et al.
5. Guna, J., Jakus, G., Pogačnik, M., Tomažič, S., Sodnik, J.: An analysis of the precision and
reliability of the leap motion sensor and its suitability for static and dynamic tracking.
Sensors 14(2), 3702–3720 (2014)
6. API Overview—Leap Motion C# SDK v2.3 documentation. https://developer.leapmotion.
com/documentation/csharp/devguide/Leap_Overview.html#hands. Accessed 4 Apr 2016
7. Sun, C., Zhang, T., Xu, C.: Latent support vector machine modeling for sign language
recognition with kinect. ACM Trans. Intell. Syst. Technol. 6(2), 1–20 (2015)
8. Zafrulla, Z., Brashear, H., Starner, T., Hamilton, H., Presti, P.: American sign language
recognition with the kinect. In: Proceedings of the 13th International Conference on
Multimodal Interfaces - ICMI 2011, p. 279 (2011)
9. Funasaka, M., Ishikawa, Y., Takata, M., Joe, K.: Sign language recognition using leap
motion controller. In: International Conference on Parallel and Distributed Processing
Techniques and Applications, PDPTA 2015, pp. 263–269 (2015)
10. Khelil, B., Amiri, H.: Hand gesture recognition using leap motion controller for recognition
of arabic sign language. In: 3rd International Conference on Automation, Control,
Engineering and Computer Science, ACECS 2016, Proceedings of Engineering and
Technology (PET), pp. 233–238 (2016)
11. Quesada, L., López, G., Guerrero, L.A.: Sign Language Recognition Using Leap Motion,
pp. 277–288. Springer, Heidelberg (2015)
12. Chuan, C.-H., Regina, E., Guardino, C.: American Sign Language Recognition Using Leap
Motion Sensor. In: 13th International Conference on Machine Learning Application,
pp. 541–544 (2014)
13. Simos, M., Nikolaidis, N.: Greek sign language alphabet recognition using the leap motion
device. In: Proceedings of the 9th Hellenic Conference on Artificial Intelligence - SETN
2016, pp. 1–4 (2016)
14. Mohandes, M., Aliyu, S., Deriche, M.: Prototype Arabic sign language recognition using
multi-sensor data fusion of two leap motion controllers. In: 2015 IEEE 12th International
Multi-conference on Systems, Signals and Devices, SSD15, pp. 1–6 (2015)
15. LIVPSIC - Livraria de Psicologia e Ciências da Educação. http://www.livpsic.com/v4/
detalhe01.php?id=1340. Accessed 8 July 2015. (in Portuguese)
16. ASP - Associação de Surdos do Porto. http://www.asurdosporto.org.pt/artigo.asp?idartigo=
1250. Accessed 8 July 2015. (in Portuguese)
17. Four Online Kids’ Games to Learn Sign Language. http://www.brighthubeducation.com/
special-ed-hearing-impairments/2910-learning-asl-with-internet-browser-games/. Accessed
8 July 2015
18. Sign 4 Me for iPad - A Signed English Translator para iPhone, iPod touch e iPad na App
Store no iTunes. https://itunes.apple.com/pt/app/sign-4-me-for-ipad-signed/id383462870?
mt=8. Accessed 8 July 2015
19. ASL American Sign Language – Aplicações Android no Google Play. https://play.google.
com/store/apps/details?id=com.teachersparadise.aslamericansignlanguage. Accessed 8 July
2015
20. Wix.com culturas_surdos created by leticia_nadia_lgp based on kindergarten. http://leticia-
nadia-lgp.wix.com/culturas_surdos#!page-8. Accessed 16 Apr 2016
21. RogerVoice, An Android App That Helps The Deaf Have A Conversation On The Phone -
Forbes. http://www.forbes.com/sites/federicoguerrini/2014/09/26/tech-that-matters-rogervoice-
will-allow-the-deaf-finally-have-a-conversation-on-the-phone/. Accessed 8 July 2015
22. Tech Tackles Sign Language – MotionSavvy - Forbes. http://www.forbes.com/sites/
karstenstrauss/2014/10/27/tech-tackles-sign-language-motionsavvy/. Accessed 8 July 2015
23. Leap Motion App Store|Leap Motion Apps for Motion Control. https://apps.leapmotion.
com/. Accessed 16 Apr 2016
zamfira@unitbv.ro
A Serious Game for Learning Portuguese Sign Language - “iLearnPSL” 595
zamfira@unitbv.ro
The Implementation of MDA Framework
in a Game-Based Learning in Security Studies
1 Introduction
In this day and age where society is made up of “digital natives”, many uses of IT
applications have emerged beyond our expectation. Technology has indeed become
crucial in assisting various aspects of human’s lives. Nowadays it is very common for
us to see technology being applied in different fields, including medicine, business,
military and education. One of these applications which very recently and suddenly
caught many attentions of scientific communities is a concept called Digital
Game-Based Learning (DGBL) [1].
By its simplest definition, Prensky defines any learning process through video
games on a computer or any other media whether online or not. In addition, he stated
applying DGBL is not strictly to the classroom, but it can be applied to teach various
things including in business or law [2]. As of now, the application of DGBL has been
tested and proven effective by many (see [3–5]). However, many are still in doubt
about the effectiveness of DGBL. Unfortunately, this doubt is considered under-
standable as it would be hard to conduct a comprehensive analysis on this issue [1].
That is why, in order to ensure the authenticity of DGBL research, the design and
development process of the game must be done in accordance to proven studies and
framework. It is also helpful to have an expert of the related field to give constant
feedbacks on the game. This research aims to monitor the implementation process of a
previously defined DGBL design. The DGBL will be specifically used to enhance
learning process in Security Studies of International Relations Major. The game was
designed using MDA Framework, whereas the implementation will be based on a
framework for creating IT-based learning method called the IT-BluTric [7].
The choosing of this topic is due to the applicability of video games in assessing the
learning content of security studies. Essentially, security studies, or also known as
strategic studies, is a subfield within International Relations that studies potential
security threats of a nation and how to solve or prevent those threats [8]. The con-
sideration process of what strategies should be used to address a particular threat is
similar to resources allocation process in many games. It was due this that this topic is
deemed suitable to be applied using DGBL.
The result of this research is an assessment of the IT-BluTric framework imple-
mentation in the development process of DGBL. This paper illustrates the whole
process of DGBL development in this research. Furthermore, feedbacks of field experts
will be given to iteration of the game in order to make sure the game still delivers the
desired learning objectives.
2 Literature Reviews
zamfira@unitbv.ro
598 J.V. Moniaga et al.
As can be seen from the pictures above, Mechanics in MDA Framework refer to the
game system such as algorithm, levels, rules, points, badges, highscores, etc. In short,
mechanics are what drive the users’ actions. Dynamics refer to what the users can see
on the screen based on their inputs.
Aesthetic on the other hand, goes down on a deeper level where game designers
have to convey the emotional and psychological messages to the users [9]. Aesthetics
are often seen as the “fun” factors of the game; they made up the emotional connection
created during the interaction between the player and the game [10].
zamfira@unitbv.ro
The Implementation of MDA Framework 599
zamfira@unitbv.ro
600 J.V. Moniaga et al.
As can be seen in Fig. 4, the 5 pillars helps IT Blueprint framework translates “the
business” in the form of information flow, to create the technology. To help gain a
deeper and better understanding of the each pillar of the IT Blueprint Framework, a list
of IT Initiatives is assigned to each pillar. IT Initiative itself is a guideline of what
features are going to be on the project. Applying these IT Initiatives into the IT
Blueprint framework by certain rules, is referred as IT-BluTric [7].
IT-BluTric will divide each IT Initiative in accordance to its relation to one of the
IT Blueprint pillars. Every initiative will then be ranked using 4 aspects: urgency,
importance, cost and timeframe (see Table 1). By doing so, developers will know
which feature needs to be done first. This also makes monitoring the project, so that it
does not go off from its original purpose, a little bit easier.
1. Urgency
Urgency refers to how soon an IT Initiative should be done. When the urgency
value of an initiative is high, it means that the initiative should be prioritized over those
with lower urgency value. Urgency value of an initiative (U) can be counted by using
project time (PT) and spare time (ST) (see Eq. (1)). The maximum urgency value is
100% [11].
2. Importance
Importance is very similar to urgency, as both aspects are used to decide what
initiative should be done first. Their difference is, while urgency focuses on whether an
initiative should be done as soon as possible, importance sees the impacts of a process.
High importance value means a process has a big impact on other processes. This
means that, processes with less importance values cannot be done unless those with
high importance values are done [11] (Table 2).
zamfira@unitbv.ro
The Implementation of MDA Framework 601
3. Cost
Cost aspect is the budget needed to finish a particular initiative/process. Cost does
not refer to the value of the budget required but rather the ability to provide them.
Higher cost value means that it will take longer for the money to be provided, meaning
that the process will take longer to finish. This can be due to many reasons, including
making detailed calculation and the organizational bureaucracy. Defining cost value
does not have any particular formula or metrics to be used, simply because the dif-
ference between each and every organizations [11].
4. Timeframe
Timeframe is the time needed to complete an initiative/process. Usually, timeframe
can be categorized into 3: short, medium and long period. In this framework, the
timeframe is defined like the table below [11] (Table 3).
3 Implementation
3.1 IT-BluTric
The vision of this research and application is to create a video game based on a
predefined design that will serve as an alternative learning method in Security Studies.
The design of the game was already defined in the previous research [6]. It mainly uses
MDA framework [10] along with expert consultations as the guideline for creating the
design (for details, see [6]).
This research implements the design into a complete game application using the
IT-BluTric framework. In order to do this, the IT Initiatives of each pillar are defined
based on the needs of the application. Then IT-BluTric is implemented in order to
determine the working order of application development. The IT-BluTric implemen-
tation can be seen in Tables 4, 5, 6, 7 and 8 below.
zamfira@unitbv.ro
602 J.V. Moniaga et al.
zamfira@unitbv.ro
The Implementation of MDA Framework 603
Table 9. Timeline
Table 9 shows every IT Initiatives in order from the first one until the last. Each
initiative entry is given estimated time to finish. This research effectively started on
July, making it 5 months since the first process is done. This is the reason some of the
processes are still yet to be done. However, the progress up until now is sufficient
enough to provide relevant data on the framework’s performance in implementing the
design of the game.
zamfira@unitbv.ro
604 J.V. Moniaga et al.
4 Results
Results shown in this research are the products of IT Initiative entries up to quality
assurance. This is the reason why most of the processes are done in the Application
pillar. The results of the scenes in the game can be found on the following figures, with
the exception of the quality assurance process. Result of quality assurance will be
feedbacks by lecturer of related subject. The feedbacks will be related to MDA
Framework [10] and 4 key characteristics of a learning game [9].
zamfira@unitbv.ro
The Implementation of MDA Framework 605
4.2 Feedbacks
Feedback results from the 3 experts regarding the game can be found in Table 11.
Experts are the experienced people who understand and know the heuristics of this
area, put themselves in the shoes of a user to test the game and give the feedbacks to
researchers [15, 16].
zamfira@unitbv.ro
606 J.V. Moniaga et al.
To guide the experts to do collect the feedbacks, DECIDE framework been chosen
because it fits to the heuristics. These are the steps of the DECIDE framework:
determine goals, explore questions, choose paradigm, identify practical issues, decide
how to deal with ethical issues, and evaluate – present data [17, 18].
From the feedbacks collected, it is clear to see that the game is still on track of its
initial goal and design. The questions asked were based on the definition of each 4 key
characteristics of a learning game [9]. Seeing as most of the questions were answered
with positive (yes) answers, it can be said that the design successfully encompassed
MDA Framework as these characteristics cover the entire framework. With these
results, game development will proceed without doing many changes to the initial
planning. The implementation of MDA framework in a game-based learning in security
studies is a perfect match because it can deliver the learning objectives.
5 Conclusion
DGBL has been proven as an effective learning method. Unfortunately, not many can
provide a comprehensive analysis on the validity of these studies. That is why, in order
to reduce the possibility of failures in developing the game, this research combine and
implement the IT-BluTric framework in the development process of the game.
zamfira@unitbv.ro
The Implementation of MDA Framework 607
References
1. Van Eck, R.: Digital game-based learning: it’s not just the digital natives who are restless.
Educause Review, March/April 2006
2. Prensky, M.: Digital Game-Based Learning. Paragon House, St. Paul (2007)
3. Erhel, S., Eric, J.: Digital game-based learning: impact of instructions and feedback on
motivation and learning effectiveness. Comput. Educ. 67, 156–167 (2013)
4. Guillén-Nieto, Victoria, Aleson-Carbonell, Marian: Serious games and learning effective-
ness: the case of It’s a Deal! Comput. Educ. 58(1), 435–448 (2012)
5. Kickmeier-Rust, M.D., Albert, D.: Educationally adaptive: balancing serious games. Int.
J. Comput. Sci. Sport 11(1), 15–28 (2012)
6. Ayu Asih Kusuma Putri, R., Wijaya, Y., Moniaga, J.V.: A design model for digital
game-based learning in international relations study developing an innovative learning
method for defense strategy course in Bina Nusantara University. In: International
Conference on Game, Game Art, and Gamification (ICGGAG) (2016). Unpublished
7. Astriani, M.S., et al.: Delivering an interactive presentation in supporting of dynamic
teaching method with an IT blueprint framework: IT initiative-ITBluTric. In: International
Conference on Information Management and Technology (ICIMTech) (2016). Unpublished
8. Elkus, A.: Professor, Tear Down This Wall: Is the Divide Between Security Studies and
Strategic Studies Permanent?” War on the Rocks, 18 April 2016. http://warontherocks.com/
2016/04/professor-tear-down-thiswall-is-the-divide-between-security-studies-and-
strategicstudies-permanent/
9. Kim, J.T., Lee, W.-H.: Dynamical model for gamification of learning (DMGL). Multimedia
Tools Appl. (2013)
10. Ruhi, U.: Towards a descriptive framework for meaningful enterprise gamification. Technol.
Innov. Manage. Rev. 5(8), 5–16 (2015)
11. Astriani, M.S., Pradono, S., Moniaga, J.V.: IT Initiative for creative interactive teaching
presentation based on IT blueprint framework. In: Advances in Educational Technologies
(2014)
zamfira@unitbv.ro
608 J.V. Moniaga et al.
12. Astriani, M.S., Pradono, S.: IT Blueprint and school. In: Proceedings of the 10th WSEAS
International Conference on Computational Intelligence, Man-Machine Systems and
Cybernetics, and Proceedings of the 10th WSEAS International Conference on Information
Security and Privacy, pp. 160–167. World Scientific and Engineering Academy and Society
(WSEAS) (2011)
13. Astriani, M.S.: IT Blueprint – jembatan bisnis dan teknologi. In: Binus Information
Communication and Technology Conference (2011)
14. Cassidy, A.: A Practical Guide to Information Systems Strategic Planning, 2nd edn.
Auerbach Publications, Boca Raton (2006)
15. Nielsen, J.: Enhancing the explanatory power of usability heuristics. In: Conference
Proceedings, CHI 1994 (1994)
16. Molich, R., Nielsen, J.: Improving a human- computer dialogue. Commun. ACM 33(3),
338–348 (1990)
17. Smith-Atakan, S.: The FastTrack to Human-Computer Interaction. Thomson Learning,
Boston (2006)
18. Scneiderman, B.: Designing the User Interface: Strategies for Effective User-Interface
Engineering. Addison-Wesley, Boston (2005)
zamfira@unitbv.ro
Industrial Virtual Environments
and Learning Process
Today, we are in the fourth industrial revolution. In this revolution, we include the
Industry 4.0. The connectivity of all the objects, in our life or in the industry with the
sensors and the actuators connected on the industrial network, create the Industry 4.0.
Other people propose another definition for the Industry 4.0: the simulation of the
industry is the new Industry 4.0. They build a virtual factory, with all the sensors,
actuators, networks, Programmable Logic Controllers (PLC’s) and so on…, they study
the simulation with the reaction of the process, refine the system, and after validation,
they can build the real factory in another part of the world.
Since the early 1990s, technologies of virtual environments - virtual reality, aug-
mented reality, multi-user 3D platforms - continue to offer new tools for a better
understanding of the complex systems in various areas, particularly in industrial
activities. The rise of communication networks, the wide dissemination of tablets and
mobile phones of the latest generation, allow a new approach of the learning process
and teaching methods. In practice, a user, or a group of users can immerse themselves
in such an environment via a visualization tool adapted to the context like a computer
screen, a projection screen, or a headset. Today, we use these devices to simulate
industrial systems for learning, training or decision-making. This new research area
results from the union of two multidisciplinary teams of researchers at Le Havre
University: the first one interested in Decision Support Software (DSS) and Industrial
Risk Prevention, the second one working on the 3D video games technologies applied
to learning and training.
This contribution is organized in four sections. In the first one, we describe a DSS
for risk prevention using Case Based Reasoning (CBR) approach linked to a
MultiAgent System (MAS). The second section presents an Intelligent Tutoring System
(ITS) fit to prevent students’ dropout risks. This module is linked to the virtual campus
GE3D described in the following section, in which we explains how a 3D virtual
campus can be used for simulating and teaching PLC’s. Finally, in the last section, we
discuss the most relevant elements of our research and the next developments.
2 Risk Prevention
This system allows to represent the observed situation but also its evaluation.
Evaluating the situation can be performed by calculating its possible consequences. This
can be carried out using previous situations whose consequences are known. So, analogy
reasoning relies on the following hypothesis: if a A situation looks like a B situation, the
consequences of the A situation ought to be similar to those of the B situation.
Case-Based Reasoning (CBR) [1] is a methodology of resolution of problems being
based on the reuse of past experiments for solving new problems. Decision Support
Systems are among the most promising applications of CBR. Thanks to the CBR, the
human abilities in problem solving are enhanced by power of the computing system.
zamfira@unitbv.ro
Industrial Virtual Environments and Learning Process 611
The reasoning approach of CBR uses past cases. A case is defined by a problem
and its solution. The problem to be solved is called target case. The solved problems
are called source cases and are stored in a case-base. The CBR cycle [2] is composed of
five steps (Fig. 2):
zamfira@unitbv.ro
612 J. Grieu et al.
system is to watch permanently the current state of industrial risky zones, called current
situation - or in our case: the target CBR - in order to allow decision makers to gather
information as soon as possible about the potential risk generated by this current
situation [7].
zamfira@unitbv.ro
Industrial Virtual Environments and Learning Process 613
In this system [10], we recognize the generic MAS engine (independent of the
domain), the knowledge part (in green) which allows to specify the domain, and finally
the 3D immersion, part necessarily dependent of the learning area. It is noteworthy that,
depending of the case, the human tutor can intervene in the process of decision as
planned by the paradigm of the CBR. We propose an intelligent tutoring system with
its 3D students’ interface and its internal decision support system designed to face the
challenge of processing the evolution of a large amount of data.
On the other hand, the disaffection of the students with the engineering studies obliges
us to change our way of teaching to encourage them to return to our training courses.
Considering this trend, our research team has built a learning platform online.
There is a lot of learning system online all over the world. But these conventional
systems are for the students who are used to work alone or who are advanced students.
After studying the results of interviews with students, we have developed GE3D [11],
which is a learning system online, in accordance with their wishes and our criteria.
Even if this tool uses some elements coming from the 3D video games, it remains a
pedagogical tool.
We use a hybrid solution, compromise between 3D real time and 3D precalculated,
which allows us to get the best of both systems for creating computer graphics. The
choice of a virtual campus rather than a traditional platform for distance learning was
obvious for the students of the department of the Genius of Electricity and Industrial
Computing from the IUT of Le Havre. These “e-native” students have adopted now for
a long time the technologies used in the video games. GE3D is a multi-users tool, with
a synchronous technology; this means that any action of a user in the virtual world will
be perceived at the same time by the other users of the platform.
The technical choices were made according to the following specifications:
• A 3D web technology with fluidity;
• An Open source technology to enable us our own developments;
• A platform with a client-server architecture;
• On the client side: a hardware and network resource-efficient system.
We will describe the use of this virtual space with a course on the PLC’s (Pro-
grammable Logic Controllers).
The typical pedagogical scenario of a course on the PLC’s is as follows:
• An appointment is given to all the students in the amphitheater (Fig. 4);
• Before coming in the amphitheater, the students download a document with blanks
that they must complete during the lecture;
• During the presentation, the teacher can use the screen for slides or videos. He can
also use a white board and a microphone for the audio part. In the same time, the
students write on their documents and ask questions if necessary with the public
chat;
zamfira@unitbv.ro
614 J. Grieu et al.
• After the course, the students respond to a multiple choice test on line. If they
succeed to pass the test, they can reach the exercise room. If not: they repeat their
test in the virtual examination room;
• In the room for the exercises (Fig. 5), the teacher gives the students some exercises
and the students propose their solutions by using the whiteboard;
• When all the exercises are completed, everybody can join the industrial room
(Fig. 6). Here, the teacher shows how the equipments are running and he stays with
the students answering to their questions;
zamfira@unitbv.ro
Industrial Virtual Environments and Learning Process 615
• When the students succeed all the exercises, they can download specifications of
various industrial processes described in the videos available in the next room
(Fig. 7);
• After programing, the students join the teacher in the real room (Fig. 8) of the
PLC’s to validate their solutions. They use a simulator of operative part, made in 3D
(Fig. 9) [12].
zamfira@unitbv.ro
616 J. Grieu et al.
zamfira@unitbv.ro
Industrial Virtual Environments and Learning Process 617
• First of all, for its cost: indeed, the acquisition of a real operative part, built with
hydraulic cylinders, engines, sensors and other actuators, represent a highly pro-
hibitive expenditure for a training center. Whereas the virtualisation, with less cost,
permits to copy out numerous operatives parts to infinity;
• Then, for the security: a virtualised machine will be safely turned on by learners
with nothing to fear of a wrong handling for their security or the equipment
integrity;
• At last, for the realism: the simulation of the industrial process allows students to
see immediately the action of their program. Previously, their attention was very
relative in front of a blinking led representing the start-up of the conveyor. Today,
when their program is validated on the simulator, they are proud of their job and all
the more so motivated to do it.
However, currently we can find some inconveniences linked to the utilisation of
those technologies:
• The modelization ad hoc of 3D objects remains a time-consuming activity;
• VR headsets currently proposed are still heavy and can cause nausea;
• The choice and the conception of interfaces allowing interactions remain a tech-
nological lock which has to be lifted;
• The lack of maturity of a learner can engender a confusion between virtuality and
reality.
At the present time, our realisations focus essentially on driving virtualised
industrial installations by Programmable Logic Controllers (Fig. 9). Thereafter, we
consider the behavioural analyse of learners thanks to sensors («Eyetracking» and
headset «EEG») and the conception of adapted Human Machine Interfaces.
We can expect that technics of treatment and massive data, the “Big Data”, will
allow us to better identify the needs and behaviours of learners by personalizing their
pedagogic process so as to reduce abandon risks.
Eventually, 3D technologies showed us their utility in comprehension of complex
systems. Inasmuch as it already participates the approach of “Simplexity” which
involves creating simple, intuitive and ergonomic interfaces, to help for
decision-making in more and more complex universes.
Acknowledgments. This project has been supported by the European Commission under the
ERDF (European Regional Development Fund) Programme through the 5.5 Action of the
GRR CLASSE Programme.
zamfira@unitbv.ro
618 J. Grieu et al.
References
1. Kolodner, J.: Case-Based Reasoning. Morgan Kaufmann Publishers, San Francisco (1993)
2. Aamodt, A., Plaza, E.: Case-based reasoning: foundational issues, methodological
variations, and system approaches. AI Commun. 7(1), 39–59 (1994). IOS Press Amsterdam,
The Netherlands
3. Jennings, N., Wooldridge, M., Sycara, K.: A roadmap of agent research and development.
Auton. Agent. Multi-agent Syst. 1(1), 7–38 (1998). Kluwer Academics Publishers, Boston
4. Moulin, B., and Chaib-draa, B.: An overview of distributed artificial intelligence. In:
Foudations of Distributed Artificial Intelligence, pp. 3–55. Wiley, New York (1996)
5. Ferber, J.: Les systèmes multi-agents. InterEditions, Paris (1995)
6. Wooldridge, M.: An Introduction to Multiagent Systems. Wiley, Chichester (2002)
7. Boukachour, H.: Système de veille préventive pour la gestion de situation d’urgence:
modélisation par organisations d’agents. Application aux risques industriels. Ph.D.
dissertation. Le Havre University, France (2002)
8. Willging, S.: Factors that influence students’ decision to dropout of online courses.
J. Asynchronous Learn. Netw. (JALN) 8(4), 105–118 (2004)
9. Hafner, K.: Software tutors offer help and customized hints (2004). http://www.nytimes.
com/2004/09/16/technology/circuits/16tuto.html?_r=0
10. Person, P., Galinho, T., Lecroq, F., Boukachour, H., Grieu, J.: Intelligent tutor design for a
3D virtual campus. In: 6th IEEE International Conference on Intelligent Systems (IS),
pp. 74–79. IEEE Publisher, Sofia (2012). doi:10.1109/IS.2012.6335194
11. Grieu, J., Lecroq, F., Person, P., Galinho, T., Boukachour, H.: A virtual campus for
technology-enhanced learning. In: Education Engineering (EDUCON), IEEE 2010, pp. 725–
730. IEEE Publisher, Madrid (2010). doi:10.1109/EDUCON.2010.5492506
12. Riera, B., Vigaro, B.: Virtual systems to train and assist control applications in future
factories. In: 12th IFAC Symposium on Analysis, Design, and Evaluation of
Human-Machine Systems, pp. 76–81. Elsevier publisher, Las Vegas (2013)
zamfira@unitbv.ro
How Game Design Can Enhance Engineering
Higher Education: Focused IT Study
Abstract. This paper seeks to report on the current state and attitudes towards
higher education (HE) curriculum for the creative (game) industry sector in
Ukraine. It is based on preliminary findings from high education and industries
surveys, which examined the competences, demanded by this important sector
of the UA economy from one hand, and, from another, offers of HEs in
developing them. Moreover, a review of the literature performed to define the
core employees’ profiles and their competences on the field job market. This
paper explores competences, professional and transversal, that are important for
the Ukrainian game industry and in what way should students be taught for
“creative” tasks. This paper offers interested parties an analysis on how HE in
Ukraine can develop relevant curriculum and deliver “industrial” education for
students who intend to operate in this sector. The study results could be helpful
for HE and policy makers to respond to current and future education needs.
1 Introduction
The world IT market is growing annually on 5–20% including game industry [1].
According Global Games Market Report [2] worldwide game industry will generate a
total of $99.6 billion in revenues in 2016, what is increasing almost 8.5% compared to
2015. Moreover, it expects that global market will grow up to 6.6% toward 2020,
eventually reaching $118.6 billion. It is no surprise that IT products in whole and
computer games particularly have become a major industry and are one of the fastest
growing application markets in Ukraine [3]. The development scene of creative sector
of world economy is expanding, therefore, amount of startups and companies based in
European countries and worldwide are increasing. Such movement influences on
accepting of new education policy which offers new grades, courses and curriculum
according the contemporary challenges and demands. For example, every EU Tech-
nical university has at least one program devoted to the game design [4] and approx.
280 bachelor and master programs on game design are available at 385 USA colleges
and universities [5].
A primary goal of the work presented in the paper is to figure out the required
knowledge and skills on the local and national game sector market in Ukraine. A side
benefit of this work will be a development of learning modules and an implementation
them in a curriculum that meets the demands and interests of the future games
engineers.
Games developers produce games for different operational system using existing
engine machines or creating new one. A game production can involve from few
employees to large studios and take several months or even years from creating ideas
and characters to programming and testing. Each stage of digital game development
involves various tasks for different roles, such as:
• Designer – creates a game flow and how play in a game. The game could be original
ideas or work from an existing concept
• artist – creates the game’s visual characters, objects and scenery, and produces
concept art and drawings (storyboards) at the planning stage
• animator – brings the characters, objects and scenery to life with computer mod-
elling and animation software during the production stage
• programmer – creates the code to make the game work. On this stage the work
could include graphics design, artificial intelligence, or gameplay software.
The literature review shows that the candidate on a position of a digital games
developer needs to demonstrate follow competences:
• excellent computer skills
• a wide knowledge and understanding of computer games
• creativity and imagination
• a logical approach to problem-solving
• good teamwork and communication skills
• flexibility and adaptability
• the ability to work under pressure and meet deadlines
• patience and attention to detail
• willingness to keep up with industry developments and learn new skills.
A survey consolidates and relates on above mentioned competences – knowledge
and skills - needed from students of engineering schools dreaming to work in digital
game industry. It was oriented mainly on a game sector as well as a high education in
Ukraine and organized in all of parts of country and performed in frame of the Erasmus+
project “GameHub: University-enterprises cooperation in game industry in Ukraine”.
The analysis of the resent survey will result by an integration of computer game
design with a software engineering course. The modules for the course will be
implemented in next steps of the project performance, along with additional course
material including syllabus, slides, projects, and other course materials specific to game
design in software engineering.
This paper is presented as follows: Sect. 2 introduces the already mentioned
GameHub Initiative as it is – the project goals and its wider and specific objectives,
target groups, and partnership; Sect. 3 outlines a profile of the IT engineers working at
game industry in Ukraine as well as development knowledge and skills of IT students
using the potential of existing national higher education; Sect. 4 describes the study on
zamfira@unitbv.ro
How Game Design Can Enhance Engineering Higher Education 621
the common (core) and specific (professional-oriented) competences for game industry
job positions and discusses preliminary results; Sect. 5 introduces the didactic approach
recommended for the use in GameHub pilot action and finally, Sect. 6 summarizes our
conclusions and introduces possible future work.
2 GameHub Initiative
The GameHub project was created to modernize the existing engineering education in
Ukraine by enhancing students’ knowledge and skills in creative game development
sector. The project was started at autumn 2015 and is co-financed by ERASMUS+
programme, Cooperation for Innovation and the Exchange of Good Practices key
action and Capacity Building in Higher Education action.
The main goal of the project is building a bridge to connect in mutually beneficial
cooperation universities and game industry by fostering and human investing the
Ukrainian emerging ICT creative business sector.
The paper delivers three main outcomes: (1) preliminary analysis of high education
and industries surveys concentrating on the competences, both professional and
common (core), important for this sector of UA economy; (2) a review of the literature
performed to define the main employee’s profiles and their competences on a job
market of the field; (3) recommendations for HE in Ukraine concerning methods and
instruments for the development of relevant curriculum and deliver of “industrial”
education for students who intend to operate in this sector.
IT market in Ukraine is increasing every year and requires more and more qualified
specialists. High salaries, great amount of working places, opportunity for career
growth (including traveling abroad) attract many young professionals and unemployed
persons to the IT industry. However, employees do not always meet the employers’
needs and possess necessary knowledge and skills. The reason is that most educational
programs in IT specialties are out of date, and, as a result, do not correspond to the
labor market requirements in rapidly developing IT sector.
In order to understand the needs of the game market in Ukraine and build con-
temporary profile of the university IT student the several surveys were performed.
zamfira@unitbv.ro
622 O. Dziabenko et al.
In order to create the knowledge and skills needed by game studios, the questionnaires
were designed. 41 game business representatives were interviewed and questioned. The
respondents evaluated common (core) and specific (professionally-oriented) knowledge
and skills which are necessary for the digital game design employee [10]. To simplify
the process, we suggested to evaluate the most popular eight job positions: content
manager, storyteller, scriptwriter, sound programmer, web-client programmer,
sketcher, 3D Character Artist/3D Environment Artist, QA tester, JS programmer [11].
For each job positions the presented below knowledge and skills were voted on the
scale not important at all and very important. The common capacity includes, e.g., to
zamfira@unitbv.ro
How Game Design Can Enhance Engineering Higher Education 623
identify and solve problems; o work in team and achieve mutual goals; to apply gained
knowledge and understanding of subject area/profession in practice; to adapt to dif-
ferent situations and flexibility; to work independently; to accept constructive feedback
on the work; to pay attention to details and quality evaluation; for creative and
imaginative capabilities; for self-education and self-development; for diversity sensi-
tivity; for effective communication/interpersonal skills; for analysis and synthesis; task
planning and time management; for excellent verbal and written communication in
foreign languages; for leadership and decision-making.
For each job position we created two competence profiles for common knowledge
and skills, the other for specific ones (Table 1). As an example, diagram on the Fig. 1
shows that leadership and decision-making as well as teamwork are dominated over the
work independently for a content manager (common knowledge and skills).
In addition, the study of curricular of the GameHub European academic partners
demonstrates that all competences may be compiled in four clusters: Design, Pro-
gramming, Creative skills, and Transversal skills.
Design combines competences used for preparation of the preliminary models and
sketchers for digital game, planning of digital game form and structure.
Programming includes competences necessary for computer programming, soft-
ware development, analysis, content development, algorithm generation, testing algo-
rithm requirements, algorithm/architecture solution.
Creative skills describe competences used for the development of digital game
sketch, image, music due to aesthetic principals and high level attraction.
zamfira@unitbv.ro
624 O. Dziabenko et al.
Fig. 1. Content Manager: competences (very important – green, important – yellow, slightly
important – red, and not important at all – blue)
Transversal skills include traditional knowledge and skills that contribute to the IT
specialist’s personal fulfilment, such as: communication skills; foreign language pro-
ficiency; basic knowledge in mathematics, physics, natural sciences; ability to study,
for social responsibility; ability for entrepreneurship, cultural literacy, creativity. Here
can be also referred skills in project management, knowledge of planning and control
methods, project monitoring and analysis.
Based on the above-mentioned GameHub surveys and conducted analyses the
competence profile of IT specialist in digital game industry is created. It includes but
not limited to:
Common competences
– Responsibility, care about quality of work;
– Adaptability and interpersonal skills;
– Efficiency and ability to self-improvement;
– Creativity, ability to system thinking;
– Focus on achieving a success.
Instrumental competences
– Capacity for research work, analysis and synthesis of technical information;
– Teamwork;
– Outstanding computer/programming skills;
– Capacity for written and oral communication in their native language.
zamfira@unitbv.ro
How Game Design Can Enhance Engineering Higher Education 625
Specific/professionally-oriented competences
– Ability to develop user’s requirements specifications to software;
– Ability to perform requirements analysis, develop specification of software
requirements, conduct their verification and certification;
– Basic understanding of the fundamentals of software simulation/design, types of
models, main concept of unified modeling language UML;
– Ability to simulate different system aspects for which the software is developed;
– Ability to develop algorithms and data structures for software products;
– Understanding of current tendencies in software structure and architecture, software
design methods;
– Basic understanding of modern psychological principals of human-machine inter-
action, methods of human-machine interface development;
– Ability to human-machine interface analysis, design and creation of prototypes;
– Ability to reusable components development and application;
– Knowledge of basic methods and techniques of visual programming;
– Ability to solve mathematical, physical and economic problems via development of
appropriate applications;
– Ability to use hardware possibilities.
The obtained competence profile allows to establish methodic approaches for the
development of didactic base for the curricular modification. In other words, conducted
competence analysis of employers’ requirements in digital game industry makes it
possible to justify the preliminary structure of training-methodic supply of GameHub
laboratory.
zamfira@unitbv.ro
626 O. Dziabenko et al.
efficiently develops several competences at the same time. The project may incorporate
a teamwork involving creative capacity and directing to the real-life problems, e.g.,
production of working prototype of STEM game for secondary education level.
6 Conclusion
In the paper we presented the obtained results that allow to create a set of common and
specific (professionally-oriented) competencies, and to determine methodical approa-
ches for the development of didactic base to improve the engineering curricula in
Ukrainian technical schools and universities. The conducted competence analysis of
employers’ requirements in digital game industry makes it possible to establish a
GameHub laboratory [14] - the structure of education equipment for building mean-
ingful final students projects with cooperation of national and international creative
industry.
Moreover, based on the above-mentioned analyses, studies and surveys we have
developed recommendations how to apply innovative teaching methods for the
development of needed common and specific competences in digital game design for
training specialists for appointed job position.
In the paper we show the required knowledge and skills on the local and national
game sector market in Ukraine is settled. The paper offers to apply “creative” tasks and
project-based approach as most effective methods for teaching students in this field.
In the future the consortium is planning to create the learning modules and mate-
rials in format of open education resources to modernize the engineering curricular
matching and satisfying to real-life tasks and objectives in creative sector of Ukraine.
The developed open education resources will be tested and evaluated through the pilot
action at six engineering schools. The results of this trial will be published on the
project website (http://gamehub-cbhe.eu/) and in the facebook group discussion wall.
The study results could be helpful for HE and policy makers to respond to current
and future education needs.
Acknowledgement. This work was partially funded by the European Union in the context of the
GameHub project (Project Number: 561728-EPP-1-2015-1-ES-EPPKA2-CBHE-JP) under the
ERASMUS+ programme. This document does not represent the opinion of the European Union,
and the European Union is not responsible for any use that might be made of its content.
We want to thank all GameHub partners who contributed to the interview, survey and dis-
cussion of analysis of labor market for game industry in Ukraine.
References
1. PwC forecasts (2016). http://venturebeat.com/2016/06/08/the-u-s-and-global-game-industries-
will-grow-a-healthy-amount-by-2020-pwc-forecasts/. Accessed 10 Aug 2016
2. Global Games Market Report (2016). https://newzoo.com/insights/articles/global-games-
market-reaches-99-6-billion-2016-mobile-generating-37/. Accessed 20 Aug 2016
zamfira@unitbv.ro
How Game Design Can Enhance Engineering Higher Education 627
3. International Factfile 2015: Ukraine, Games industry news, The Market for computer &
video games, 9 November 2015. http://www.mcvuk.com/news/read/international-factfile-
2015-ukraine/0158786. Accessed 20 Aug 2016
4. Animation Career Review (2016). http://www.animationcareerreview.com:8080/careers-
animation. Accessed 20 Aug 2016
5. The 2016 Essential Facts About the Computer and Video Game Industry, Entertainment
Software Association (ESA), April 2016, Ipsos MediaCT for ESA. http://essentialfacts.
theesa.com/Essential-Facts-2016.pdf. Accessed 20 Nov 2016
6. Official data of the State Statistic Service of Ukraine. http://www.ukrstat.gov.ua. Accessed 3
Nov 2016
7. IT specialist portrait in Ukraine. https://dou.ua/lenta/articles/it-portrait-2015/. Accessed 12
Aug 2016
8. Official data of the Association of the Ukrainian Outsourcing Companies. Exploring
Ukraine. IT Outsourcing Industry. http://hi-tech.org.ua/exploring-ukraine-it-outsourcing-
industry-the-volume-of-it-outsourcing-services-provided-in-ukraine-has-grown-by-a-factor-
ten/. Accessed 15 Nov 2016
9. Official data of the Association: Information Technologies in Ukraine. http://itukraine.org.
ua/analitychni-materialy. Accessed 3 Nov 2016
10. International standard of education classification. http://www.uis.unesco.org/Education/
Documents/isced-fields-of-educationtraining-2013RU.pdf. Accessed 3 Nov 2016
11. National standard of education classification (project). http://naps.gov.ua/uploads/files/sod/
NSKO.pdf. Accessed 3 Nov 2016
12. GameHub report: Task Analysis; Development of Competence Profiles. http://gamehub-
cbhe.eu/project-results/. Accessed 20 Nov 2016
13. GameHub report: Developed Didactical Approach in Training. http://gamehub-cbhe.eu/
project-results/. Accessed 20 Nov 2016
14. GameHub report: Eleborated Technico-Pedagogical Requirement on game Laboratory.
http://gamehub-cbhe.eu/project-results/. Accessed 20 Nov 2016
15. Tuning Academy. http://www.unideusto.org/tuningeu/publications.html. Accessed 20 Nov
2016
zamfira@unitbv.ro
Physioland - A Serious Game
for Rehabilitation of Patients
with Neurological Diseases
1 Introduction
The number of people affected by neurological diseases, that is, diseases related to the
Nervous System, whose basic functional unit is the neuron or nerve cell, is increasing
daily. By receiving and transmitting the nerve impulses, the neurons conduct the
collected information [1]. The occurrence of lesions in neurons, whether genetic,
congenital or acquired, triggers cellular dysfunctions, compromising the transmission
of electrical signals, culminating in neuronal death and, consequently, configuring a
neurological disease [2]. Some examples of neurological diseases are multiple scle-
rosis, stroke, Friedreich’s ataxia and Parkinson’s disease.
Although different in their genesis, because the causes of neuronal damage are
different, the neurological diseases that have been mentioned before have something in
common: the transmission of the electrical signals for the movement cease, the muscles
lose the activity and atrophy, compromising the mobility of the affected patients [3].
Neurons are mostly amitotic cells, so when they are injured, they cannot be repaired
by cell division. However, it is common to observe a recovery, although partial, of the
lost functions, even in cases where the consequences of the injuries are severe, which
shows that the Nervous System has the capacity to develop mechanisms that allow it to
adapt the possibilities of the individual to the challenges of the environment. Thus,
when there is loss of neurons, new alternative neural circuits are sought which can
replace the injured structures. These adaptive mechanisms are the manifestation of an
intrinsic property of the Nervous System, neuroplasticity or neuronal plasticity, which
can be defined as the ability it has to modify its structural and functional organization,
as long as it is subjected to repeated stimuli [4].
It seems to be consensual, in the literature on this subject, that the practice of motor
exercises induces neuronal plasticity. Therefore, people with mobility problems as a
result of neurological diseases should have adequate treatment programs that include
functional activities, as believing that their training encourages neuroplasticity, patients
can achieve a positive clinical evolution, which will give them independence and,
consequently, a better quality of life [5].
Therapeutic resources based on movement, applied to patients with neuronal
lesions, can stimulate new connections with the Central Nervous System, contributing
to its reorganization: new dendrites can sprout; it is possible to extend branches already
present; existing synapses can be altered or new synapses can be created; changes in
axons may occur; new neurotransmitters can be produced [6]. It is here that physical
therapy becomes essential to promote sensory stimuli able to encourage neuroplasticity
and, thus, to contribute to the recovery of the functional movements of patients who
have suffered sequels due to neurological diseases [7].
This paper is organized as follows: Sect. 2 introduces the Physical Therapy; Sect. 3
refers to Serious Games and Motivation, Sect. 4 presents the purpose of this work;
Sect. 5 mentions the Physiotherapeutic Exercises adapted to electronic game situations;
Sect. 6 details the Monitoring of Physical Therapy Exercises Using Image Processing
Techniques; Sect. 7 describes the development of the Physioland Game, namely the
General Architecture, the Physioland Concept and the Special Features of Develop-
ment; Sect. 8 focus on the Results obtained from the experience; Sect. 9 shows the
Conclusion of this work.
2 Physical Therapy
zamfira@unitbv.ro
630 T. Martins et al.
4 Purpose
As mentioned, serious games entertain players and reinforce healthy movements, while
create engaging and pleasant environments, providing challenging and rewarding
experiences. However, we have verified that there is a gap with regard to serious games
specifically designed to support the traditional physical therapy. To overcome this gap,
it became a general objective of this work to develop a serious game, which was called
Physioland, to motivate and monitor the practice of physical therapy of patients with
neurological disease. The non-invasive system uses image processing techniques to
monitor patients and it adjusts to electronic game situations some exercises proposed
by traditional physical therapy. This game is intended for all people with neurological
disorders in mild or moderate condition, having or not balance.
zamfira@unitbv.ro
Physioland - A Serious Game for Rehabilitation of Patients with Neurological Diseases 631
5 Physiotherapeutic Exercises
After analyzing several exercises, with health professionals, six exercises were selec-
ted. They are interesting from a technological and physiotherapeutic point of view, and
with different dynamics to be adapted to electronic game situations, in order to motivate
patients to perform them. The chosen exercises were as follows (Fig. 1):
• Glenohumeral joint abduction/adduction (Fig. 1(a));
• Glenohumeral joint flexion/extension (Fig. 1(b));
• Radioumeral joint flexion/extension (Fig. 1(c));
• Hip joint abduction/adduction (Fig. 1(d));
• Cross-movement (Fig. 1(e));
• Pulleys (Fig. 1(f)).
Fig. 1. Set of exercises chosen for the Physioland Game. (a) – Glenohumeral joint
abduction/adduction; (b) – Glenohumeral joint flexion/extension; (c) – Radioumeral joint
flexion/extension; (d) – Hip joint abduction/adduction; (e) – Cross-movement; (f) – Pulleys
The glenohumeral joint abduction can be defined as the movement occurring in the
frontal plane, around a horizontal axis directed dorsoventrally, and that moves the arm
away from the midline of the body. The glenohumeral joint adduction is the movement
in the opposite direction to that of abduction, that is, it takes place in the same frontal
plane, around the same axis, but it approximates the arm of the midline of the body [21].
The glenohumeral joint flexion occurs in the sagittal plane, being performed for-
wards and upwards, around a transverse axis. If the movement is in the same plane, but
in the opposite direction, representing the return of this flexion, is called the gleno-
humeral joint extension [21].
The radioumeral joint flexion occurs in the frontal plane and consists of the
movement of the arm towards the shoulder, which results in a decrease of the
arm-forearm angle. In the opposite direction, it is performed the radioumeral joint
extension, which provides a growth of the same angle, since the arm moves away from
the shoulder [21].
zamfira@unitbv.ro
632 T. Martins et al.
The hip joint abduction is performed in the frontal plane, around a horizontal axis in
the anteroposterior direction, and consists of the lateral elevation of the leg, that is, it
moves away from the midline of the body. As with the glenohumeral joint, also for the
hip joint, the abduction return movement, which approaches the midline leg of the
body, is called the hip joint adduction [21].
Cross-movement is a diagonal movement, with each arm alternately moving for-
ward and up or down, attempting to reach a target appearing on the opposite side,
placed randomly by the physiotherapist.
The exercise that uses the pulleys consists of a combined movement of the
glenohumeral joint abduction/adduction and the radioumeral joint flexion/extension
and, therefore, it is performed in the frontal plane.
zamfira@unitbv.ro
Physioland - A Serious Game for Rehabilitation of Patients with Neurological Diseases 633
team realized that in the technique of performing each exercise there are four main
characteristics to detect if it is correctly performed: angles defined by different body
segments, alignment, compensation and speed [24].
Regarding the angles, the ideal values are specific to each exercise. Being A, B, and
C three joints, with their coordinates it is possible to define two vectors ~ v1 and ~v2 ,
according to Eqs. (1) and (2):
!
~
v1 ¼ BA ¼ A B ð1Þ
!
~
v2 ¼ BC ¼ C B ð2Þ
The amplitude h of the angle of the two vectors is given by Eq. (3):
! !
1 BA : BC
h ¼ cos ! ! ð3Þ
BA BC
The alignment, with few exceptions, requires that some joints align with each other in
the exercise execution plan. It is necessary to compare the coordinates of the involved
joints, relative to a given axis. If h is this axis, and A and B are any two joints, the
referred comparison will be made by determining the module of the difference between
the coordinates of A and B, relative to h, that is, by calculating the value of the
following expression (4):
j hA h B j ð4Þ
After making the comparisons two by two, the largest of the values obtained is
recorded. Ideally, for the alignment to be correct, this value should be as close to 0
(zero) as possible, which means that the coordinates of all the compared joints, relative
to the axis in question, are relatively close to each other, which symbolically is
translated by the expression (5):
hA ffi hB ð5Þ
Compensation refers primarily to leaning the body to the right or to the left (lateral),
forward or backward. The procedure for checking if there is or there is not compen-
sation is the same that was described for alignment. It may happen that the joints
involved are other, and that the axis to be considered is also different.
The average angular speed, x, of running an exercise, in radians per second, in a
time interval Dt, in seconds, in which the angle described by a body segment undergoes
a change Dh, in radians, is given by Eq. (6). It should remain noticeably constant
during the exercise performing, which corresponds to an average acceleration
zamfira@unitbv.ro
634 T. Martins et al.
Dh
x¼ ð6Þ
Dt
Dx
c¼ ð7Þ
Dt
These equations will be used to determine the speed and the acceleration of the exe-
cution of each monitored exercise. An exception is the cross-movement, for which only
compensation is monitored, because it is a very free movement.
7 Physioland Game
The Physioland has a medieval concept, a topic that can interest the extended target
audience of the game – people with neurological disorders in mild or moderate con-
dition. It has a 3D environment, and it was developed in the game engine Unity 3D [25].
It was our main task to develop a narrative that was easily understood by neuro-
logical patients, which they could easily interpret. It was also decided to create game
situations that should be approached, as much as possible, of what the player is doing,
when he/she performs the physiotherapeutic exercises, without distracting him/her and
without losing his/her interest and motivation, this is, game situations directed to the
needs and restrictions of the patients.
zamfira@unitbv.ro
Physioland - A Serious Game for Rehabilitation of Patients with Neurological Diseases 635
Fig. 3. Physioland screenshots: (a) Physioland loading screen; (b) Main menu; (c) Settings
zamfira@unitbv.ro
636 T. Martins et al.
patients must be always accompanied by their health professional who is responsible for
all the settings and for putting the game to work. The patient should only run the game.
The game can be run in play or free mode, but the first is the most versatile and
complete. When a health professional chooses this mode (the clinic login is required in
the settings and Internet connection), it appears the list of health professionals of the
clinic that is authenticated. After the selection of the health professional, the list of
patients of the selected professional is displayed and, when one of these is selected,
he/she can see the summary table of the patient. The levels are then loaded with their
settings for that patient. At the beginning of each level, a screen with information is
displayed along with an animation that shows how the exercise should be performed.
At the end of each level a summary is presented with the score obtained. After the
sequential execution of all levels intended for the player, it is redirected to the main
menu.
In this game mode, all information about the player is loaded from the remote
database, as well as all data coming from the performance of each player are sent to the
health professional. According to the specifications of each level and to the players, the
challenges can be performed with the upper or lower limbs, on the right side, on the left
side or on both sides, for 2.5 min or 5 min each one.
zamfira@unitbv.ro
Physioland - A Serious Game for Rehabilitation of Patients with Neurological Diseases 637
The second level, «sunset», refers to the exercise of the glenohumeral joint
flexion/extension, which leads the player to pick up coins that, just as on the previous
level, are positioned in the arch described by the limb hand that matches the player and
the character (Fig. 4(b)). The logic of the score is the same as the previous level, as is
the procedure that occurs when an incorrect position is detected by the player.
In the third level, the «boat», the player will have to perform the exercise of the
radioumeral joint flexion/extension to pick up the coins that are in the arch described by
the hand of the limb that the character is exercising (Fig. 5(a)), which is, again, the one
of the opposite side of the player. The patient score procedure follows the standards of
the levels already described and the patient must guarantee a minimally correct posture
to be able to execute the game.
The fourth level was called «cascade» and refers to the exercise of the hip joint
abduction/adduction. It is the only level of the game that provides the patient with the
exercising of the lower limbs. The character has to pick up the coins that appear in the
arch described by the foot of the limb that is being exercised (Fig. 5(b)), opposite to the
one of the player. The shorter the coins are of the midline of the character’s body, the
greater is their value. The most valuable coin is obtained when the patient completes
the abduction. Then, new coins come up and he begins to perform the exercise in the
opposite direction, that is the adduction. In this sense, the coins appear with values in
descending order. The situation repeats until the stopwatch reaches zero.
zamfira@unitbv.ro
638 T. Martins et al.
The fifth level, «bridge», gives the patient an exercise with freer movements
(Fig. 6(a)). As the game progresses, the main character has to pick up the coins that
appear randomly in front of him (more to the right, to the left, to the top, to the bottom,
more or less distant), which requires the player to perform the cross-movement exer-
cise. The main character (and therefore the player) must alternate between the left upper
limb and the right, regardless of where the coin appears. However, this model is not
rigid, it can be executed with only one arm, or use the other if the patient cannot reach a
coin with the first, since it is important that the problems of each patient are respected.
Whenever one coin is collected, another appears elsewhere, after a few seconds.
Theoretically, whenever the player picks up a coin, he/she should pick up his/her arm
to the starting position. Since not all patients can do it, the game does not check this
situation, making it more versatile. If the patient is in compensation, which, with
respect to performance, is the only variable monitored, the error procedure is the same
as in previous exercises.
«Fishing» is the nickname for the sixth level (Fig. 5(b)). The character, who is in a
small boat, has to pick up the coins that are falling from the top, alternately, to the left
and to the right. However, on each side, there is a margin of randomness in the fall of
the coins. To overcome this challenge, the patient has to move the boat to the side
where they will appear, performing the exercise of the pulleys. This challenge can be
run on several levels of difficulty (easy, medium, difficult, and very difficult). These
levels of difficulty are reflected in the speed of the falling of coins. All coins have the
same punctuation value. If the patient drops a coin into the water, he/she is penalized,
and the score is withdrawn. Whenever the patient takes an incorrect posture, the alert is
given and the coins disappear during this time, a situation similar to what happens in
the other levels.
On the right side of the game screen is a graph corresponding to the acceleration of
the patient’s movement, which varies in color: green, yellow, orange and red. It is
possible to observe, on the left side of this graph, an arrow that moves up and down, as
it concerns the acceleration of the movement in one or in the other direction of exe-
cution. The patient can be guided by this graph to confirm that he/she is achieving a
speed control.
zamfira@unitbv.ro
Physioland - A Serious Game for Rehabilitation of Patients with Neurological Diseases 639
As previously indicated, Unity 3D was the game engine used to create Physioland.
It enables the development of games in 3D environments and programming in C# and
Javascript language, which is very flexible and has libraries with which it is possible to
integrate the Microsoft Kinect SDK.
Among the software used to develop the graphical components of Physioland, there
are: Blender, a complex but versatile open-source tool for 3D development; Adobe
Fuse, for the creation of characters and their composition, texturing, rendering and
skinning; Adobe Photoshop, for the texturing and creation of 2D elements of the
graphical interface; IClone Pro, for animation; IClone 3dXchange, for the rigging and
export of models for Unity 3D.
The Steinberg Cubase software, an audio workstation, was used for Physioland
sound-plastic process, and video editing was achieved with Adobe Premiere Pro, a
professional software for editing audiovisual contents.
zamfira@unitbv.ro
640 T. Martins et al.
8 Results
9 Conclusion
The serious game Physioland was developed to motivate and monitor patients with
neurological diseases in their physical therapy practice. The use of the Unity 3D game
engine in combination with the Microsoft Kinect sensor for the detection of physio-
therapeutic movements proved to be a very reliable solution and with very good
accuracy for the proposed objective, without freezes and with real-time response.
The results obtained with the questionnaire to the eleven patients show that
Physioland is a game able to challenge the patient to do more and better, to motivate to
continue his/her treatment, to encourage to complete the exercises and to help to
abstract from the annoyance that therapy causes. It is a game that all the patients
interviewed prefer to use for physical therapy exercises, to the detriment of traditional
practice. When they were asked to give an overall appreciation of Physioland, the word
“motivator” was the one most used by them.
Thus, it can be said that the developed game was able to respond positively to the
goal defined for this research: to develop a serious game, based on image processing
techniques, to motivate and monitor the practice of physical therapy of patients with
reduced mobility, as a consequence of neurological disease, providing a good com-
plement to traditional physical therapy. In the course of the experience, the patients
showed more and more enthusiasm, wishing to play Physioland over and over.
zamfira@unitbv.ro
Physioland - A Serious Game for Rehabilitation of Patients with Neurological Diseases 641
References
1. Santos, K.G.L.: Bases anatomofisiológicas do corpo humano I, Rio de Janeiro. Universidade
de Castelo Branco, Brasil (2010). (in Portuguese)
2. Shieh, C.-C., Coghlan, M., Sullivan, J.P., Gopalakrishnan, M.: Potassium channels:
molecular defects, diseases, and therapeutic opportunities. Pharmacol. Rev. 52(4), 557–594
(2000)
3. Villar, F.A.S.: Alterações centrais e periféricas após lesão do sistema nervoso central.
Considerações e implicações para a prática da fisioterapia. Rev. Bras. Fis. 2(1), 19–34
(1997). (in Portuguese)
4. Ribeiro Sobrinho, J.B.: Neuroplasticidade e a recuperação da função após lesões cerebrais.
Acta Fisiátrica 2(3), 27–30 (1995)
5. Borella, M.P., Sacchelli, T.: Os efeitos da prática de atividades motoras sobre a
neuroplasticidade. Revista Neurociências 17(2), 161–169 (2009). (in Portuguese)
6. Mulder, T., Hochstenbach, J.: Adaptability and flexibility of the human motor system:
implications neurological for rehabilitation. Neural Plast. 28(1–2), 131–140 (2001)
7. Oliveira, C.E.N., Salina, M.E., Annunciato, N.F.: Fatores Ambientais que influenciam a
plasticidade do SNC. Revista Acta Fisiátrica 8(1), 6–13 (2001). (in Portuguese)
8. Peres, C.P.A.: Estudo das sobrecargas posturais em fisioterapeutas: Uma abordagem
biomecânica ocupacional. Universidade Federal de Santa Catarina, Florianópolis (2002). (in
Portuguese)
9. Smith, S.T., Talaei-Khoei, A., Ray, M., Ray, P.: Electronic games for aged care and
rehabilitation. In: Proceedings of the 11th International Conference on e-Health Networking,
Applications and Services, Sydney, NSW (2009)
10. Martins, T., Carvalho, V., Soares, F.: Application for physiotherapy and tracking of patients
with neurological diseases – preliminary studies. In: IEEE 2nd International Conference on
Serious Games and Applications for Health (SeGAH), Vilamoura, Portugal (2013)
11. Martins, T., Carvalho, V., Soares, F.: Monitoring of patients with neurological diseases:
development of a motion tracking application using image processing techniques. Int.
J. Biomed. Clin. Eng. (IJBCE) 2(2), 37–55 (2013)
12. Alcará, A.R., Guimarães, S.É.R.: A instrumentalidade como uma estratégia motivacional.
Revista de Psicologia Escolar e Educacional 11(1), 177–178 (2007). (in Portuguese)
13. Sprinthall, N.A., Sprinthall, R.C.: Psicologia educacional - Uma abordagem desenvolvi-
mentalista, Lisboa: McGraw-Hill (1993). In Portuguese
14. Kaufmann-Sacchetto, K., Madaschi, V., Barbosa, G.H.L., Silva, P.L., Silva, R.C.T., Filipe,
B.T.C., Souza-Silva, J.R.: O ambiente lúdico como fator motivacional na aprendizagem
escolar. Cadernos de Pós-Graduação em Distúrbios do Desenvolvimento, vol. 11(1), pp. 28–
36 (2011) (in Portuguese)
15. Burke, J.W., McNeil, M.D.J., Charles, D.K., Morrow, P.J., Crosbie, J.H., McDonough, S.
M.: Optimising engagement for stroke rehabilitation using serious games. Vis. Comput. Int.
J. Comput. Graph. 25(12), 1085–1099 (2009)
16. Neves, D.E., Santos, L.G.N.O., Santana. R.C., Ishitani, L.: Avaliação de jogos sérios casuais
usando o método GameFlow. Revista Brasileira de Computação Aplicada 6(1), 45–59
(2014). (in Portuguese)
17. Sik-Lányi, C., Brown, D.J.: Design of serious games for students with intellectual disability.
In: Proceedings of the 2010 International Conference on Interaction Design and International
Development, Bombay, India (2010)
zamfira@unitbv.ro
642 T. Martins et al.
18. Martins, T., Araújo, M., Carvalho, V., Soares, F., Torrão, L.: PhysioVinci – a first approach
on a physical rehabilitation game. In: Proceedings of the Fifth International Conference on
Serious Games Development and Applications, SGDA 2014, Berlin, Germany (2014)
19. Malone, T.W.: What makes things fun to learn? Heuristics for designing instructional
computer games. In: Proceedings of the 3rd SIGSMALL Symposium and the First SIGPC
Symposium on Small Systems, New York, NY, USA (1980)
20. Talug, D.Y.L: Lifelong learning throughout today’s occasions namely social media and
online games. In: Procedia – Social and Behavioral Sciences, 4th World Conference on
Educational Sciences, WCES-2012, Barcelona, Spain (2012)
21. Sharkey, J.: The Concise Book of Neuromuscular Therapy: A Trigger Point Manual. Lotus
Publishing, Chichester (2008)
22. Shotton, J., Girshick, R., Fitzgibbon, A., Sharp, T., Cook, M., Finocchio, M., Moore, R.,
Kohli, P., Criminisi, A., Richard, A., Kipman, A.Blake: Efficient human pose estimation
from single depth mages. IEEE Trans. Pattern Anal. Mach. Intell. 35(12), 2221–2840 (2013)
23. Shotton, J., Fitzgibbon, A., Cook, M., Sharp, T., Finocchio, M., Moore, R., Kipman, A.,
Blake, A.: Real-time human postura recognition in parts from single depth images. In: IEEE
Conference on Computer Vision and Pattern Recognition (CVPR) 2011, Colorado Springs,
CO, USA (2011)
24. Martins, T., Carvalho, V., Soares, F.: Tracking of physiotherapy exercises using image
processing techniques. In: Controlo 2016 - Proceedings of the 12th Portuguese Conference
on Automatic Control, Guimarães, Portugal (2016)
25. Martins, T., Araújo, M., Lopes, O., Carvalho, V., Soares, F., Matos, D., Marques, J.,
Machado, J., Torrão, L.: PhysioVinci – Solução Integrada para Reabilitação Física de
Pacientes com Patologias Nerurológicas, Video Jogos 2014, 6–7 Novembro 2014, Barcelos,
Portugal (2014). (in Portuguese)
zamfira@unitbv.ro
Human Computer Interfaces, Usability,
Reusability, Accessibility
zamfira@unitbv.ro
The Development of ICT Tools
for E-inclusion Qualities
An Early Case Study
Dena Hussain(&)
Abstract. With the diversity and increasing use of different information and
communication technologies (ICT) in the educational sector, new pedagogic
approaches are also being introduced and have had a major impact on the
educational sector, focusing on different perspective including improved edu-
cational methods and in both schools and homes, information and communi-
cation technologies (ICT) are widely seen as enhancing learning, fulfilling their
rapid diffusion and acceptance throughout developed societies. But the need to
utilize ICT tools to support and guide educators in finding the right support for
students with special individual needs is still a challenge, investigating different
challenges that are presented to teachers in their working environment is an
ongoing matter. One of these challenges that teacher face frequently is creating
an inclusive environment. An “inclusive education” is a process of strengthening
the capacity of the education system to reach out to all learners involved. It
changes the education in content, approaches, structures and strategies, with a
common vision that covers all children of the appropriate age range. Inclusion is
thus seen as a process of addressing and responding to the diversity of needs of
all children. Therefore an inclusive education system can only be created if
schools become more inclusive, in other words, if they become better at edu-
cating all children in their communities with their individual needs. Therefore,
creative forms of communication should be encouraged to promote personalized
care, hence the focuses of this research is to investigate the use of data process
flow map with the aim to guide the teacher towards an inclusive way of
thinking.
1 Introduction
The utilization of ICT tolls has been investigated and introduced in many studies in
different context. The potential of social inclusion and exclusion that technology can
offer, and the way in which technology can facilitated to access information sources,
learning opportunities and personal agencies can be investigated [1]. The World
Declaration on Education for All, adopted in Jomtien, Thailand (1990) [2], sets out an
overall vision: universalizing access to education for all children, youth and adults, and
promoting equity. This means being proactive in identifying the barriers that many
encounter in accessing educational opportunities and identifying the resources needed
to overcome those barriers [2]. Flexible teaching-learning methodologies necessitate
shifting away from long theoretical, pre-service-based teacher training to continuous
in-service development of teachers [2]. A survey regarding ICT in education was
commissioned in 2011 by the European Commission Directorate General Communi-
cations Networks, Content and Technology to benchmark access, use of and attitudes
to ICT in schools in the EU27, Croatia, Iceland, Norway and Turkey, the conducted
survey investigating the use of ICT in education. More than 70% of teachers surveyed
at all grades expressed a positive or very positive opinion about the relevance and
positive impact of ICT to support different students’ learning processes (working
collectively, autonomously, practicing, etc.) and objectives (motivation transversal
skills, higher order thinking skills, etc.) [3]. The possibility of people participating in
the Information and Knowledge Societies is dependent on the availability and
affordability of ICTs and relevance of contents and services, but also on their acces-
sibility: ‘users must be able to perceive, understand and act upon ICT interfaces’ [3].
The objective of this study is to investigate the utilization of Information and
Communication Technologies (ICT) in creating a digitalized process that can assist
educators in finding the right support for pupils with special individual needs, were
generalized teaching methods cannot be applied and student needs are challenging to
recognize. ICT is a particularly valuable tool for children with special needs and can
improve their quality of life, reducing social inclusion and increasing participation.
The aim of the platform is to guide the teacher towards an inclusive way of
thinking, creating a balance between three main factors, which are the student, the
environment and associated activities, resulting in full participation which is the main
concept for this research project to create an inclusive environment for every child with
special educational needs, creating a unified model for inclusion.
This study is part of a European project where the objective of the research is to utilize
the Index of inclusion as a mind map, several associates are involved, including a
Municipality representing Sweden as strategic partnership, together with municipalities
in Germany and Iceland. The project involves different education schools (pupils aged
8–18), primary schools (pupils aged 5–12) and secondary schools (pupils aged 11–18).
The average age of pupils involved in the project is 9.5 years in the primary school and
14.5 in the secondary. The overall project focus is to learn from each other by sharing
experiences of the inclusion work carried out in each country, hence teachers act as a
gateway and their skills development and curriculum resources need increased support.
The fact that the drive towards equity in education through the support of accessible
ICT is a main concept hence is the main research goal in this project. The focus of this
paper, is to create an ICT tool “Digi-Flow” that can assist educators to find the right
support for students with special educational needs. The goal is to create an ICT tool
“Digi-Flow” in such a way that can help determine what is best for children with
zamfira@unitbv.ro
The Development of ICT Tools for E-inclusion Qualities 647
Several fundamental outcomes have been achieved during the early stages of this
research, including a full pedagogical assessment for the Index of inclusion, bench-
marking the concept of “inclusion” between all European partners, creating a unified
model that can be adapted to all countries. The development of the ICT tool
“Digi-Flow” was influenced by gathering different reflections and information which
was collected via different investigations and surveys that where performed, giving a
clearer objective for the tool and the data needed to be included.
zamfira@unitbv.ro
648 D. Hussain
Since the ICT platform was created with partnerships from three different European
countries it was important to understand different national requirements and regulations
therefore three data clusters were considered:
1. Country legal, regulatory and programmatic commitments.
2. Country capacity to implement and apply the introduced solution.
3. Country actual results for children with special needs.
The objective of this research is to create an effective platform which can help
determine what is best for children with special needs, therefore the most important
element is to create a logical structure of questions. The designed platform will utilize
the Index of inclusion as a mind map in a unified form between all three countries, and
as a design process following the Inclusion stages- process flow for the digital tool. In a
previous study the dimensions of inclusion where identified and categorized into three
main categories [4], which are:
1. Equivalence: the school’s capability to see/recognize and understand the pupils
preconditions and needs.
2. Accessibility: the school’s capability to adapt teaching, localities and social com-
munity from a diversity of needs.
3. Participation: the school’s capability to stimulate pupils to ‘take part’; learning to be
lead, to lead oneself and learning to lead others.
To insure quality and effectiveness an auditing process was used during the development
of the ICT tool, the main objective of this process was to confirm the different data
collected and to help verify the different platform requirements and specifications during
developed. Creating a more efficient and effective environment for all partners included
in the project hence reduce and forms of redundancy in both the development method
and data collected. The behavior of this process was integrated within the development
method used, therefore had an incremental nature. As shown in Fig. 2, the audit process
consist of several main stages and sub stages which are:
zamfira@unitbv.ro
The Development of ICT Tools for E-inclusion Qualities 649
4. Planning: The aim of this phase was to help create a focus point, selecting specific
features from a list of different requirement definitions, and to identify which
requirement set to verify and develop further into specifications. Sub stages
included:
(a) Select priority from list
(b) Review objectives
(c) Set standard
5. Data Collection: During this phase the defined requirement sets where expressed in
different data forms, reviewing the objective of the different data needed to be
included and why. This was achieved via three sub stages:
(a) Design audit
(b) Collect data
(c) Analyze data
6. Reporting: The objective of this phase was to verify and validate requirements that
where translated into specification via prototyping via collecting feedback from all
partners and participants.
7. Implementation and Monitoring: All feedback gathered in the previous stage was
assessed and evaluated, redefining requirements and introducing changes when
needed, and therefore reviewing initial requirement standard and creating action
plans.
8. Review and Re-audit: The aim of this phase was to review all decision making and
create additional plans.
Early results show that the need for such as tool has been confirmed, but also that
the potential users for the tool can vary. Using this process also helped identify which
type of data can and must be included in the ICT platform and why, such data included
information regarding the child´s perspective and related social background. As shown
in Fig. 3, different ideas where gathered and analyzed to help create the data sets
required for the ICT tool “Digi-Flow” [4].
zamfira@unitbv.ro
650 D. Hussain
The results also show different user groups, indicating that the user group which
found most need for such a platform are special educators, whom work directly with
children with special needs on different levels. As shown in Fig. 4.
The study also emphasizes that the tool can be utilized as a platform to improve
communication not just for the different educators working with children with special
needs but also as a platform for communication between the school and parents.
The developed platform consists of different questions which are divided into
categories, weights and bases. The categories are used as a factors to link resources and
the bases are to filter the questions. Base in this project means two things.
9. Base-Questions (or parent-questions) can have questions depend on them, meaning
sub-questions (or children-questions).
10. Base-Questions act’s as the filtering process.
As the platform fetches a question it examines if it’s a base-question. If it is,
depending on the answer provided, it will identify any sub-questions linked to it. The
‘weights’ in this ICT platform are used to determine the importance of the questions.
Weights are stored in every answers and effect the questions in a ‘positive’ or ‘nega-
tive’ way. Every question and resources are linked to a category. It maintains good
structure and simplifies understanding and goal to what the objective of the question is.
The main categories of full participation are displayed with a numerical representation
of their relevancy to the current evaluation. The results obtained from different user
groups utilizing the tool includes, a main resource page, which contains links and
information of persons and country organizations that can be contacted. Additional
studies and prototyping is in progress as part of this study and are part of future results
verifying the advantages of ICT tools in this context. To what extent the ICT tool is
spread will be measured in long-term perspectives and assessed accordingly.
zamfira@unitbv.ro
The Development of ICT Tools for E-inclusion Qualities 651
4 Conclusion
As a conclusion, using ICT tools to link schools and different recourses can deliver
substantial educational benefits, for both teachers and children with special needs. By
assisting teacher in the process towards an inclusive environment helps create a sus-
tainable and effective platform which can help determine what is best for the students
and helping the teacher determine the needed actions and link resources to relevant
information.
References
1. Sheehy, K.: ICT and special educational needs: a tool for inclusion. Br. J. Learn. Disabil. 33
(4), 206–207 (2005)
2. Policy Guidelines on Inclusion in Education Published by the United Nations Educational,
Scientific and Cultural Organization 7, place de Fontenoy, 75352 Paris 07 SP, France,
UNESCO (2009)
3. Policy Guidelines on Inclusion in Education Published by the United Nations Educational,
Scientific and Cultural Organization, UNESCO, p. 7 (2013)
4. Hussain, D.: “The utilization and development of ICT tools for inclusion qualities in cases of
special need children.” Book of industry papers, posters and abstracts. In: International
Conference on Health and Social Care Information Systems and Technologies (2016)
zamfira@unitbv.ro
Insights Gained from Tracking Users’
Movements Through a Cyberlearning
System’s Mediation Interface
Abstract. Cyberlearning has the ability to connect learners from diverse set-
tings to learning resources regardless of the learners’ proximity to traditional
classroom environments. Tracking users’ movements through a cyberlearning
interface provides data that can be used both to interpret students’ level of
engagement in the learning process and to improve the cyberlearning system’s
user mediation interface. The Online Watershed Learning System (OWLS),
which serves as the end user interface of the Learning Enhanced Watershed
Assessment System (LEWAS), is an open-ended guided cyberlearning system
that delivers integrated live and/or historical environmental monitoring data and
imagery. Anonymous user tracking in the OWLS helped to identify students
from various courses as ‘groups of users’ across the world and assisted in
providing information about the importance of various components of the
mediation interface. A pilot test of this tracking capability was conducted in two
first-year engineering courses at Virginia Western Community College during
the fall 2015 semester. During this pilot test, tracking data was collected from a
total of roughly 80 students from a total of four course sections. The data
collected included the amount of time that each student spent using each
component of the OWLS, the paths that he or she used to navigate through these
components and how frequently each student returned to the OWLS. Sugges-
tions for system modifications based on comparison of the time students spent
using various system components with students’ post-test evaluation of the
educational value of these components are included. To address the limitation of
the data collected during the pilot study, which could not identify a user across
different devices, a user login system is being developed for investigating
individualized learning. The current system will address the need to understand
in real-time the learner-specific pathways of content and progression, and these
learners’ levels of engagement within the system.
1 Introduction
2 Literature Review
Engagement can be defined as the interaction between the students and their learning
environment, and the literature emphasizes that students’ learning is directly correlated
to how meaningfully engaged they are in a task [8]. Engagement can have three
interrelated dimensions: behavioral, emotional and cognitive [9]. According to Kuh
[10], time on a task is a construct for measuring engagement, including several other
constructs such as, student involvement, social and academic integration, good prac-
tices in graduate education, etc. For traditional classrooms, engagement has been
measured with class attendance [11], which is the only easily quantifiable visible
indicator of time on task. Attendance or participation in class has also been determined
as an important variable for measuring success [11] and behavioral engagement [9].
Similarly, for cyberlearning systems, interaction with the interface can be a measure for
behavioral engagement [12].
Various approaches for monitoring interface interactions have been used in the past.
Traditionally, observational methods such as observations, video recording, question-
naires, etc. [13, 14] have been used to monitor users’ interactions from an external
viewpoint. However, given that one of the strengths of cyberlearning systems is that
they can be used anywhere at any time, this type of monitoring is often not possible.
Therefore, it is advantageous to store/log the digitized traces of the users from within a
cyberlearning system [15]. These traces can be represented by the time sequence of
actions completed by a user/learner, such as, mouse clicks, typed keys, navigation
through web-pages, etc. [16]. Analyzing large amounts of these traces to digitally track
students’ usage of a cyberlearning tool assists in identifying their preferences and the
bottlenecks faced by each learner [5]. This also aids in identifying the patterns of
zamfira@unitbv.ro
654 D.S. Brogan et al.
enquiry of various students while solving a given problem [17] and in identifying
students’ participation in the learning environment by providing the degree of usage of
the system by the student [18]. This participation explains the level of engagement of
the students with the cyberlearning platform [18] and identifies their study habits [19].
Hence, by implementing a user tracking system within a cyberlearning tool, the two
components of the active learning process, i.e., student activity and engagement of an
individual user [20], can be better understood. These deep insights obtained from user
tracking data allow educators to (1) predict students’ performances, (2) understand the
efficacy of the learning materials, and (3) validate/evaluate the teaching strategies.
These insights will improve the quality of education and lay the foundation for a more
effective education system [21].
The patterns of students’ activities within a cyberlearning environment provide
evidence to cluster similar students and to develop predictive student models or task
models. This information aids in making personalized, intelligent, online learning
systems, which are efficient for engaging various types of learners [21]. For example,
according to the knowledge level of the learners, adaptive tests and personalized
learning materials can be recommended as scaffolding to help learners complete tasks
[22, 23]. Therefore, integrating a cyberlearning system, such as the OWLS, with user
tracking capability is an innovative approach that can be incorporated to advance the
current engineering education system. This addresses the need to understand in
real-time what each learner is doing within the learning environment as well as the need
to know learner-specific pathways of content and progression [4]. This allows
advancement of technology that adapts learning materials based on assessment of an
individual’s learning experiences and his or her level of engagement with the OWLS.
Anonymous tracking to store/log the digitized traces of users was added to the OWLS
to monitor the paths that users take through the OWLS, how long they spend using
various components of the OWLS and how frequently they return. This information
provides insights into the importance of various components of the OWLS. This
tracking was accomplished using Google Analytics. Additionally, the UUID.js library
was used to assign an alpha-numeric universally unique identifier (UUID) to each
device-browser pair. This UUID was stored in a tracking cookie along with the time in
ms. This approach does have some tracking limitations. For example, a single user
accessing the OWLS from multiple browsers on one or more devices will appear as
multiple users because the UUID is assigned to the device-browser pair rather than to
an individual. Other limitations are that users whose cookies are erased after a session
cannot be tracked from one session to the next, that some users may have disabled
tracking cookies completely, and that browser updates often erase tracking cookies.
Due to these limitations, user logins are suggested for future versions.
zamfira@unitbv.ro
Insights Gained from Tracking Users’ Movements 655
Once the data is collected, it must be analyzed, and one challenge in analyzing the data
collected by Google Analytics is to cluster and identify groups of users. This was
accomplished by collecting 15 different pieces of identifying information from each
OWLS page visit, i.e., the UUID, the UTC time at page load in ms, the local URL, the
user’s current country, region (e.g., state) and city, the device category (e.g., desktop,
tablet, mobile), the operating system and version, the browser and version, the screen
resolution, the internet service provider (ISP), the referring website, and if the next
webpage viewed by the user was outside of the OWLS. For analysis, each page view
from October 9, 2015 to May 25, 2016 was assigned a known user ID and a user
group. Known LEWAS team members were identified and separated from other groups
of users. Remaining known users and user groups were identified by noting their
physical location and the use or not of a school ISP. A known user appearing in
multiple locations was assigned to the highest ranking group he or she appeared in from
the list of user groups.
During this period there were 11,231 page views within the OWLS. Figure 1 shows
the number of pageviews per day for the largest user groups. Note that students in
particular courses use the OWLS extensively for no more than a few days before
moving on to other topics but the LEWAS team members (red) have a regular back-
ground presence. Thus, it is important to remove known LEWAS team members’
pageviews from analysis of the intended users.
One of the groups of users in Fig. 1 was a pilot test of roughly 80 students from a total
of four sections of two first-year engineering courses at Virginia Western Community
College (VWCC, purple). Students in these courses participated in the user tracking
pilot test on October 13, 20–21 and 26–30, 2015. A total of 173 device-browser pairs
zamfira@unitbv.ro
656 D.S. Brogan et al.
(UUIDs) were recorded from these implementations. However, one UUID was
excluded from analysis as being that of an instructor who visited the OWLS before
introducing it in various course sections. One challenge to analyzing this data is that the
campus computers that students used erased tracking data at browser exit. This explains
both a lack of multiple sessions for a single UUID and the large number of UUIDs
relative to the numbers of students enrolled in these courses (n = 78). It also prevented
analysis of how often users returned.
Some insights can be gained from the ways that these users interacted with the
OWLS. Approximately 27.9% of these users viewed three pages within the OWLS
before leaving. At the other extreme, 10 users (5.8%) viewed 17 or more pages with the
two longest user paths being 40 and 44 pages before leaving the OWLS. Figure 2
shows two concurrent course sections on the day that they were first introduced to the
LEWAS. The blocks with smaller heights at the right ends of users’ interactions with
the OWLS are of unknown temporal duration and are indicated by a reduced vertical
thickness of constant length. These are users’ final pages in the OWLS. The fact that
several users transitioned between OWLS pages at approximately the same times
suggests that they were following the actions of the instructor on the projected display.
Thus, designed instruction, rather than open-ended exploration, drove which pages
students visited when first introduced to the OWLS. This also accounts for the large
percent of users who viewed three pages in the OWLS before leaving. Accordingly, the
homepage, the interactive graph and the data download page were the three most
visited OWLS pages by the students in these courses.
Fig. 2. Use of the OWLS on October 21, 2015 in two courses at VWCC with users 1-20 clearly
in course 1 and users 26–45 clearly in course 2
In order to evaluate the components and features of the OWLS, the post-tests given to
students in the pilot test at VWCC included an item worded “What was the learning
value of the following components of the OWLS (circle your choices):” followed by
ranking levels for several OWLS features. For the purposes of comparing the features
of the OWLS, the ranking levels were assigned integer numeric values of 1 to 4 for
each group, and these values were averaged for each group to determine scores for each
zamfira@unitbv.ro
Insights Gained from Tracking Users’ Movements 657
feature with a score of 1 being the worst and a score of 4 being the best. The results are
shown in Fig. 3. These results imply that access to real-time data ranks as the most
important of these features followed by the anywhere/anytime access that cyberlearning
systems provide. These data availability features are followed by the data visualization
features (live camera and interactive graphs) and supporting information features (case
studies, overhead view/map and background information).
In order to compare user tracking data with students’ self-perceived learning, the
OWLS web pages were grouped according to the OWLS features. Figure 4 shows the
numbers of pageviews for these features and the corresponding ranks (in red) for those
features students rated for their learning value. Because the students viewed the
Interactive Graph and Live Camera during class time when they were directed to view
them according to the results shown in Fig. 2, they gained increased exposure to these
pages, which may have increased their rankings of these components compared to
features that they did not interact with very much. This is a potential source of bias in
the ranking of OWLS features.
Fig. 4. Number of page views from the full VWCC implementation for each OWLS feature.
The numbers on the right are the rankings of the OWLS features given students in these courses
following “real-time data” (1) and “anywhere/anytime access” (2).
zamfira@unitbv.ro
658 D.S. Brogan et al.
The OWLS user tracking was able to identify groups of users (who each typically used
the OWLS for no more than a few days) from courses at several different institutions
across the globe. It was also able to show how students who used the OWLS during
class time followed the actions of the instructor rather than exploring the system in an
open ended way. Furthermore, this directed use of certain components may have biased
students’ rankings of these features of the OWLS. This tracking also showed how much
time students spent on each page. However, security measures in the computer lab at
VWCC prevented the OWLS from tracking repeat visitors to know how frequently
people returned to the system. Further research is needed to resolve these issues.
Planned future OWLS tracking enhancements include addition of a user login
system with user tracking capabilities. Each user will be identified using his or her
username, and each of their requests to the server will be stored with timestamps in a
database for further analysis. These requests may include navigating to a different web
page or changing the page focus to various features within a page. Storing the tracking
information in a database of LEWAS will aid in securing student information and
addressing the security issues of proprietary products like google analytics. With this
system in place, each student in a class will be tracked while solving a particular
problem using the OWLS. The students’ trace information in the database will be
explored to identify their activity streams and their levels of engagement with the
OWLS while problem solving.
Additionally, if students’ use of the OWLS is directed by the tasks that they are
asked to complete with it, it is possible to create profiles for different groups of users
based on the ways that they use the OWLS. This can be used as training data for future
courses to predict how a particular group of users will interact with the OWLS to
complete a given task. This predictive modeling can be used as a foundation to develop
group-specific interface modifications. After the user login system is implemented, this
process can be adapted to develop a personalized learning environment for each user.
References
1. Alvarez, I.B., Silva, N.S.A., Correia, L.S.: Cyber education: towards a pedagogical and
heuristic learning. ACM SIGCAS Comput. Soc. 45, 185–192 (2016)
2. London, J.S.: Exploring cyberlearning through a NSF lens. In: 2012 Paper Presented at
ASEE Annual Conference and Exposition, San Antonio, Texas. https://peer.asee.org/21371
(2012)
3. Borgman, C.L.: Fostering Learning in the Networked World: The Cyberlearning Oppor-
tunity and Challenge. DIANE Publishing, Darby (2011)
4. Madhavan, K., Lindsay, E.D.: Use of information technology in engineering education. In:
Johri, A., Olds, B.M. (eds.) Cambridge Handbook of Engineering Education Research.
Cambridge University Press, Cambridge (2014)
5. Johri, A., Olds, B.M.: Situative frameworks for engineering learning research. In: Johri, A.,
Olds, B.M. (eds.) Cambridge Handbook of Engineering Education Research. Cambridge
University Press, Cambridge (2014)
zamfira@unitbv.ro
Insights Gained from Tracking Users’ Movements 659
zamfira@unitbv.ro
Practical Use of Virtual Assistants and Voice
User Interfaces in Engineering Laboratories
1 Introduction
provides a development environment and app store where developers can publish and
distribute their voice based applications while the Amazon Web Services Internet of
Things (AWS IoT) platform allows the connection of devices to cloud based services
[5]. These flexible, highly functional development environments and backend archi-
tectures allow the use of the Amazon Echo and similar voice enabled services in a
range of diverse application areas including home automation and education.
Practical engineering laboratories for undergraduate students are evolving incre-
mentally driven by the availability of more affordable instrumentation and hardware
kit, the growth of the Internet, access to virtual and remote laboratories and a move
towards student-centered pedagogies but are at their core fundamentally unchanged [6].
This paper explores the use of virtual assistants in electronic and electrical engineering
laboratories to tutor students; guiding them through experiments; presenting supple-
mentary teaching resources when requested; accessing, controlling and configuring test
instrumentation and hardware and providing feedback through summative and for-
mative assessment. A case study and practical working example of a virtual assistant
approach is demonstrated based on the modification of an existing remote laboratory
and teaching resources for fundamental electronic and electrical engineering circuits
suitable for the first year of an undergraduate degree. The process of integrating test
instrumentation, the board under test, a switching matrix, additional teaching resources
and the Amazon Echo virtual assistant is discussed. Section 2 of this paper provides an
overview of the Amazon Echo platform and the creation of voice user interfaces using
the Alexa Skills Kit. Sections 3 and 4 discusses challenges related to the re-purposing
of an existing laboratory for voice interactions and provides a practical example of this
process. Section 5 looks at the practicalities of implementation, assessment and the
creation of a help system to provide feedback to the student. Section 6 presents the
conclusion and possible future work in this area.
Amazon Echo is a smart speaker with cloud based voice recognition and speech syn-
thesis capabilities based on natural language processing (NLP) algorithms and
speech-unit selection technology [7]. It consists of a cylinder speaker with a seven-piece
microphone array and internet connectivity. The Echo device is capable of processing a
range of voice commands and user interactions (via cloud based services) and can be
used as a home automation hub to access and control smart devices. When operating in
default mode the Echo continuously listens to all speech in its general vicinity and
responds/becomes active when it detects the use of a “wake word” i.e. “Alexa” or
“Echo”. The voice commands interactions detected after activation are sent to the cloud
for processing by the Alexa Voice Service(AVS)/developer services and relevant
responses generated (Fig. 1). Third party developers can create voice experiences/
custom skills that extend the capabilities of any Alexa-enabled device using the Alexa
Skills Kit (ASK) which is a collection of self-service APIs, tools, documentation and
code samples. User created custom skills have an invocation name which is a key word
used by the end user to initiate a set of voice interactions/responses with the Echo
device.
zamfira@unitbv.ro
662 M.J. Callaghan et al.
The voice interactions and responses are defined by an interaction model (Fig. 2)
which manages communications between the parties involved using an intent schema,
slots and sample utterances [8]. Intents are the core functionality of your skill and are a
list of common actions your skill can accept and process. Slots are parameters or values
passed with an intent. Sample utterances specify the spoken words and phrases users
can say to invoke intents.
The architecture of a typical online laboratory allows the user, located in a separate
geographical location, to access, control and conduct experiments remotely (Fig. 3).
The hardware control element is usually facilitated by the use of GPIB (General Purpose
Interface Bus) or similar communication standards connected to a switching matrix
allowing test instrumentation and experimental boards to be selected, connected and
configured during experiments using a software client-server approach [9, 10]. Figure 4
shows a range of experiments and accompanying learning outcomes/objectives for an
existing online remote laboratory which focused on teaching applied fundamental cir-
cuit theory and suitable for students on the first year of an undergraduate electronic and
electrical engineering degree.
zamfira@unitbv.ro
Practical Use of Virtual Assistants and Voice User Interfaces 663
The lab, related teaching resources, test instrumentation and the circuit board under
test were accessible through the web using a client/server approach where the host PC
acts as server, running scripts and hosting a database to store instrumentation config-
urations and general settings. An application running on the host computer and utilizing
the Keysight IO Libraries Suite [11] facilitated access to the physically connected
hardware (Fig. 3) and exchanged data with the client. The approach taken for this
project was to reuse the existing local communication and control protocols but to
remove the frontend web/client based UI and remote access functionality from the
laboratory (Fig. 5) and replace it with a voice driven Virtual Assistant using the
Amazon Echo hardware and built with the Alexa Skills Kit and backend Amazon web
services [12].
Fig. 4. Learning outcomes and test instrumentation for existing remote laboratory
zamfira@unitbv.ro
664 M.J. Callaghan et al.
The Series Parallel laboratory (Fig. 4) was selected in the first instance to develop as a
prototype voice assisted practical to demonstrate the viability of the approach and to
understand the process of re-purposing existing teaching resources and material for use
in this context. The existing laboratory was game and time based where the student was
given values of the input voltage Vin and resistors R2/R3 and a target output voltage Vo
to achieve. The student used the formulas provided for the equivalent resistance Req
and voltage out Vo to calculate the correct value of resistor R1 to achieve the target
output voltage. A score was awarded dependent on how close the calculated value was
to the target value of Vo and the time taken to complete the calculations [13].
When the laboratory was in progress, the game user interface/client communicated
with the underlying hardware, switching matrix and instrumentation (Fig. 3) and
connected the selected value of resistor R1 to physically complete the circuit. The game
was initially designed in Microsoft Excel (Fig. 6) and has a high level of replay value
and longevity as each time it was played, different combinations of resistors R2 and R3
were selected randomly and connected from an existing bank of physical components
resulting in a range of different target values of voltage out to achieve (Fig. 7).
Using the existing laboratory as a starting point the next step in the process was to
create a structured series of interactions suitable for a voice driven experience which
included an overview of the laboratory, access to help if needed, control and config-
uration of the instrumentation and circuits in the hardware layer, assessment and
provided feedback to the student (Fig. 8). This structure was used to create the inter-
action model for the Series Parallel skill and to identify the intents, prompts, slots,
sample utterances and communications with the hardware layer (Fig. 9) which would
then be used for development and implementation using the Alexa Skills Kit.
zamfira@unitbv.ro
Practical Use of Virtual Assistants and Voice User Interfaces 665
Fig. 7. Hardware instrumentation and component values for Series Parallel laboratory
zamfira@unitbv.ro
666 M.J. Callaghan et al.
zamfira@unitbv.ro
Practical Use of Virtual Assistants and Voice User Interfaces 667
zamfira@unitbv.ro
668 M.J. Callaghan et al.
Echo skills are hosted on Amazon web services and require an Amazon developer
account and have two main components, Amazon Lambda and the Alexa Skills kit.
Amazon Lambda is an event driven cloud computing platform which runs and executes
code on demand when triggered. Running an Echo skill requires the creation of a
Lambda function which is executed from an event source (in this case the Amazon
Echo). Skills are created using the Alexa Skills kit and require a name that will be
displayed to users of the Amazon app store and an invocation name which allows the
user to activate the skill through the Echo hardware. Skills can be written in Node.js,
Python or Java using the interaction model part of the interface to tell the skill which
intents (in a JSON structure) it supports and the sample utterances that will trigger each
intent (Fig. 10). The skill is then linked to the Lambda function using the Lambda
Amazon Resource Name (ARN) and can be tested using the service simulator or the
physical hardware device before publishing on the Echo app store.
Using this approach the Series Parallel skill was created and made available on the
physical Amazon Echo device [14] which was placed in the laboratory (Fig. 11). The
student starts the interaction with the Echo using the app invocation name. The Echo
then welcomes the student to the laboratory, provides an overview of the experiment
and sets out their objectives (Fig. 9). It then guides the student through the laboratory
with a series of voice prompts e.g. help provision (formulas for equivalent resistance
zamfira@unitbv.ro
Practical Use of Virtual Assistants and Voice User Interfaces 669
Req and voltage out Vo) and the selection of the value of resistor R1 based on the given
values of resistors R2, R3 and input voltage Vin. The student responses to the voice
prompts and their responses are processed by the Echo/Alexa Skills kit based on the
sample utterances (Fig. 9).
zamfira@unitbv.ro
670 M.J. Callaghan et al.
This paper explored the feasibility of using virtual assistants and voice user interfaces
in campus based engineering laboratories to tutor and assess students. A practical case
study was presented based on the modification of an existing remote laboratory where
the remote access functionality was removed and replaced with a local control system.
The working example shown demonstrated how this approach could be used to guide a
student through an experiment, providing supplementary teaching resources and help
when requested while accessing and controlling test instrumentation and hardware. It
also demonstrated how voice user interfaces could be used for summative and for-
mative assessment and to provide feedback to students. This area of research is set to
grow rapidly as virtual assistants and related devices become mainstream driven by low
cost consumer hardware and cloud based services. Future work on this project will
focus on developing formal, structured approaches to the creation of virtual
assistant/voice user interfaces for engineering laboratories, extending the approach to
all of the practical experiments shown previously, investigating how this approach
could be integrated with existing and widely used remote laboratory infrastructures and
frameworks and exploring possible uses of these technologies to improve accessibility
and access to teaching resources for students with disabilities.
zamfira@unitbv.ro
Practical Use of Virtual Assistants and Voice User Interfaces 671
References
1. Bouziane, A., Bouchiha, D., Doumi, N., Malki, M.: Question answering systems: survey and
trends. Proc. Comput. Sci. 73, 366–375 (2015). ISSN 1877-0509
2. Khillare, S.A., Pundge, A.M., Mahender, C.N.: Question answering system, approaches and
techniques. Int. J. Comput. Appl. 141(3), 34–39 (2016)
3. Microsoft: Conversational intelligence (2016). https://www.microsoft.com/en/mobile/
experiences/cortana/. Accessed 30 Nov 2016
4. Amazon: Started with the Alexa Skills Kit (2016). https://developer.amazon.com/public/
solutions/alexa/alexa-skills-kit/getting-started-guide. Accessed 30 Nov 2016
5. Amazon: Amazon Web Services Internet of Things (2016). https://aws.amazon.com/iot-
platform/how-it-works/. Accessed 30 Nov 2016
6. Feisel, L.D., Rosa, A.J.: The role of the laboratory in undergraduate engineering education.
J. Eng. Educ. 94, 121–130 (2005)
7. Black, A.W., Taylor, P.: Automatically clustering similar units for unit selection in speech
synthesis. In: Proceedings of Eurospeech 1997, Athens, Greece, pp. 601–604 (1997)
8. Amazon: Amazon Echo Interaction Model (2016). https://developer.amazon.com/public/
solutions/alexa/alexa-voice-service/reference/interaction-model. Accessed 30 Nov 2016
9. Lindsay, E., Liu, D., Murray, S., Lowe, D.: Remote laboratories in engineering education:
students’ perceptions. In: Proceedings of 18th Annual Conference Association for
Engineering Education (AaeE 2007) (2007)
10. Lowe, D., Murray, S., Lindsay, E., Liu, D.: Evolving remote laboratory architectures to
leverage emerging internet technologies. IEEE Trans. Learn. Technol. 2(4), 289–294 (2009).
doi:10.1109/TLT.2009.33
11. IO Suite: Keysight IO (2016). http://www.keysight.com/en/pd-1985909/io-libraries-suite?
cc=GB&lc=eng. Accessed 30 Nov 2016
12. Amazon: Amazon web services (2016). https://aws.amazon.com/. Accessed 30 Nov 2016
13. Callaghan, M.J., McCusker, K., Losada, J.L., Harkin, J., Wilson, S.: Using game based
learning in virtual worlds to teach electronic and electrical engineering. IEEE Trans. Ind.
Inform. 9(1), 575–584 (2013)
14. Amazon Alexa/Echo demo: Practical lab for Series Parallel circuit (2016). https://www.
youtube.com/watch?v=tDpx9jZ-WKs&spfreload=10. Accessed 30 Nov 2016
zamfira@unitbv.ro
Approaching Emerging Technologies: Exploring
Significant Human-Computer Interaction
in the Budget-Limited Classroom
James Wolfer(B)
1 Introduction
There has been an explosion of sensor, presentation, and display technology
available for exploration in Human-Computer Interaction. While much of this
technology is readily available, approachable, and/or inexpensive, such as cell
phone or Web display, other technology remains relatively expensive in the con-
text of classroom instruction. This is especially true for classes that are only
taught occasionally at any given institution.
One class that has particular interest in emerging technologies is the Human-
Computer Interaction course. Designed to explore the requirements acquisition,
design, development, deployment, and outcomes measurement of user interfaces
the class blends the assessment of software-driven interfaces such as web pages
an e-commerece sites with emerging hardware interfaces [1]. This work describes
a curated collection of hardware and software designed to provide a laboratory
experience exposing the principles of select emerging technologies to computing
students. By carefully choosing the devices used to cover important technologies
we provide a significant laboratory experience within budget constraints.
c Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6 63
zamfira@unitbv.ro
Approaching Emerging Technologies 673
2 Brain-Computer Interfacing
One of the emerging technologies that holds potential to have impact in areas
ranging from entertainment to medicine is that of Brain-Computer Interfacing
(BCI). Historically the purvue of medical research, recently BCI capability has
become approachable at the consumer level. There are several BCI Interfaces
available commercially at this point in time. Perhaps the most prominent is the
Emotiv line of interfaces [2]. While capable, the Emotive series of devices are
relatively expensive. As an alternative we elected to work with the NeuroSky
Mindwave electroencephlograph (EEG) sensors [3].
Figure 1 shows the NeuroSky Mindwave. While limited to a single sensor,
this lightweight headset is capable of collecting raw EEG output between three
and 100 Hz with a sampling rate of 512 Hz. In addition, the Mindwave produces
EEG power spectrums for commonly defined bands such as Alpha, Beta, Delta,
etc. Quality analysis data is also available. The device interfaces to the computer
via Bluetooth, and is powered by a single AAA battery.
NeuroSky provides SDK’s for most major operating systems, and the open
source community provides several alternatives for language and device interfac-
ing [4]. In addition, there are a variety of applications that can readily capture
the raw data. One such application, for the Android system, is eegID, by Isomer
Programming, LLC [5]. This application acquires a comprehensive set of data
including signal quality, raw EEG values, EEG values in volts, NeuroSky propri-
etary data such as attention level, meditation level, and a range of EEG signal
zamfira@unitbv.ro
674 J. Wolfer
3 Haptics
Another fundamental technology for interacting with the computer, haptics,
exploits the sense of touch. Perhaps the most common experience of haptic feed-
back is that of the vibrational cellphone ring signal. More sophisticated interfaces
provide a realistic sense of touch, and can be used to model phenomoma ranging
from spring-mass demonstrations to arterial pulse palpation [6,7]. Using haptics
in an educational setting has been described in [6].
There are many haptic devices available. With six degrees of freedom, the
Sensable Phantom [8] series of devices are among the most popular for research
purposes, are very capable, but are correspondingly expensive limiting their
availability for occasional classroom use. A variety of projects have been used to
provide less expensive alternatives. For example, Carniero et al., describe a one
degree of freedom haptic device suitable for educational activities [9].
For much of our haptics work we elect to use a Novint Falcon [10], a haptic
game controller. The Novint Falcon, shown in Fig. 3 provides three degrees of
translational freedom, has a maximum force of over nine Newtons, and can
sustain a refresh rate of 1000 Hz. In addition to the Novint drivers and SDK,
there are open source alternatives that allow programming the device across a
variety of computer and operating system platforms [11]. In addition, high-level
scenegraph tools are available supporting rapid prototyping with the Falcon [12].
Student projects involving the Novint Falcon include using it as the basis for the
haptic representation of aortic surface and pressure profiles [7,13].
zamfira@unitbv.ro
Approaching Emerging Technologies 675
Augmented and Virtual Reality are approaching mainstream, with Oculus pro-
viding high-end hardware to the consumer market [14]. As a less expensive alter-
native for classroom use we adopt a Google Cardboard [15] approach. Leveraging
the incredible power of the modern smart phone we provide a series of experi-
ments and demonstrations of stereographic, or 3D, content.
The smart phone, mounted in an optical container such as that shown in
Fig. 4a can is used to project 3D images as well as sound from the smart phone
speakers. There are a variety of applications online that take advantage of this
arrangement to provide a VR experience at moderate cost [16]. In addition to
entertainment, projects like RADSVRx represent a serious effort to deploy this
technology for cost effective medical training [17].
Custom content creation is another aspect of such a system. Developing syn-
thetic content is relatively straightforward using standard development tools for
computer graphics. An example, rendered in red-cyan anaglyphic to make it
suitable for viewing with colored glasses, of an artery extracted from computed
tomography is shown in Fig. 5a.
Custom content is more challenging for photographic material. Recently,
small “toy” beam-splitters have become available to clip onto cell phone
zamfira@unitbv.ro
676 J. Wolfer
Fig. 4. VR examples
cameras for just such purpose as shown in Fig. 4b [18]. Figure 5b shows a side-
by-side stereographic image, along with the corresponding anaglyphic image
(Fig. 5c) of Colossus, the worlds first programmable computer at The National
Museum of Computing in the UK [19]. Viewing the side-by-side images on a
smart phone cardboard interface provides a sharp, color, image of the computer
in depth.
The smart phone also forms the basis for projects in augmented reality, com-
bining the synthetic and real environments. Often used to augment human capa-
bility, student projects include geographic and structural mapping of the cam-
pus providing interactive identification of buildings and other structures while
walking.
We now turn to some of the readily available interfaces for general interaction.
Of course there are many that are used daily and taken for granted. Examples
include the keyboard, mouse, and trackpad provided with modern computers.
Other examples include, touch surfaces on tablets and mice, as well as a variety
game controllers with impressive capabilities, such as the Microsoft Kinect. In
this section we profile a series of devices less familiar to the students.
It could be argued that the cell phone is the most successful HCI device in history.
In addition to basic communication capability, the modern cell phone features
a host of sensor information readily available for subsequent analysis. Examples
include location, orientation, acceleration, rotation, pressure, proximity, light,
sound, and magnetism. In addition, many cell phones have displays with resolu-
tions rivaling computer displays. The sensor capabilites of cell phones have been
used for applications ranging from navigation to mental health assessment [20].
There are a variety of approaches to access the sensor data from cell phones.
For students adept at software development programs can be written to extract
zamfira@unitbv.ro
Approaching Emerging Technologies 677
(a) CT Artery
Fig. 5. VR examples
the data. This has the advantage of complete control of the application, with
the disadvantage of long development times. In the interest of obtaining data
quickly for post-acquistion processing and analysis we elected to use existing
phone apps to extract the data. One such app for Android, AndroSensor [21],
records all of the sensor data to a comma-separated value (csv) file suitable for
import into a spreadsheet or program for subsequent processing. A sample plot
is shown in Fig. 6.
zamfira@unitbv.ro
678 J. Wolfer
When evaluating the keyboard, students liked the idea of the projection
keyboard, but they viewed it as a novelty. They found it awkward to type at
accustomed speed due to lack of expected mechanical feedback and they found
the audible key-clicks helpful, they also found them irritating with time. Thus
the projective keyboard filled a vital role demonstrating a contemporary device
with significant limitations as an object for their first critical review.
The Leap Motion Controller (Fig. 7b) is a gesture capture device that sensi-
tive enough to acquire the position of fingers on the hand within it’s view [22].
Leap provides an SDK to support software development on the device which
makes the device extremely useful for advanced students. Examples provided by
Leap include interactive Chess, using the hand in more-or-less natural motions,
zamfira@unitbv.ro
Approaching Emerging Technologies 679
such as pinch and release, to move individual game pieces. Student projects using
the Leap Motion Controller used the SDK to develop interactive grasping and
moving in a mouse-like interface.
6 Instructional Use
As indicated in the description of the technologies contained in the collection
supporting the HCI class, many of these technologies supported student projects
with duration ranging from a single semester to multi-semester undergraduate
re- search. In addition to their use in projects, the technology served several
ongoing instructional roles. Specifically, the served as the basis for classroom
demonstrations with interactive exploration of the strength and limits of each
technology followed by discussion on limitations and improvements.
The emerging technologies described here also formed the basis for a series of
individual assignments. For example, in HCI usability design principles form an
important component of the discipline. Encapsulating both theory and empirical
observations, these principles form an assessment basis. Over the course of a
semester each of the devices and technologies are assessed by students in terms of:
7 Conclusion
This work has described the assemblage of a carefully curated set of devices
representing a selection emerging technologies appropriate to form the hardware
basis for a human-computer interface course. When combined with appropriate
design principles, targeted assignments, and sustained projects they combine to
provide significant experience at a reduced cost when compared with research-
grade equipment.
zamfira@unitbv.ro
680 J. Wolfer
References
1. Preece, J., Sharp, H., Rogers, Y.: Interaction Design. Wiley, Hooboken (2015)
2. EMOTIV Brainware. http://www.emotiv.com
3. NeuroSky. http://neurosky.com
4. Singh, S.: NeuroPy. http://neurosky.com
5. Isomer Programming: eegID. http://www.isomerprogramming.com/downloads/
android-apps/eegid
6. Restivo, M.T., Cardoso, A., Lopes, A.M.: Online Experimentation Emerging Tech-
nologies and IoT. IFSA Publishing, Barcelona (2016)
7. LeClair, A., Wolfer, J.: Haptic representation of aortic pressure waveforms using
synthetic ECG derived time intervals. In: Proceedings International Conference on
Online Experimentation (2015)
8. Sensable Phantom Omni. http://www.dentsable.com/haptic-phantom-omni.htm
9. Carneiro, F., Quintas, P., Abreu, P., Restivo, M.T.: Design and test of a 1 DOF
haptic device for online experimentation. Int. J. Online Eng. 12(4), 55–57 (2016)
10. Novint: Falcon. http://www.novint.com/index.php/novint
11. Danieau, F.: Libnifalcon. https://github.com/libnifalcon/libnifalcon
12. Open Source Haptics: H3D H3DAPI. www.h3dapi.org
13. Gordon, S.L., Wolfer, J.: Vascular-haptic interaction: a student project case study
in computer graphics. In: International Conference on Engineering and Computing
Education, March 2009
14. Oculus: Oculus rift. https://www.oculus.com/
15. Google VR: Google cardboard. https://vr.google.com/cardboard/
16. Wake County Public Schools: Google cardboard virtual reality. http://www.
wcpss.net/cms/lib/NC01911451/Centricity/Domain/3791/Google%20Cardboard
%20VR.pdf
17. Hernandez-Rangel, E.: RADSVRX, Radiology Advanced Educational System with
Virtual Reality Experience. http://www.alexandriavr.com/radsvrx
18. Amazon: 3D cellphone camera lens. https://www.amazon.com/Universal-Mobile-
Camera-iPhone-Samsung/dp/B01CP2VFPE/ref=pd lpo 107 tr t 3? encoding=
UTF8&psc=1&refRID=FQB721ASYKJZ8GPXVMZB
19. The National Museum of Computing. http://www.tnmoc.org
20. Saeb, S., et al.: Mobile phone sensor correlates of depressive symptom severity in
daily-life behavior: an exploratory study. J. Med. Internet Res. 17(7), e175 (2015)
21. Fivasim: AndroSensor. http://www.fivasim.com/androsensor.html
22. Leap Motion: Leap motion controller. https://www.leapmotion.com/G
zamfira@unitbv.ro
Touching Is Believing - Adding Real Objects
to Virtual Reality
Abstract. This article presents the idea of adding real objects represen-
tations to Virtual Reality as a way to improve the immersive experience.
To this end, a low-cost hand tracking device and an instrumented cube
based on the use of inertial measurement units is presented. Some pre-
liminary results that show the use of the hand tracker for the animation
of a virtual hand model are shown. The fusion of inertial measures with
a vision-based marker detector outputs will be performed with the help
of a Kalman filter to provide smooth and bias corrected estimates of
the object pose. The developed solutions offer the flexibility of interact-
ing either with local or remote systems, as they have been designed as
wireless Internet connected objects.
1 Introduction
Virtual Reality (VR), thanks to the recent introduction on the market of low cost
visualization devices, has attracted consumers, companies, researchers, media
producers, game developers, etc., who are seeing opportunities for the most
various application fields.
Besides the interest in content creation, there have been two main areas of
development for supporting VR, which are displays and motion trackers. On
one side the display devices try to achieve the best quality in terms graphics,
but also in terms of capturing the view point motion that result from the user
movements. As most head mounted displays (HMD) are still linked to computers
through wired connections, this restricts the movement freedom of the users. The
increasing power of portable and wearable computers are making possible to use
wireless HMDs with embedded computing capabilities, that by consequence do
not impose such movement restrictions. This has resulted in various devices
that are being released by companies such as HTC,1 Oculus,2 Samsung,3 or
Microsoft.4 These products and their predecessors have pushed researchers to
1
HTC and Valve virtual reality systems: www.htcvalve.com.
2
Oculus virtual reality systems: www.oculus.com.
3
Samsung VR: http://www.samsung.com/global/galaxy/gear-vr/.
4
Microsoft Hololens: https://www.microsoft.com/microsoft-hololens/en-us.
c Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6 64
zamfira@unitbv.ro
682 P. Menezes et al.
create Virtual Environments Applications and Toolkits, for example [9,10], for
various purposes.
Naturally, several schools and educators are enthusiastically trying to under-
stand how to explore it as an effective learning tool. In fact these technologies
may create opportunities for the students to improve or reinforce the learn-
ing of specific subjects, for example using virtual laboratories, enabling them
to practice at home before or after performing the real laboratory experiments.
Besides the new rehearsal possibilities that VR opens for preparing the students,
there are experiments which may not be performed directly due to the associ-
ated dangers, costs or impracticalities. In some of these cases simulation may
be a solution or remote control of the experiment may be the only possible or
acceptable solutions.
By coupling realtime simulations with 3D graphic engines, it is possible to
make the student visually observe the experiments evolution by mocking the
real systems. In the case of remotely controlled experiments, traditionally VR-
related devices may also enable us to explore the idea of telepresence in tele-
manipulation or teleoperation tasks. This brings the student to the center of the
task to be performed, using as one of two different approaches with respect to
the way the user perceives himself: (1) an embodied agent that sees by himself
and acts using some robotic body, or as (2) an embarked pilot that controls the
device, robot, or vehicle from the inside. Both ideas share the principle that the
user sees him/herself as being present at the remote or virtual place. The main
differences come from the way that he/she perceives the physical body and they
way its actions are controlled.
For the embodied case, the physical (robotic) or virtual body must be seen
as the own body, in a way that the nearby elements must be perceived as being
proximal to one’s body. The control of the (acquired) body must be intuitive and
automatic, or in other words for controlling it if suffices to think about doing
and not about acting upon some input (button, joystick) for doing it, i.e. not
requiring any kind of high-level reasoning. The embarked pilot case, is slightly
different in the perception it provides. The physical body is no longer perceived
as one’s body, but instead it is seen as a vehicle that the operator controls
from the inside, like a car, a plane, etc. The perception of being inside of a
vehicle, on some kind of virtual cockpit, with the possibility of looking in most
frontal directions as needed, enable the reuse of the mental models constructed
for driving cars, as an example. In these cases, the operator can see him/herself
driving the vehicle through some set of controls, which may include steering
wheels, joysticks, or other, as appropriate. Therefore, parts of the vehicle no
longer need to be perceived as the own body. But perceiving the own body
and its spacial and contact relations with the cockpit devices are required to
make use of the those mental models. In previous works developed at ISR-UC,
both embodied and embarked approaches were developed and evaluated with
convincing results [1,9], where it became clear that the need to perceive the
spatial relationships, namely the position of the controls with respect to the
hands of the operators, is of vital importance for an intuitive operation. This
zamfira@unitbv.ro
Touching Is Believing - Adding Real Objects to Virtual Reality 683
takes us to the need to the need to track the operator’s hand, and eventually
body, postures or even the related 3D point clouds to generate its representation
with more or less fidelity, but precise enough to induce the required presence
perception. Recently some new products have been introduced in the market
with the capability of being tracked and represented in the virtual environment
like the controllers like Reactive Grip5 or full body suits like HaptX Skeleton6 .
The motion tracking methods based on accelerometer, gyros or magnetic
sensors, either individually or combined, are well documented in literature, with
their use mostly restricted to orientation tracking due to the drifting problems of
position estimation using these sensors. On another side, vision-based tracking
of objects and markers has also its limitations in terms of occlusions and rate
especially when dealing with low-cost solutions that do not make use of synchro-
nous active markers and or high-speed cameras. This paper presents a low-cost
solution that makes use of the fusion visuo-inertial sensor data to enable the
tracking of instrumented objects to be included in the virtual reality scenarios.
This aims at fulfilling an important missing feature in most immersive setups,
which is the ability to touch the objects the user is interacting with.
Fig. 1. Left: hand tracking device animating a hand model. Right: user hands repre-
sented by point cloud for the control of an underwater robot.
5
Tactical Haptics Reactive Grip available at: http://tacticalhaptics.com/products/.
6
AxonVR HaptX Skeleton available at: http://axonvr.com/.
zamfira@unitbv.ro
684 P. Menezes et al.
One solution can be the use of a cockpit that completely mapped into the
VR scene, along with a user body representation so that he may perceive himself
approaching and touch its elements. Finally, there is one final detail that needs
to be handled to fulfil all the necessary elements for a complete sense of presence,
which is the own body parts perception inside the VR representation, notably
hands and forearms. Here there are two possibilities, both represented in Fig. 1:
(1) use models to represent these body parts, or (2) add his own body parts
representations to the view. For the first case, we need to be able to track not
only the hands, but also the fingers so that we can consistently animate the
hand models . For the second approach we can use some 3D capturing device
(e.g. KinectTM or StructureTM ) to generate the necessary point cloud, that after
the appropriate segmentation may be used to represent those body parts.
Fig. 2. Left: Wireless Internet connected hand tracker device. Right: hand model with
virtual bone structure.
Direct kinematic equations are used to provide the position and orientation
of each finger tip.
Ei = b Tw × w Tf bi × f bi Tti (1)
where Ei is a matrix containing pose (position and orientation) of the fin-
gertip (i = 1,2,...,5), b Tw is the transformation between the wrist, w, and the
base, b, reference frame, which is typically the forearm, and w Tf bi is a matrix
that represents the transformation between wrist and the ith finger reference
frame, f bi , and finally f bi Tti the transformation between the ith finger base and
corresponding fingertip, ti .
zamfira@unitbv.ro
Touching Is Believing - Adding Real Objects to Virtual Reality 685
With the proposed hand model, and in order to obtain the pose of each
joint, in terms of hardware, the ideal solution would be to place an inertial
measurement unit (IMU) in each one and read the corresponding pose as the user
moves his hands around. This would result in a solution with 16 IMUs, making
the final prototype complex and probably not comfortable to use. Therefore,
in order to decrease the amount of sensors placed on the hand, a study was
performed [3] on the most common hand gestures and motions and the impact
these have on finger articulation. In this way we determined that, we are able to
use only one sensor for each finger without losing enough gestures and motions
that would disrupt the immersion, resulting in a solution with only 6 IMUs.
The developed prototype (Fig. 1) has two main goals, low production cost and
wireless capabilities (not constraining user motions). In the experimental studies
undertaken, we were able to obtain a Root-Mean-Square Error (RMSE) values
of 0.0586 (quaternion units) and 0.9585◦ (Euler angles).
Inertial tracking systems calculate the relative change of a moving target in posi-
tion, Δpn , and orientation, Δqn , between two consecutive sampling times, from
acceleration and angular velocity given by the IMU. Several pose representa-
tions are present in inertial navigation, we use the standard Euclidian coordi-
nate system for position, p, and quaternions [2] for orientation, q. By solving
the quaternion differential equation [2] with Euler method we end up with the
quaternion update equation:
1
qn+1 = qn ⊗ (1 + Ωn δt) (2)
2
where Ω = [0, ω1 , ω2 , ω3 ]T is the quaternion representation of the angular
velocity in the moving frame, δt is the sample interval and ⊗ denotes quater-
nion multiplication. Double integration of the kinematic motion equation with
reference frame position, p, and acceleration, a, leads to the position update
equation:
1
pn+1 = pn + vn · δt + · an−1 · δt2 (3)
2
Considering the acceleration, am , is measured in the moving frame, the accel-
eration must be transformed into the reference frame, before the position update
calculation. In addition, a gravity compensation is needed since the accelerom-
eter is sensitive to it. This leads to the following acceleration update equation:
an = qn ⊗ am ∗
n ⊗ qn − g (4)
zamfira@unitbv.ro
686 P. Menezes et al.
Fig. 3. Left: Wireless Internet connected cube with IMU, touch sensor and fiducial
markers. Right: diagram of sensor fusion.
4.1 Results
Figure 4 shows a of the fusion of the IMU and a visual marker-based pose esti-
mation. As expected the IMU tends to accumulate errors due to poor sensor
bias estimation, but can provide data at a much higher rate than a convencional
video camera. The visual-marker pose estimation is very stable during static
poses or, slow movements, but fails when the markers are occluded or become
blurred during fast movements. This represents a perfect combination between
zamfira@unitbv.ro
Touching Is Believing - Adding Real Objects to Virtual Reality 687
the two techniques as the case where one has weaknesses is the one where the
other presents the best results.
Using the wireless touch sensitive cube, presented in Fig. 3, in a similar way as
the isometric objects, like Spacetec Spaceball TM presented in [7], the perception
of the compliance of objects presented in the virtual environment can be altered.
This perception modification can be achieved by amplifying or diminishing the
ratio between the force applied and the visual perception experienced through
the HMD, leading the subject to perceive the object as a stiffer or softer object.
The manipulation of this perception can be used in many educational appli-
cations, for example, while studding classical mechanics.
Pos. X
55
Camera
IMU+Camera
50
Pos(mm)
45
40
6 7 8 9 10 11 12 13
time(s)
Pos. Y
120
Camera
100 IMU+Camera
80
Pos(mm)
60
40
20
0
6 7 8 9 10 11 12 13
time(s)
Pos. Z
500
Camera
IMU+Camera
480
Pos(mm)
460
440
420
6 7 8 9 10 11 12 13
time(s)
Fig. 4. Pose estimation using the fusion of a visual marker pose extraction and IMU
pose estimation
zamfira@unitbv.ro
688 P. Menezes et al.
References
1. Almeida, L., Patrão, B., Menezes, P., Dias, J.: Be the robot: human embodiment
in teleoperation driving tasks. In: Ro-Man 2014: The 23rd IEEE International
Symposium on Robot and Human Interactive Communication, Edinburgh (2014)
2. Chou, J.C.K.: Quaternion kinematics and dynamic differential equations. IEEE
Trans. Robot. Autom. 8(1), 53–64 (1992)
3. Cobos, S., Ferre, M., Ortego, J.: Efficient human hand kinematics for manipulation
tasks. In: Direct, pp. 22–26 (2008)
4. Dominjon, L., Lécuyer, A., Burkhardt, J., Richard, P., Richir, S.: Influence of
control/display ratio on the perception of mass of manipulated objects in virtual
environments. In: IEEE Proceedings Virtual Reality, pp. 19–26 (2005)
5. Gustus, A., Stillfried, G., Visser, J., Jörntell, H., van der Smagt, P.: Human hand
modelling: kinematics, dynamics, applications. Biol. Cybern. 106(11–12), 741–755
(2012)
6. Kato, H., Billinghurst, M.: Marker tracking and HMD calibration for a video-
based augmented reality conferencing system. In: Proceedings 2nd IEEE and ACM
International Workshop on Augmented Reality, pp. 85–94 (1999)
7. Lécuyer, A., Coquillart, S., Kheddar, A., Richard, P., Coiffet, P.: Pseudo-haptic
feedback: can isometric input devices simulate force feedback? In: Proceedings of
the IEEE Virtual Reality Conference, pp. 83–90 (2000)
8. Pretto, A., Grisetti, G.: Calibration and performance evaluation of low-cost IMUs.
In: 18th International Workshop on ADC Modelling and Testing (2014)
9. Sanchez, J.G., Patrão, B., Almeida, L., Perez, J., Menezes, P., Dias, J., Sanz, P.:
Design and evaluation of a natural interface for remote operation of underwater
robots. IEEE Comput. Graph. Appl. 11(99), 1 (2016)
10. Sanfilippo, F., Hatledal, L.I.: A fully-immersive hapto-audio-visual framework for
remote touch, November 2015
11. Welch, G., Bishop, G.: SCAAT: incremental tracking with incomplete information.
In: Computer Graphics and Interactive Techniques (SIGGRAPH), pp. 333–344
(1997)
zamfira@unitbv.ro
The Importance of Eye-Tracking Analysis
in Immersive Learning - A Low Cost Solution
1 Introduction
Immersive learning systems are attracting the interest of may people and natu-
rally that of teachers and students. The most appealing features of these systems
are, on the one hand, the interactive level of control that the trainee/student has
on the virtual environment, and on the other hand, the amount of information
that it can provide about the user, which can be recorded for later analysis or
processed in real-time if needed. To acquire such information a number of sen-
sors may need to be installed in the surrounding space, on the user’s body, or
both. For the systems where the user can freely move in space, a combination of
body tracking system and head mounted display (HMD) is typically required,
to provide accurate data of user’s body and head movements and poses. For
other systems where the user is sitting down performing the experiment, only
the HMD’s information may be enough provide data related with the evolution
of the user’s head position and orientation along time. However, in order to track
the user’s gaze direction, special purpose modifications to the HMD are required
and this is normally a challenge due to spatial constrains, illumination artifacts
caused by the displayed scenes, etc. In most learning scenarios it is very impor-
tant to track the user’s gaze direction inside the virtual environments, because of
its importance to the instructor/teacher, as gaze patterns may help to determine
if the trainee/student struggled to understand the subject, and act accordingly,
or if he just payed attention to the important details in the right time. Some
examples of the use of this can go from adjusting the displayed information, i.e.
giving extra information to users who seem to be struggling [7], to providing the
c Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6 65
zamfira@unitbv.ro
690 P. Menezes et al.
zamfira@unitbv.ro
The Importance of Eye-Tracking Analysis in Immersive Learning 691
It is common sense that vision plays a central role in any interaction activities
developed by humans and most animals. One of the reasons is that through
vision we can learn very different types of information about a subject without
reaching and touching it. In addition it enables us to sneak a glance at some
interesting object and continue performing some main task without interruption.
Let’s consider an example: While driving a car, we keep our eyes on the road,
but as some speed limit sign appears we need to check if our current speed is
below the limit or not. For this we peek at the speedometer on the dashboard to
get the intended information, but return immediately the attention to the road.
The design of panels and dashboards is normally guided by the knowledge
of the human vision related cognitive processes. This influences the choice of
the distribution of instruments, or the way to attract the attention to the right
information at the right moment.
In complex systems the number of gauges, together with the complexity of the
task itself, may render the operator activities very stressful and demanding. In
many of such cases it is not viable to have beeps or blinking lights to attract the
attention to the information displays, as they tend to turn into noise that may
contribute to degrading the driver performance, and lead to a desensitisation
process that makes the operator to stop noticing them.
For this reason during the training process the operators/pilots must learn
some mandatory procedures which include the periodic verification or reading of
some displays or gauges in predefined sequences. This is typically important to
confirm if the related quantities are within acceptable limits or if some corrective
operation must be executed.
During normal training sessions, the instructor may observe directly if the
trainee does, as expected, the periodical checking of the important displays, or
consults the appropriate one prior to do some action or manoeuvre, as a way to
choose the appropriate commands or verify the related safety.
zamfira@unitbv.ro
692 P. Menezes et al.
pose, unlike in the real world, where a pilot or operator can see most, if not
all, of the dashboard, just by moving the eyes left and right. The use of current
HMDs still imposes the rotation of the head to visualise the remaining parts, as
a result a trainee may easily forget about reading what is out of sight. (2) The
HMD occludes the eyes of the trainee, so the instructor cannot see them. This
indeed does not allow an instructor to know where the trainee is looking at. As
a result it is not possible to know if the appropriate instrument readings were
done when they were supposed to happen.
The first mentioned limitation could be overcome from using higher resolution
displays, but these impose higher demands on the graphics unit and required
bandwidth. As already applied in some HMDs for reducing the cybersickness
effects, the detail of the peripheral regions is reduced. But the peripheral region
is defined with respect to the display center and not to the gaze direction. This
is called foveated rendering and some companies have been working in it [5], still
considering very high resolution displays receiving full resolution video, even if
a great part of it is blurred. An interesting solution would be to reduce the
dimensions of the generated rasterizations and vary their position on the screen
so that they are always centred on the user gaze direction. From this it becomes
obvious that the introduction of an eye-tracking mechanism would serve as a
basis for solutions of both of these problems. Hereafter we will focus on the
second case.
4 Eye-Tracking in Learning
An eye-tracker suitable for use inside an HMD can be a valuable tool for VR-
based training processes. For sessions conducted by an instructor, he/she can
have a view of what is the focus of attention of the trainee in every instant in
time. This can be used, for example, to understand if there are some elements
that take longer for the student to understand that require some intervention or
explanation from the instructor [1].
With the help of this, the instructor can also observe, either directly, with
the help of heat maps, or another method, if the trainee has done the required
readings of the instruments at the appropriate times or in the right sequence
before some particular actuations, as is defined on the procedures to be learnt.
In the cases that the system is to be used to practice in an unsupervised
manner, the procedures to be learnt may be used to generate a set of rules
that will serve to evaluate the trainees performances. Here, the gaze information
can be used to automatically check if the verifications were done properly, and
if the required gauges were read as expected prior to some actuations. This
enables the identification of different cases: (1) The trainee did not read the
necessary gauges/displays and by consequence did not perform a required action,
performed the action not correctly, or even performed the action correctly by
chance. (2) The trainee did the required readings but has chosen the wrong
actions or incorrect way of doing them. (3) The trainee did the required readings
and used them to choose the appropriate actions.
zamfira@unitbv.ro
The Importance of Eye-Tracking Analysis in Immersive Learning 693
Several researchers and developers have acquires Oculus Rift pre-release devel-
opment kits DK1 or DK2, to anticipate the development of new applications for
these devices or research purposes. When compared with those that preceded
them, these devices have very interesting characteristics in terms of field of view
and development support. The configuration of the device, in terms of available
space and the detachable lenses, creates a perfect opportunity for introducing
the necessary modifications to support an eye tracking solution.
A preliminar analysis of the characteristics of the device and the typical usage
conditions was necessary to elaborate a solution that respects the requirements
of introducing none or minimal interference to the user during usage.
To get an adequate view of each of the user eyes, a pair of cameras was
installed under the respective lenses, in a position that minimises the occlusion
of the viewed scene. And in order to provide good illumination to the user’s eyes,
infrared illumination was chosen, that facilitates the segmentation procedure by
increasing the contrast between the pupil and the iris (dark pupil effect). To place
the cameras on the HMD a spacer was designed and produced using 3D printing.
Figure 1 shows the assembled the lens, spacer and the infrared illumination ring,
and the details of the parts used. It should be noted that the chosen cameras
are low cost endoscopy-like ones, which as in most cameras on the market, came
with an IR suppressing filter. This had to be removed, and in turn replaced with
one that attenuates the visible light reflected on the eyes, while allowing the IR
light to pass.
Fig. 1. Left: Camera board and sensor with 90◦ bent connections; Lens; IR filter (film);
Lens on spacer with IR ring; Spacer with camera below; Right: Oculus Rift DK2 with
left lens out, showing spacer and camera installed; on the right side we can see the
illumination ring in mounted on the respective lens.
zamfira@unitbv.ro
694 P. Menezes et al.
The resulting setup can be seen on the right side of Fig. 1, where the left
illumination ring and lens was left out to show the respective spacer and camera
in place. It can also be seen the cabling of the cameras and illumination ring.
Using the installed hardware it’s now possible to have clear near-IR images
of each of the user’s eyes. These images are initially filtered using a low pass
Gaussian filter. A common approach is to perform pupil segmentation applying
a binarization on grey-level images, which tend to compress to the same range,
distinct zones that may produce different responses on the various colour chan-
nels of RGB cameras but tend to have the same luminance levels. This typically
leads to the generation of artifacts that need to be removed using morphological
operations or other.
A careful analysis of histograms obtained from the 3 colour channels showed
that (as expected) the red channel could provide the best segmentation results,
due to the proximity to the IR band.
After the segmentation, the contours of the pupil are extracted by finding
the chain of connected border pixels. Due to the eye movements the shape of
the pupil area can vary from a circular to an elliptic one. Therefore the use
of an ellipse fitting algorithm enables the extraction of the pupil center at the
intersection of the two axes. As no instantaneous variations of the threshold
value are expected in normal conditions, failure to detect the elliptical region
indicates the occurrence of a blink. The basic algorithm is described in Fig. 2.
Fig. 2. Top: Block description of the processing algorithm; Bottom: Example result of
detection of eye pupil and ellipse fitting
As the interest of the use of this type of eye tracker is to know to which region
of the scene the user is looking at, there is a need to obtain a mapping between
coordinates of the observed pupil center and the screen coordinates of the point
the user is looking at.
zamfira@unitbv.ro
The Importance of Eye-Tracking Analysis in Immersive Learning 695
S = f (P), (1)
where S = (Sx , Sy ) are screen coordinates of the point the user is looking at and
P = (Px , Py ) are the corresponding image coordinates of the pupil center. Note
that different mapping functions will be obtained for the left and right eyes.
The calibration of the eye tracker for obtaining a mapping function between
eye- and screen-coordinates was done using both a linear and a polinomial
approaches. The linear mapping consists in reducing the above function to
S = FP, where F is a 3 by 3 matrix, whose coefficients can be estimated using a
least squares solution. The calibration errors were then compared for the linear
global, linear per quadrant, and polynomial cases. The following tables shows
the obtained values for the errors in pixels between the target locations and
those obtained via the mapping functions.
The difference between the linear and the linear per quadrant is that the
for the second the display region is divided in 4 quadrants producing 4 separate
mapping matrices.
zamfira@unitbv.ro
696 P. Menezes et al.
These results show as expected that the global linear mapping presents the
worst results, but when aplied locally (per quadrant) can improve substancially.
The polynomial approach shows as expected much better results. The reason is
clearly understood from the fact the iris locations are obtained from the projec-
tion of a point from an approximately spherical surface onto a plane through a
projective projection.
Using the obtained mapping functions it is now possible to infer the region
of the scene the user is looking at, as can be seen in the example shown in Fig. 3.
Fig. 3. Example of the output of eye tracker (red cross) mapped onto a plane cockpit
shown to the user.
References
1. Ha, J.S., Byon, Y.J., Baek, J., Seong, P.H.: Method for inference of operators’
thoughts from eye movement data in nuclear power plants. Nuclear Eng. Technol.
48(1), 129–143 (2016)
2. Menezes, P., Chouzal, F., Urbano, D., Restivo, T.: Augmented reality in engineering.
In: 19th International Conference on Interactive Collaborative Learning and 45th
International Conference on Engineering Pedagogy (2016)
3. Móro, R., Bieliková, M.: Utilizing gaze data in learning: from reading patterns detec-
tion to personalization. In: UMAP Workshops (2015)
zamfira@unitbv.ro
The Importance of Eye-Tracking Analysis in Immersive Learning 697
4. Oglesby, J.L.M.: An adaptive visual learning approach for waterborne disease pre-
vention in rural West Africa. Master’s thesis, W. Kentucky University (2016)
5. Patney, A., Kim, J., Salvi, M., Kaplanyan, A., Wyman, C., Benty, N., Lefohn, A.,
Luebke, D.: Perceptually-based foveated virtual reality. In: SIGGRAPH Emerging
Technologies (2016)
6. Patrão, B., Menezes, P.: An immersive system for the training of tower crane oper-
ators. In: Exp.at 2013: The 2nd Experiment@International Conference, Coimbra,
Portugal (2013)
7. Song, J.C., Nam, S., Song, K.S., Park, S.Y.: Development of a personalized learning
system using gaze tracking system. WSEAS Trans. Comput. 14, 264–271 (2015)
zamfira@unitbv.ro
Simulation
zamfira@unitbv.ro
Augmented Reality-Based Interactive
Simulation Application in Double-Slit
Experiment
1 Introduction
Due to limited opening hours and lacking physical materials, it is difficult to access to
relevant references from the laboratory in the middle or high school. As a consequence,
many experiments cannot be carried out, especially in the teaching of physics, such as
the double-slit experiment. Therefore, students in such school learn physics still
through the traditional learning models, including attending a lecture, taking notes on
what the teachers write on the blackboard, memorizing facts from books or slides.
However, abstract concepts such as optical wave are formidable to high school stu-
dents, and their imaginative abilities are limited, so it is a challenge for them to imagine
the experimental progress just based on the information they got from books or slides.
This could be a barrier to in-depth understanding of the subject matter. Yet, these
problems could be solved by introducing alternative teaching resource such as Aug-
mented Reality-based learning tools (Jamali et al. 2015).
2 Literature Review
AR-based application in teaching and learning is most applicable in the following two
cases: (a) when the phenomenon is not easy to be simulated in reality, such as
inquiry-based micro-particles interactive experiments (Cai et al. 2014); (b) when real
experiment is limited by various factors which is hard to deal with, such as the convex
imaging experiment (Cai et al. 2013), as it is dangerous to keep a lighted candle in a
classroom.
Creating a mixed and enhanced reality, AR has compelling features for educational
purposes, such as learning content in 3D perspectives, offering learner with senses of
presence and immersion and visualizing the invisible (Wu et al. 2013). Additionally,
these features coincide with ideas in education theories. For instance, the theory of
situated learning insists that the actual and complete knowledge is acquired in real
learning situation, which AR technology could create by bridging virtual and real
worlds. Behaviorism which holds learning is the result of association formed between
stimuli and responses is another example. Within an AR-based learning environment,
learners could receive corresponding feedback immediately as they interact with the
environment or objects in it, while stimulus-response ties are forming and corre-
sponding knowledge is grasped. Besides, in an AR-based learning environment,
learners could gradually construct their recognition structures by conducting various
activities, which satisfies both Piaget’s assumption and practice of “bring laboratories
into classed” and the argument of constructivism that “learning is embedded in
authentic social experiences” (Cai et al. 2013).
A considerable number of AR-based learning and teaching tools are developed and
used in physics teaching (Castillo et al. 2015; Cai et al. 2013; Cai et al. 2016; Kauf-
mann and Meyer 2008), what’s more, a great number of researchers have designed
suitable activities to test the influence generated by using these tools in students’
learning performance (Akcayir et al. 2016; Cai et al. 2016; Wang et al. 2014).
Kaufmann and Meyer (2008) had introduced an AR-based application in teaching
mechanics. They developed a computer game to simulate experiments in the field of
mechanics. Involved in the 3D virtual world created by this application, students
engaged themselves in their own experiments. What’s important is that this application
zamfira@unitbv.ro
Augmented Reality-Based Interactive Simulation Application 703
offered students a considerate number of tools to measure mass, force and other
physical property of an object, during and after the experiments.
In the convex imaging experiment (Cai et al. 2013), learners need to (1) operate
2D-code cards to change the object distance and the distance between the object and
the lens; and (2) imagine that the 2D-code cards are the experimental facilities. The
learning effects could have been compromised due to the increased cognition load
caused by the information migration. The experiment would have been more inter-
esting if not only the virtual objects were integrated into a real scenario with AR, but
also the learner’s interactive operation behaviours were the same as the real experi-
mental condition.
AR-based application has been also developed for teaching magnetism. Cai et al.
(2016) implemented an AR and motion-sensing learning technology to teach the fields
of magnetic, where the magnetic model and magnetic induction line are simulated and
presented in real time. It demonstrated that the AR-based motion-sensing software can
improve students’ learning attitude and learning outcome.
As mentioned above, AR technology can improve development of simulation
systems and foster students’ learning of science. Therefore, our research targets
double-slit experiment, which phenomenon is not to be observed and is difficult to
carry out in most high schools. It is for this reason that we decide to develop an
AR-based interactive simulation application. With video recording the process of such
experiment, learners just can observe the phenomenon instead of interacting with them
by changing relevant parameter. Furthermore, Constructivism advocates that “knowl-
edge originate from activities and recognition starts from practice”. In the proposed AR
environment, learners can change relevant parameter with markers to observe the
variable phenomenon, furthermore, comprehend the process of such experiment. Our
research aims to design and develop a physical AR cognitive tool named DSIAR for
double-slit experiment and measure its reliability and usability.
zamfira@unitbv.ro
704 T. Wang et al.
Camera embedded
in mobile divece
Capture and track
Co rend
marker, and get the
m er
information of it
po in
sit g
in
g
Compare
Adjust the model
Database according to
of original infomation
markers
(a) (b)
(c) (d)
used to capture and present the real world and the virtual models while the camera
embedded in it detects markers.
DSIAR runs on an android mobile smart device (cellphone or tablet). In particular,
it focuses on in-depth understanding of relation between phenomenon and relevant
parameters (including distance between slits, distance between slits and fluorescent
zamfira@unitbv.ro
Augmented Reality-Based Interactive Simulation Application 705
zamfira@unitbv.ro
706 T. Wang et al.
expecting them to share their feelings as well as comments on this application in the
experimental operation.
From the interview, we can draw the following conclusions:
(1) The changing of phenomena with the operation of cards (adjusting relevant
parameters) is conform to reality.
After experiencing this AR-based application, the teacher expressed a wish to apply
it in her class. “It’s a wonderful teaching aids for this chapter. With it, the content will
not be dull and abstract.” which suggests that the simulation matches reality.
(2) All students felt that this application is very novel and interesting.
In the pilot testing, we offered the students with the AR-based simulation appli-
cation which they never experienced before. For them, therefore, this is very innovative
and interesting. They expressed that “Great, it’s a brand-new interaction way that I
can operate the object on the screen with my hands”; “I have never learning anything
with such application, if it were applied in class, my friends will be very interested!”
Students were impressed by the application, and it attracted their attention.
4 Conclusion
zamfira@unitbv.ro
Augmented Reality-Based Interactive Simulation Application 707
Based on preliminary results of pilot testing, DSIAR can have a positive influence
on assisting teaching and learning, attracting students’ attention and stimulating their
interests. Although the development of Augmented Reality-based interactive simula-
tion application is finished, however, the sample size is not large enough. In order to
further explore the effect of AR-based simulation application, future work will involve
a large sample under rather more naturalistic conditions to collect enough data to verify
the effect and potential of AR-based simulation application, while an inquiry-based
learning activity will be designed.
References
Akcayir, M., Akcayir, G., Pektas, H.M., Ocak, M.A.: Augmented reality in science laboratories: the
effects of augmented reality on university students’ laboratory skills and attitudes toward science
laboratories. Comput. Hum. Behav. 57, 334–342 (2016). doi:10.1016/j.chb.2015.12.054
Castillo, B., Iván, R., Sánchez, C., Guadalupe, V., Villegas, V., Osiris, O.: A pilot study on the
use of mobile augmented reality for interactive experimentation in quadratic equations. Math.
Probl. Eng. 2015, 1–13 (2015). doi:10.1155/2015/946034
Cai, S., Chiang, F.-K., Wang, X.: Using the augmented reality 3D technique for a convex
imaging experiment in a physics course. Int. J. Eng. Educ. 29(4), 856–865 (2013)
Cai, S., Chiang, F.K., Sun, Y., Lin, C., Lee, J.J.: Applications of augmented reality-based natural
interactive learning in magnetic field instruction. Interact. Learn. Environ. 25(6), 1–14 (2016)
Cai, S., Wang, X., Chiang, F.-K.: A case study of augmented reality simulation system
application in a chemistry course. Comput. Hum. Behav. 37, 31–40 (2014). doi:10.1016/j.
chb.2014.04.018
Gausemeier, K., Bruseke, U., Wortmann, R.: Virtual and augmented reality in education and
training: an interactive, multimedia training and information system for use in an exhibition.
In: Third International Conference on Virtual Reality and Its Application in Industry,
vol. 4756 (2003). 304-313428
Jamali, S.S., Shiratuddin, M.F., Wong, K.W., Oskam, C.L.: Utilising mobile-augmented reality
for learning human anatomy. Procedia Soc. Behav. Sci. 197, 659–668 (2015). doi:10.1016/j.
sbspro.2015.07.054
Kaufmann, H., Meyer, B.: Simulating educational physical experiments in augmented reality. In:
Applications of Mixed Reality, pp. 1–8 (2008)
Nincarean, D., Ali, M.B., Halim, N.D.A., Rahman, M.H.A.: Mobile augmented reality: the
potential for education. In: 13th International Educational Technology Conference, vol. 103,
pp. 657–664 (2013). doi:10.1016/j.sbspro.2013.10.385
Wang, H.-Y., Duh, H.B.-L., Li, N., Lin, T.-J., Tsai, C.-C.: An investigation of university
students’ collaborative inquiry learning behaviors in an augmented reality simulation and a
traditional simulation. J. Sci. Educ. Technol. 23(5), 682–691 (2014). doi:10.1007/s10956-
014-9494-8
Wu, H.-K., Lee, S.W.-Y., Chang, H.-Y., Liang, J.-C.: Current status, opportunities and
challenges of augmented reality in education. Comput. Educ. 62, 41–49 (2013)
zamfira@unitbv.ro
Developing Metacognitive Skills for Training
on Information Security
1 Introduction
We would wonder it is good to think aloud, work in small groups, or preferably two
students, to encourage team complicity, to participate with other groups on time, to
exchange opinions maybe by thoughtlessly opening up their creative ideas, to share
experiences in the same sense, to discuss solutions and to propose critical thinking. But
all this we do not have usually covered by a lesson list or concept contents or in a
laboratory practice agenda for our subject (tell yourself Information Security or another
subject), that we teach in the classroom.
Metacognition is known as the educational phenomenon of thinking about thinking,
which provides a way to explore reasoning behind concepts. Each person has different
metacognitive abilities that are acquired while someone grows and matures in life
(Willingham 2007). Some of these skills are to monitor one’s own learning, to know
what one does not know, to predict performance, to plan, to manage study time and to
cognitive resources to achieve a successful outcome. So, in order to maximize learning,
students need to know how metacognitive aspects can improve their skills on a subject
including content, procedural and conditional knowledges.
Studies on metacognition and cognitive monitoring point out that young students
have a very limited knowledge of their own cognitive processes. Metacognitive
knowledge is thought to consist of the thoughts that each has of oneself, but of others
too, in the process of learning. This is can be said in the results of any kind of intellectual
activity. Thus, metacognitive experiences are conscious experiences. Then a nuclear
idea is to meet the challenge of finding effective ways to teach metacognitive knowledge
and cognitive monitoring skills. Here, we would like to deal with metacognition as an
emerging boom of scientific researchers in recent years, but applied to education in
Information Security.
Our pretensions with this paper are double. First, to study the relevance of
metacognition and its interest in information security education. Secondly, once this
research context has been established, it will provide a framework for educational
intervention that will allow us to advance a little in this line of research. To achieve this
in the remainder of the text, we introduce some related works that we have considered
relevant in Sect. 2, the research methodology in Sect. 3, to move to Sect. 4 with the
results and discussion, a case of study in Sect. 5 and finally conclusions.
2 Related Works
In related work, we would suggest to take into account three points of reference in
order to focus this paper: what about training on security of information, about
classroom didactics and what is related to metacognition.
Around our subject, there are inquiries, such as Puhakainen and Siponen (2010),
which suggests that training in information security needed educational approaches
based on theory and not excessively falling into anecdotal or superficial contents. It was
considered to employees as a measure to solve a known problem related to training, but
we can consider these reflections to a certain extent in university environment to
properly plan a course didactically.
In addition, in information security courses, it is significant practical tasks to learn
by doing and to participate actively in classroom, even further in working out
non-technical learners or young people learning first information security. Really
cybersecurity is a concern into a classroom, but also outside: at home and on the streets.
It is a cross-cutting issue that involves all groups in society and almost all aspects of
human activity with the rapid spread of information technologies.
Some authors have emphasized that it is possible to bring technical topics closer to
several collectives by designing a relevant instructional and motivation course and
assessing it in the area of social sciences (non-technical skills). As Cano et al. (2016)
have analyzed, the nature of this transcends, per se, the engineering branch and coexists
with Law, Business management, Sociology, Psychology and Criminology, among
others.
Nowadays, a key to keep in mind is that students who begin their career are really
digital natives. In another way but that came to mean something similar, Prensky (2001)
said students are all native speakers of the digital language of computers, video games
and Internet. This study had classified students into two groups: people born before or
after 1980, in order to from this way to accept that there are digital native or only
immigrants. An interesting debate on digital natives has been very illustrative on two
academic levels about it (Bennet et al. 2008). Firstly, on the questioning of the existence
of a generation of digital natives and how to define it including age range and other
characteristics; and secondly, how education should change given these evidences.
zamfira@unitbv.ro
710 J. Cano et al.
On another front, since the first seminal studies about metacognitive capabilities
mentioned inin Flavell (1976; 1979) as well as Brown (1978), scientific interest in
metacognitive development has increased enormously in recent years. As a matter of
fact, according to them, metacognition deals with knowledge of processes and products
of cognition applied to oneself, through an active supervision of cognition, i.e., on
learning-relevant properties of information or data. And as they said, there is
metacognition if we notice that we have more problems learning one thing than
another, note to revise something before finally accepting it, write down something
before I can forget it, and so on. A lot of these metacognitive activities are needed in
problem solving techniques. The concern to self-inquire about the state of our
knowledge of something, in the process of solving a problem is essential, in situations
of a laboratory, in schools, or in daily life, as Brown mentioned. To get here,
metacognition demands introspection of one’s own performance to learn and discern
between our point of view and that of others. It should include the ability to make
efficient problem-solving decisions and, in this way, to be able to predict the limitations
of capacity that one has, to know heuristic routines and their usefulness that can help us
in the learning process, to identify and characterize the problem that is addressed, to
plan and to program the appropriate problem, to monitor and to supervise how effective
routines are, and to dynamically evaluate our cognitive activities to face success or
failure in a strategically-timed completion. Accordingly, it could be said a bunch of
decision-making keys to resolve problems.
Some years later, we can find a study of the US Army (Geiwitz 1994) that proposes
a conceptual model for the training of metacognitive abilities that highlights the rela-
tionship between these abilities and cognitive performance as well as suggestions for
their evaluation. It has also been interesting to us by the review of the advances in the
matter particularly in its decade back and its justification by means of an experiment
with questions on physics, tension and acceleration. In (Reed and Giessler 1996)
another experience can be seen applied to a software development course. For this,
authors present twelve graduated students and they intend to set up a comparison
between metacognition and experience for working hypermedia environments.
As for our preferred line on the subject of Information Security, especially at a
non-graduate level, it should be emphasized that traditional lecture approach has been
observed as a controversy about how teachers choose right topics and awaken the
interest of learners who play a more passive role in the classroom. But in spite of being
the typical way to teach Information Security, other cognitive strategies can be given
validly, such as Tutorial approach for self-learning, Scribes approach assigning
student-scribes for taking careful notes and subsequently presentation, Mentor
approach of having experts instruct, Project approach which can include an activity at
the end, an experiment or another classroom preparation, Synergy Approach in which
research and teaching attempt to attract the student’s attention, and Attack/defend
approach where students are divided into offensive teams and defensive teams with
different goals, as shown in Yurcik and Doss (2001).
A leading author on this subject, M. Bitshop in her “Education in Information
Security” (Bitshop 2000), described the objective of university education in Informa-
tion Security like learning general principles, how to apply them, and at best, teachers
could take case studies and generalize them to understand.
zamfira@unitbv.ro
Developing Metacognitive Skills for Training 711
However, with all these studies in the mixer, we still feel that there is a gap between
metacognition and educational aspects in Information Security, which has in addition
the complexity of being a multidisciplinary subject.
At the beginning, our methods exposed in this work start with doing a comprehensive
study or review of the literature regarding metacognitive abilities. That is why we have
decided to seize the initiative and introduce a drawn scheme facilitating a suitable
understanding of the context where we move. This has been important for us since, as
far as we know, there is insufficient research on the metacognitive aspects of education
in Information Security issues, covering both technical and non-technical aspects,
which may be influenced by multiple disciplines like Computer Science, Telecom-
munication, Law, Business Management or Criminology, among others.
Our first hypothesis was that it can be built a scenario for knowing how significant
has been scientific research on Metacognition over time and what is the current state. It
seems reasonable that the right environment should provide a context where we, or
another teaching team, can find out some guidelines and activities for students to be
able to develop a positive learning experience.
Secondly, our hypothesis is that should be possible to integrate somehow into
Information Security a parallel metacognitive curriculum. We planned the hypothesis in
the sense of training students, included undergraduate ones, and it should incorporate
metacognitive learning together regard of cognitive content. Moreover, we believe that
practical activities could be a catalyst to reach these skills, in any way empowering
students to be responsible for their own learning and internalizing the importance of
developing metacognitive processes. Asking ourselves matters covered by this per-
spective, we can design instructional and motivational activities and strategies to allow us
to implement metacognitive processes. Then we could sum up a bunch of metacognitive
practices in education in order to contribute to the curricula but also for lifelong learning.
The methodology of the instructional design consists of three blocks. First of all, an
activity is presented with an introduction for capturing the attention and enthusiasm of
students themselves. Course objectives and a personal and professional motivation
should be presented. Secondly students are divided into pairs or three-member groups
if any, facilitating cooperative work. Talking out loud is advisable if they do not speak
more than two at a time or do not cross conversations, but in any case, do not interrupt
if teacher speaks to class crowd. Finally, each student should individually make a
results-based evaluation of the activity.
Metacognitive instruction will essentially consist of taking the time each student
takes at each moment during the teaching-learning process. For this, in the line of time
monitoring and the temporary awareness of the process of thinking, they must give a
description as well as spending minutes of each small enough piece of activity that they
might consider to be the proposed activity.
This way of learning to do the activities have a clear didactic vision. We aim to
approach learning techniques based on problems (ABP) or case studies (ABC).
The strategy pursues autonomous learning with necessary teaching guide proposing a
zamfira@unitbv.ro
712 J. Cano et al.
similar exercise in which can be found in the real world. With these techniques based on
problems, knowledge of cybersecurity as content is reached. But also, meta-cognitive
skills such as creative thinking, meaningful learning and learning to learn.
At the spatial level, learning monitoring will consist of making drawings or graphic
diagrams, which can help with solving problem, such as assembly, disassembly,
construction, list of steps to do, and much better taking pictures to remember something
that it could be difficult to write down. This one reinforces the idea that a picture is
worth more than a thousand words in many cases. Given the digital environment we
see in our classrooms, we could think it is easy to find our universities plenty of
students with smart phones with camera today.
Practical activities on computer security actually emphasize active learning
dynamics. These processes are difficult to perform through lectures and seminars
(understood as an oral presentation of a topic by a teacher in classroom). In terms of
skills, it is beneficial to be able to make practices in a lab format, using a laboratory for
just this, or at least an upgraded computer room for that purpose. That allows students
to develop a non-formal learning environment which facilitates especially significant
innovations in such a dynamic way. We believe that teaching improves if we try to
insert task through simulations, case studies, practical problems, workshops and, within
this situation, computer forensic experiments. In short, that may give a meaningful
experience to students.
zamfira@unitbv.ro
Developing Metacognitive Skills for Training 713
zamfira@unitbv.ro
714 J. Cano et al.
To address this concern, we were selecting the top nine reporting specialties today
in order to have a historical view of the relevance that each one has had over time. For
this, we have compiled a series of data, with the intention of observing if we find some
change in evolution that allows us to make a valuation. This can be seen in two parts,
firstly taking note of the proportion of our topic metacognition in scientific articles for
each of the last five years (period 2010–2015). Secondly, we wrote down taking a
journey back in 5-year steps until the 80s, i.e. milestones for 2005, 2000, 1995, 1990
and 1985. Below this compilation is expressed in a bar graph of Fig. 3.
In the Fig. 3 it is possible to see that there are empty cells when there are no data to
show. Thus, when cumulative sums of percentages are greater than one hundred per-
cent, there are articles that deal with several disciplines at a time.
Gradually the academic interest in metacognitive issues has been linked to Psy-
chology and Social Sciences, effectively from the early years in the 80 s. At the end of
that decade Medicine as well as Arts and Humanities are beginning to wake up. Also,
Neuroscience and Engineering in small proportion. This has been a fact that will
remain until today.
zamfira@unitbv.ro
Developing Metacognitive Skills for Training 715
As can be seen in Fig. 3, it was not until the mid-1990s that computer science
became aware of the concept. With the advent of the year 2000, IT issues appear with
greater force than Engineering itself and curiously in the early years of the decade
(2010) is positioned behind the pioneering branches of Psychology, Social Sciences
and Medicine, even Humanities.
All these results make us confirm that Computing is experiencing an emergent
creativity to apply metacognitive concepts, at least in the light of the results of the
articles compiled from literature.
But at this point, IT field is broad enough to allow us to ask for Information
Security. Actually, we can see that it is great to have an instrument to position our-
selves to what extent it is. Furthermore, from this relevant knowledge on metacognition
finding, there is a gap to study that would be interesting to develop in academia and
among practitioners.
zamfira@unitbv.ro
716 J. Cano et al.
zamfira@unitbv.ro
Developing Metacognitive Skills for Training 717
autonomous students become. The monitoring will be executed according to the close
observation of these strategies until their completion.
Our activity within the frame of work is rounded with a self-assessment which
implements each student, really the small group (2 or 3 members) to which it belongs.
The assessment is compared to classmates, at least two of them, so that they can
compare the execution and monitoring of metacognitive strategies and cognitive
outcomes.
5 A Case of Study
zamfira@unitbv.ro
718 J. Cano et al.
To get an idea of the evaluation performed in the activities, 5 out of the 29 students
were excellent and 5 were poor, but in general they have got an averaged score of 7.7 in
the whole of classroom. Figure 7 shows the general results obtained graphically.
zamfira@unitbv.ro
Developing Metacognitive Skills for Training 719
6 Conclusions
In summary, students may feel that they are facing real problems related to Information
Security and Communications in a guided environment. Knowledge learning is
important and a demanding goal, but responding to activities we have developed
meta-cognition strategies that enrich the fact of active learning by doing. The strategies
we have set out to be aware of how we learn, have been satisfactory and we feel that
they complete and perfect our educational mission over and above other considerations.
That is, not only content is developed, but metacognitive skills that the students are
going to take for their long life and even more can serve as a model to meet other
challenges.
Finally, our contribution really begins a path where we have tried to cover a gap
with regard to metacognition in a field as multidisciplinary as Information Security.
References
Bennett, S., Maton, K., Kervin, L.: The ‘digital natives’ debate: a critical review of the evidence.
Br. J. Educ. Technol. 39(5), 775–786 (2008)
Bishop, M.: Education in information security. IEEE Concurrency 8(4), 4–8 (2000)
zamfira@unitbv.ro
720 J. Cano et al.
Brown, A.L.: Knowing When, Where, and How to Remember: A Problem of Metacognition
(1978)
Cano, J., Hernández, R., Ros, S.: Bringing an engineering lab into social sciences: didactic
approach and an experiential evaluation. IEEE Commun. Mag. 52(12), pp. 101–107 (2014)
Cano, J., Hernández, R., Ros, S., Tobarra, L.: A distributed laboratory architecture for game
based learning in cybersecurity and critical infrastructures. In: 2016 13th International
Conference on Remote Engineering and Virtual Instrumentation (REV), Madrid, pp. 183–185
(2016)
Geiwitz, J.: Training Metacognitive Skills for Problem Solving (No. ASC-TR-051-3). Advanced
Scientific Concepts Inc., Pittsburgh (1994)
Flavell, J.H.: Metacognitive aspects of problem solving. Nat. Intell. 12, 231–235 (1976)
Flavell, J.H.: Metacognition and cognitive monitoring: a new area of cognitive–developmental
inquiry. Am. Psychol. 34(10), 906 (1979)
Puhakainen, P., Siponen, M.: Improving employees’ compliance through information systems
security training: an action research study. MIS Q. 34(4), 757–778 (2010)
Prensky, M.: Digital natives, digital immigrants: part 1. Horiz. 9(5), 1–6 (2001)
Reed, W. M., Giessler, S.F.: Prior computer-related experiences and hypermedia metacognition.
Comput. Hum. Behav. 11(3), 581–600 (1996)
Willingham, D.T.: Critical thinking. Am. Educ. 31, 8–19 (2007)
Yurcik, W., Doss, D.: Different approaches in the teaching of information systems security. In:
Proceedings of the Information Systems Education Conference, November 2001
zamfira@unitbv.ro
Optimization of the Power Flow in a Smart Home
Abstract. With the IT technology, the traditional power grid is being upgraded
to the smart grid (SG) with two-way communication and power flow between
utilities and customers. In addition, SG includes new technologies in distributed
energy generation (DEG) and distributed energy storage (DES), advanced meas‐
urement and sensing, controls, cyber security, consumer-side energy manage‐
ment, and environment protection. Thus, it shows the advantages in efficiency,
reliability, and security. A smart home is a mini power system with the renewable
energy resources and the local energy management. Therefore, the emission and
the power consumption can be reduced while the system efficiency will be
improved. In this paper, particle swarm optimization is used to manage the power
flow in a smart home with the objectives in the minimum cost and maximum
comfort. Results from two homes with different size of PV systems are compared
and discussed. The PV size for a stand-alone home is determined.
1 Introduction
With the IT technology, the traditional power grid is being upgraded to the smart grid (SG)
with two-way communication and power flow between utilities and customers. In addi‐
tion, SG includes new technologies in distributed energy generation (DEG) and distrib‐
uted energy storage (DES), advanced measurement and sensing, controls, cyber security,
consumer-side energy management, and environment protection. Thus, it shows the
advantages in efficiency, reliability, and security. DEG plays a more and more important
role in SG through peak demand reduction, congestion alleviation, and reliability. It is
mainly from solar, hydro, wind, biomass, and geothermal renewable energy, and 13% of
electricity was from the renewable energy in 2015. In the U.S., 37% of electricity is
consumed in residences, where the heating and cooling account for about 48% of the
utility bill in the home [4]. Similar to power grid, the residential buildings are being
upgraded to smart homes with incorporated devices to achieve the goals of the homes,
such as energy consumption, comfort, security, and home-based health care [1]. A smart
home typically includes HVAC system, electrical vehicles (EVs), home backup battery,
and other appliances. It also includes the DEGs with renewable energy except hydro and
biomass. Due to the possible noise of the wind turbines and particular regions required by
the geothermal, they both show the restriction for the installation close to residential build‐
ings. Solar energy, the only one left, is clean and sustainable with little maintenance for the
zamfira@unitbv.ro
722 L. Zhang and X. Xiong
photovoltaic (PV) system. Although the PV panels are not effective in cold regions, their
advantages still make them as a viable source of electricity generation. Furthermore, with
the technology development and consumer awareness, the cost of the residential installed
PV project fell from $5.71/W in 2010 to $3.09/W in 2015 and the national residential PV
capacity increased from 250 MW in 2010 to 2,250 MW in 2015 [2–4].
Besides the development in renewable energy, there is a significant change in the
transportation. Today, more than 180, 000 electric vehicles (EVs) are on the roads
worldwide and they are attractive to the families due to their high efficiency and low or
zero emission. On the market, there are Tesla Model S, Honda Fit, Ford Focus, and
Nissan Leaf and their miles per gallon equivalent (MPGe) rating is around 100 according
to the U.S. Environmental Protection Agency [5–7]. These EVs can be Hybrid Electric
Vehicles (HEV), Plug-in Hybrid Electric Vehicles (PHEV), and pure Battery Electric
Vehicle (BEV) [8]. With the enough spare generating capacity in the current power grid
infrastructure, the EV number is expected to be over 20 million by 2020 [9, 10]. At a
smart home, EV charging in the night is convenient and doesn’t need to go to the gas
station. A vehicle-to-grid (V2G) car with fully charged batteries (85 kWh as the capacity)
could provide the 10 kW grid for 4.25 h for 50% discharge. Thus, a large amount of DES
can supply power to the critical facility in response to power shortage due to storms and
other disasters [11].
For the residential DEG, the outputs from the stochastic PV systems strongly depend
on the weather and have significant negative impacts on grid voltage and frequency
stability. Therefore, the PV system has no scheduling freedom. In order to compensate
for the negative impact of DEG on the grid, techniques in DES have been developed to
mitigate this variability for short periods. DES includes rechargeable battery, super
capacitor, and hot water tank technologies. They have different ratings for power and
discharge time [12, 13]. For a smart home, home backup batteries are possible choices
to mitigate the power-quality related issues by scheduling the batteries’ charging and
discharging. On the market, the home backup batteries available is $3,500 for a 10 kWh
batteries and $25,000 for a 100 kWh [11]. When combined with rooftop photovoltaic
(PV) system, the utility cost can be completely eliminated. In addition, the vehicles stay
in garages for most of time and the EVs’ large batteries can also be used as distributed
energy storage (DES) to smooth out the fluctuating production of electricity and to
improve stability of the power system. Similar to DEG, DES has its own behavior pattern
and the control strategies for DES are based on such behavior as well as the energy
conversion efficiency. For instance, to charge rechargeable battery at constant current,
the charge voltage increases and its capacity is affected by ambient temperature, etc.
The power from the grid, DEG, or DES will be consumed by different devices and
appliances at a smart home. Their demand on electricity is a random variable with a
probability distribution in an operation time window and it can be categorized into
different types or patterns based on whether it is can be separated, shifted, interrupted,
decreased, or cancelled [14]. However, such pattern of a demand may change. For
example, the charging of an electrical vehicle can be interruptible and shiftable with in
the midnight. But this charging can be non-interruptible and non-shiftable if a long
distance trip is planned for the next early morning. With the advanced sensing and
metering, residential energy management systems can adjust consumption levels as
zamfira@unitbv.ro
Optimization of the Power Flow in a Smart Home 723
demand response to respond the electricity price, correct voltage sags and flickers, or to
help stabilize the system frequency [15]. This energy management can also help utility
companies and power plants reduce cost through reductions in peak demand.
In this paper, Sect. 2 is focused on the physical and mathematical model of a smart
home and the details in the optimization of the power flow are provided. Section 3 shows
the simulation results and the discussion of the power flow and Sect. 4 includes the
conclusion.
The structure of a smart home is shown in Fig. 1. This home consists of several rooms.
In addition, there are a roof-top PV system, a set of home backup battery, an AC system,
and an electric vehicle in the garage. One central control panel is used for the energy
management in the house.
For the temperature control in each room, there are at least one electronic grill
damper, a thermometer, a motion detector, and one RF transceiver. The damper is a
valve or plate that regulates the flow of air inside a duct to the room. Thus, there are
multiple zones in a smart home.
Among different devices, heating, cooling, and lighting counts for over 50% of the
utility bill. While a 2.5 ton central unit (about the size for a typical 1,500 to 2,000 square
foot home) uses about 8.7 kW. For a 5 ton, the power is around 17 kW. Two stage
cooling, Two-stage cooling means the air conditioner or heat pump has a compressor
with two levels of operation: high for hot summer days and low for milder days.
Comfortable summer room temperature 25 °C and the winter room temperature 23 °C.
Zoning allows resident to precisely control the temperature in every room of the
home. With a thermometer in each room, an automatic air vent grill damper in the duct
is used to control airflow to that room. Motion detectors will help the central controller
to determine whether the room is occupied or not. Thus, the temperature is differentiated
and energy is saved. Most large homes feature multi-zone heating and air conditioning
systems with one single furnace or one central air conditioner. With electronic dampers
and thermostats, the rooms will be maintained at different temperatures. Thus, the energy
is saved as unoccupied parts of the home are kept at a temperature setback. In fact, most
zamfira@unitbv.ro
724 L. Zhang and X. Xiong
traditional multi-zone duct systems don’t save energy and their reliability is low. If the
residents are away, the heating or ac should not be shut down in order to maintain the
temperature and control humid to minimize the damage to furniture and the wall
compounds in Sheetrock.
0.25
Price ($/kWh)
0.20
0.15
5 10 15 20 25
Time (hour)
2.2 Loads
In this model, there are two types of loads: flexible and fixed. The power supplied to the
fixed load is guaranteed but not to the flexible load. Thus, the flexible loads can be
reduced or completely cancelled and its original profile is shown in Fig. 3. In this work,
there are two kinds of the fixed loads implemented, one is also shown in the Fig. 3 and
it is due to the devices, such as router and home security system, which always run. The
second one is shiftable loads and it can be regarded as separated and fine-grained tasks.
Each task will be completed in one time slot. This kind of loads includes the dish washer,
washer, and dryer. The EV battery is also a shiftable load for charging with a specific
time window. In out calculation, the daily total usage is around 27.8 kWh, which is close
to the average daily American home electricity consumption [20]. There are two peaks
in the load profile at 7AM and 6PM similar to ref [21].
zamfira@unitbv.ro
Optimization of the Power Flow in a Smart Home 725
2.0
desired fixed
desired flexible
1.5
Load (kW)
1.0
0.5
0.0
5 10 15 20 25
Time (h)
In this work, only AC unit operation is discussed in the summer but not household
heating in winter because nearly half of houses use natural gas [22]. Power demand from
the AC unit is a combination of the fixed and flexible loads. If the room is occupied, the
temperature has to be maintained as the desired temperature and this demand is fixed.
For the room unoccupied, the temperature can be kept at a little bit higher than the desired
temperature and this load is flexible.
Q = w1 C + w2 U (1)
Here, C is calculated through the price and the power flow between the grid and the
home. This power can be positive if it flows from the grid to the home and negative if
it flows from the home to the grid. U is proportional to the shortage of the power supplied
to the flexible loads. w1 and w2 are weighting factors and the sum of them is 1. The
objective of the control is to minimize the Q and the room temperature settling time.
In this paper, an optimal energy management strategy, through particle swarm opti‐
mization (PSO), is to minimize Eq. (1) for the most cost-effective and comfort-aware
service to a residential consumer. In the implementation, there are one array for the
response of the flexible loads, one array for the power flow between the home and the
grid, and two arrays for charging and discharging strategy of the home backup battery
and the EV battery. The particles, a potential solution to the problem, consist of these
four arrays as the positions. In case there is no home backup battery, the element in the
corresponding array is set to be zero.
zamfira@unitbv.ro
726 L. Zhang and X. Xiong
Figure 4 shows the load profiles on the summer solstice. Compared with the Fig. 3, there
is difference between the scheduled fixed load and the real load on the summer solstice
because of the extra power for the central AC unit for cooling. The peak values in both
cases are around 3.25 kW, which is 1.5 kW higher. While the real flexible loads are the
same in the day time but not in the night time. More power is supplied to the flexible
load at the home with 100 m2 PV system.
6 45
shiftable load
flexible load (a)
5 fixed load 40
ambient temperature
4
Temperature (oC)
35
Load (kW)
3
30
2
25
1
0 20
5 10 15 20 25
Time (hour)
6 45
shiftable load (b)
5 flexible load 40
fixed load
ambient temperature
4
Temperature (oC)
35
Load (kW)
3
30
2
25
1
0 20
5 10 15 20 25
Time (hour)
Fig. 4. Loads and the ambient temperature on the summer solstice for the houses with 50 m2 (a)
and 100 m2 (b) PV systems
In the summer, the temperature of the occupied rooms is set as 23 °C. Otherwise, it
is set as 27 °C. In Fig. 5, the power for cooling to the bed rooms and the other rooms is
the same before 20 O’clock in both cases. Since the size and the heating properties of
the rooms are close, the temperature is maintained at 27 °C. With the same dead band
zamfira@unitbv.ro
Optimization of the Power Flow in a Smart Home 727
as 1 °C, the temperature of the rooms fluctuate simultaneously. After 20 O’clock, the
living rooms and the kitchen are occupied, the temperature is kept at 23 °C, the power
increases. However, the temperature of the bed rooms are still maintained 27 °C. After
21 O’clock, the temperature set point of the bed rooms are changed to 23 °C and the
temperature in the other rooms is set at 27 °C.
0.2
0.1
0.0
30
28
Temperature (oC)
26
24
22
20
5 10 15 20 25
Time (hour)
bed room
0.3
living room (b)
Power (kW)
0.2
0.1
0.0
30
28
Temperature (oC)
26
24
22
20
5 10 15 20 25
Time (hour)
Fig. 5. The power and the temperature in the bed room and the living room 50 m2 (a) and
100 m2 (b) PV system
Figure 6 shows the power flow from the homes with a 50 m2 and 100 m2 PV system
on the summer solstice. This day is the longest with the sunrise at 5 o’clock and sunset
at 18 o’clock. The PV system generates electricity from 5 o’clock to 18 o’clock. Since
the ambient temperature is high, the peak power is only 5 kW and 10 kW for the two
cases. The power from the PV system is supplied to the loads and the backup battery.
For the home with a 50 m2 PV System, the PV output power is supplied to the loads and
zamfira@unitbv.ro
728 L. Zhang and X. Xiong
the EV battery after 15 o’clock. Between the sunset at 18 O’clock and 21 O’clock, battery
starts to be discharged. Then, power from the grid, with the peak between the prices of
$0.136/kWh and $0.139/kWh, is the only source to meet the demand of the home. For
the home with a 100 m2 PV system, the peak PV output power is 10 kW between 10 and
15 O’clock and the power flow is similar to that in the home with 50 m2 PV system.
After sunset, the backup battery is discharged and the power flows to the load and the
EV battery. Different to the first case, there is no power needed from the grid. Thus, a
home with a 100 m2 PV system can be a stand-alone system.
20
EV battery
backup battery
15 shiftable loads
flexible loads
fixed loads
10
PV
Power (kW)
grid
5
-5
(a)
-10
5 10 15 20 25
Time (hour)
25
20
15
10
5
Power (kW)
0
EV battery
-5 backup battery
-10 shiftable load
flexible load
-15 fixed load
-20 PV (b)
grid
-25
5 10 15 20 25
Time (hour)
Fig. 6. The power profiles on summer solstice with the PV size as 50 m2 (a) and 100 m2 (b)
zamfira@unitbv.ro
Optimization of the Power Flow in a Smart Home 729
4 Conclusions
In this paper, particle swarm optimization is used to manage the power flow in a smart
home with the objectives in the minimum cost and maximum comfort. Results from two
homes with different size of PV systems are compared and discussed. The fixed loads
are guaranteed with the power supply and the temperature of the occupied room is
maintained. The PV size for a stand-alone home is determined.
References
zamfira@unitbv.ro
730 L. Zhang and X. Xiong
zamfira@unitbv.ro
A Virtualized Computer Network for Salahaddin
University New Campus of HTTP Services
Using OPNET Simulator
1
Department of Computer Science and Engineering,
University of Kurdistan-Hewler, Kurdistan, Erbil, Iraq
tarik.ahmed@ukh.edu.krd
2
Computer Science and IT Department, Salahaddin University-Erbil, Erbil, Kurdistan, Iraq
ammar.hasan@su.edu.krd
Abstract. Factors of any computer network must be studied carefully and the
specialty of any type of computer network must be taken into consideration to
design an effective and a competent computer network and to avoid future dead‐
locks that might take place. Computer network factors are many as they can show
its performance and capabilities. The examples of these factors are time delay,
throughput, bandwidth, transfer of data, packet transfer, packet delay, http
transfer, congestion, collision, Ethernet specifications, VoIP etc. This paper is
depended on OPNET for developing a computer network simulation for Sala‐
haddin University-Erbil measuring HTTP (Hyper Text Transfer Protocol) service
values. Moreover, the simulation is made without high speed medias and devices
to make the design reliable and with superior performance sufficient to satisfy the
university users’ requests.
1 Introduction
This research starts with a practical and virtual example for designing a computer
network that will be installed using the Optimized Network Engineering Tool (OPNET)
simulator for the new university Campus. The Core switch (Cisco 6500) is used to
interconnect all other switches that are fixed in each college. The switch can only connect
LAN’s of the same type (Ethernet to Ethernet, FDDI to FDDI, or Token Ring to Token
Ring).
The factors that are mentioned above must be considered to achieve maximum
workable efficiency and non-stoppable productivity. In addition, a network without any
collision and with less congestion can be achieved. This project is conducted through
the OPNET simulator (the version of modular 14.5). The duration of the simulation is
set to 20 min. This period is sufficient to the network devices to get stabilized after the
addresses were distributed to all workstations.
zamfira@unitbv.ro
732 T.A. Rashid and A.O. Barznji
This is clearly demonstrated from the obtained results in this paper. The results are
practical, almost like real, and especially are depended on the devices and medias which
are used for interconnecting the network. The OPNET provides a complete environment
which supports the distributed systems and the modeling of communication networks.
Both behavior and performance of modeled systems can be analyzed by performing
discrete event simulations. The environment of OPNET integrates different technologies
and it can be used for different aspects in education systems. This covers simulation,
data collection, designing a model, and data investigation.
OPNET allows the selection of any device or component needed to be set and used
in the design, then configures and programs each one individually per the application
simulated in the design. There is also capability to address these devises for establishing
connections among them with all other nodes then providing troubleshooting to all the
network through the RUN command containing all its devices and medias. And finally,
the OPNET monitors the communication to get feedbacks for determining the perform‐
ance of the network. By using the simulation of OPNET, the designer can choose devices
and components for different and multiple models, categories, vendors, and factories to
determine the best components for each computer network. This work also aims at
examining the used components in the design and implementing them to avoid deadlocks
and congestions in the network. Each device can be chosen and configured individually
on the OPNET. Also, all the connections or links in the design must be considered by
the designer of the network per their technical specifications.
2 Background
This research is depended on the OPNET to design a computer network for Salahaddin
University-Erbil Campus measuring HTTP service values to reach reliable performance
sufficient to the users. The researchers in [1], designed a simulation via the OPNET for
Mosul University. Three applications were added to test the network design. These appli‐
cations were FTP (File Transfer Protocol), HTTP (Hyper Text Transfer Protocol) and VoIP
(Voice over Internet Protocol). The results showed that their proposed model had a positive
efficiency on designing and managing the targeted network and can be used to view the data
flow. The researchers in [2] presented a design and an implementation of hybrid network for
different IP routing protocols in low load campus network. In this network model, generic
LANs models and WLANs were used. They created a simulated environment where many
applications are in use at a time. The network model was based on the OPNET IT GURU
Academic Edition. The OPNET was used to develop a new model suitable for Campus/
University environment. The model was tested against various types of applications (FTP,
ATM, and Remote Login) in Hybrid Networks. Two routing protocols RIP and IGRP were
used to check the performance of Hybrid network for different applications. The OPNET
simulation showed the impact of IP routing protocol for hybrid networks for different types
of applications. The researchers in [3] designed a computer network infrastructure to
support various administrative and academic activities for Tarumanagara University. The
built infrastructure covered all areas of the building and the floor. With increasing range of
services, Tarumanagara University needed to enhance local networks to accommodate their
zamfira@unitbv.ro
A Virtualized Computer Network for Salahaddin University 733
needs. The new network infrastructure that was built must guarantee the quality of serv‐
ices. The services are reliable, scalable and support future expansions. The network design
was the most important and critical parts before developing a new network infrastructure.
Analysis of user and network requirement was done to design the network.
The above research works demonstrate that there is an actual need to design the network
efficiently then simulate it via the OPNET [4, 5] before starting to build it on practical
fields. From the obtained results, several decisions can be taken to make the design more
efficient and reliable regarding the used media and devices, switch categories, transmission
technologies and protocol types.
zamfira@unitbv.ro
734 T.A. Rashid and A.O. Barznji
provides the internet service to the network project. The server can be connected to a
router to achieve more security and to protect entire network from the intruders. Firewall
is used to protect internal network computers from the intruders and hackers. The capa‐
bility of this design to add wireless distribution services is in such an easy way via just
replacing the college switches by a wireless router to distribute wireless services that
maintain the speed and connectivity. Table 1, shows the used components in this project.
The cloud is used to provide services for the hosts of the entire network, and then the
network is configured to produce many services HTTP, VoIP, packet transferred per second
and delay in the profile definition. Similarly, each computer that we wish to get result should
be configured to simulate that service, and then get the results in the end.
The computer network is installed in an area of 3000 m2 or (3 km2), the colleges are distrib‐
uted around the central building of the university. The presidency location assumed to be as
one of the college networks (See Figs. 1, 2 and 3). Then, 16 college networks can be
obtained, it is sufficient for the Salahaddin university hierarchy configurations.
zamfira@unitbv.ro
A Virtualized Computer Network for Salahaddin University 735
In this way, it is noticed that each college switch is of 900 m distance from the central
switch, and this fulfills the cable media specification of physical layer. Fiber optic cable
1000 BaseX can connect computers disparities of 950 m by each other.
The college networks consist of 20 computers distributed around each college loca‐
tion. The college computer network should be wired by STP cable interconnecting each
computer with the central switch of 64 Port switch ready for future expansions as shown
in the Fig. 2.
All the college switches are connected to the main switch by a fiber optic cable 1000
Base X. This is connected to the main router and the servers of firewall, and the appli‐
cation by a switch 5.3. The Main Router Profile Definition in the profile Definition is
configured the application that we wanted to simulate in this network project. This
application should be configured on each node (PCs), or else, the Pc cannot get benefit
from that specific configured application. Also, the server should be configured to get
this application profiles.
Addresses can be configured to all devices in the layer 3 (network). In this project,
IP addresses are configured to each device in the network project.
zamfira@unitbv.ro
736 T.A. Rashid and A.O. Barznji
This paper determines several results about the ability of using the OPNET to simulate the
Salahaddin University-Erbil campus by all its sites (locations) and type of services like HTTP
and FTP with measuring the network parameters of time delay for Ethernet and the
throughput. The Ethernet Delay was decreased slightly until it was finally stabilized at
0.00007 s as shown in Fig. 4. The Cloud Switch queuing delay for the point-to-point from
the cloud to switch starting by 0.0000078 as shown in Fig. 5. The Global statistics Traffic
Sent Byte/sec reached 50000 byte/sec as shown in Fig. 6. The voice traffic sent stabilized
at 480000 Byte/sec. The Global statistics VoIP Traffic Sent Byte/sec reached 500,000 Byte/
sec. It is noticed from the curves obtained that the file transfer HTTP data Traffic received
response which was started with zero, then increased reaching 85000 byte/sec, after 4 min
from the beginning, and the simulation stabilized at approximately 17000 byte/sec as shown
in Fig. 7. The Global statistics RIP HTTP data Traffic Sent reached 105 bit/sec for maximum
value, then decreased after an interval of 20 min reaching only 2 bit/sec as the network
stabilized after obtaining IP addresses by each of the connected devices (See Fig. 8). The
Server to Switch point to point throughput in bit/sec for the network speed reached
3000000 bit/sec, then stabilized at 2700000 bit/sec as shown in Fig. 9. This high bit rate was
sufficient to provide the services to all the university users, besides, the cable capability to
carry this load as the specification mention in Table 1, in addition to what mentioned in the
Sect. 3, the cable (router and switch) that used in this network portion has the capacity to
handle or carry a bit rate of 100000000 bit/sec. Global statistics web browsing heavy Server
HTTP Traffic received bit/sec for web browsing stabilized at 4000 Byte/sec (See Fig. 10).
All these determined values are sufficient to serve the users of the university for several
applications like HTTP. Thus, it is important for future work to study, and make simula‐
tions for other applications such as FTP, VoIP etc.
Fig. 4. The Ethernet Delay is decreased slightly until it gets stabilized at 0.00007 s.
zamfira@unitbv.ro
A Virtualized Computer Network for Salahaddin University 737
Fig. 5. The Cloud Switch queuing delay for the point-to-point from the cloud to switch Starting
by 0.0000078.
Fig. 6. The global statistics Traffic Sent Byte/sec reached 50000 byte/sec.
zamfira@unitbv.ro
738 T.A. Rashid and A.O. Barznji
Fig. 7. The global statistics web browsing HTTP traffic received Byte/sec reached
17000 Byte/sec.
Fig. 8. The network stabilized after obtaining IP addresses by each of the connected devices.
zamfira@unitbv.ro
A Virtualized Computer Network for Salahaddin University 739
Fig. 9. The Server to Switch point to point throughput in bit/sec for the network speed reached
3000000 bit/sec then stabilized at 2700000 bit/sec.
Fig. 10. Global statistics web browsing heavy Server HTTP Traffic received B/S for web
browsing stabilized at 4000 Byte/sec
6 Conclusion
In this paper, different factors are studied and examined to determine a computer network
performance and its capabilities. Examples of these factors are time delay, throughput,
bandwidth, transfer of data, packet transfer, packet delay, http transfer, congestion,
collision, Ethernet specifications, VoIP etc. The OPNET is used for developing a
computer network simulation for Salahaddin University-Erbil measuring HTTP service
values. In addition, the model is made without high speed medias and devices to make
the design dependable and satisfactory for handling the university users’ requests. The
zamfira@unitbv.ro
740 T.A. Rashid and A.O. Barznji
OPNET can produce more and better abundant, comprehensive and accurate results for
networks (and its components such as routers, switches, etc.) than other programs which
are used for topological representation for these networks. In other words, the OPNET
takes into consideration the specifics of technical features for each component used for
connecting the networks. For example, the way of connecting each component or device
via what type of cable or what features the cable has or the kind of pros and cons this
cable has.
References
1. Hammoudi, M.: Building model for the University of Mosul computer network using OPNET
Simulator. Tikrit J. Eng. Sci. 18(2), 34–44 (2011)
2. Sharma, S.: Design and implement the hybrid network for different IP routing protocols and
comparative study thereof. Inf. Assur. Secur. Lett. 1, 035–040 (2010)
3. Mulyawan, B.: Campus network design and implementation using top down approach: a case
study Tarumanagara University. In: Proceedings of the 1st International Conference on
Information Systems For Business Competitiveness (ICISBC) (2011)
4. Jaswal, K., Jyoti, K.V.: OPNET based simulation and investigation of Wimax network using
different Qos. Int. J. Res. Eng. Technol. 3(5), 575–579 (2014)
5. Mehta, P., Baghla, S.: Performance evaluation of heterogeneous networks for various
applications using OPNET modeller. Int. J. Recent Innov. Trends Comput. Commun. 3(6),
4003–4006 (2015)
zamfira@unitbv.ro
Online Engineering
zamfira@unitbv.ro
GIFT - An Integrated Development and Training System
for Finite State Machine Based Approaches
1 Introduction
With our hybrid online lab GOLDi (Grid of Online Lab Devices Ilmenau), described in
several papers [1–6], we support the design process of digital control systems, which
usually consists of the conceptual formulation and the design of the control algorithm
to finally achieve a validated control (see Fig. 1). For the functional description, we offer
different specification techniques by using noncommercial development tools for
various web-based control units in the remote lab:
• a Finite State Machine (FSM) based design on the basis of digital automata - executed
within a client-side FSM interpreter
• a software-oriented design in C or assembler executed on microcontrollers
• a hardware-oriented design in hardware description languages or schematic block
design by using FPGA’s
Simulation and visual prototyping help to identify functional errors before starting
practical work on real physical systems (the electro-mechanical models) in the lab room
with one of the selected control units.
zamfira@unitbv.ro
744 K. Henke et al.
While several professional or semi-professional design tools are available for soft‐
ware and hardware oriented approaches, the FSM based design is only partially
supported. That’s why we enhanced the functionality of our GOLDi system by the GIFT
(Graphical Interactive Finite State Machine Toolset) system which will be described in
detail within this contribution. The GIFT system is an integral part of the GOLDi remote
lab infrastructure. The first conceptual ideas for this GIFT system were presented during
REV2016 [7].
In the following, we will demonstrate how students can use the GIFT system to:
• specify the given task by using Finite State Machines
• simulate their design within the GIFT system to achieve a faultless solution and
• export the generated next-state and output equations to the GOLDi remote lab infra‐
structure to test their solution under real environmental conditions with electrome‐
chanical hardware models.
Based on the flexible grid structure of the GOLDi system an experiment consists of two
components: on the one hand there are various control units (e.g. FSM, microcontroller,
FPGA). On the other hand, there are the electro-mechanical physical systems (e.g.
zamfira@unitbv.ro
GIFT - An Integrated Development and Training System 745
The Design Environment is used to specify the user developed design by using a spec‐
ification technique determined by the control unit. For FSM based designs, this Design
Environment is the GIFT system which provides methods for:
• general FSM administration
• input the FSM as automaton graph or transition table
• handling of transition conditions and coding of states
• simulation of single and parallel automata
• generation of the next-state (z) and output (y) equations
• export of z/y equations to the ECP of the GOLDi system
• generation equations for D and JK flip-flops
Other examples for Design Environments are Atmel Studio [8], in case of microcon‐
troller based control unit for a software-oriented implementation, or Quartus Prime from
Altera [9], in case of FPGA based control unit, for a more hardware-oriented design. The
Design Environment is not limited to specification - it provides also support for syntax check
of the design in the specification language as well as automatic tools for verification and
validation. An important feature is also the possibility to generate output data (e.g. GIFT
export data or Atmel *.hex, Altera *.pof binary files), which can finally be loaded into the
Execution Environment.
zamfira@unitbv.ro
746 K. Henke et al.
Fig. 3. I/O monitor (e.g. sensor signals) of the ECP for the 3-axis portal
zamfira@unitbv.ro
GIFT - An Integrated Development and Training System 747
• choose an individual initial situation for the experiment by manipulating the physical
system via mouse or keypad
By using Finite State Machines as specification technique to realize various control tasks, the
graphical interactive FSM toolset GIFT can be used directly within the GOLDi design flow
(see Fig. 4).
As an example, to introduce Finite State Machines for students in the first semester, we will
discuss a design task by using the electro-mechanical model “3-Axis portal” from the
GOLDi remote lab infrastructure:
“On one spindle of a 3-Axis Portal crane, a tool carriage can be moved to the right and
to the left. Limit switches provide input information on the left end position (xl) as well as the
right end position (xr) of the tool carriage (xr, xl).
The motion can be controlled via the output variables (yl, yr) between
• motion to the left (yl = 1, yr = 0),
• motion to the right (yl = 0, yr = 1) and
• stop (yl = yr = 0).
An additional input variable xs signalizes
• stop motion (xs = 0) or
• movement (xs = 1) to the left or right.
After a possible break, the movement in the original direction should be continued.”
In the following we demonstrate how students can specify their design based upon the given
task by using Finite State Machines, simulate their design within the GIFT system to achieve
a faultless solution step by step and can finally export the generated next-state and output
equations to the ECP within the GOLDi remote lab infrastructure.
zamfira@unitbv.ro
748 K. Henke et al.
Figure 5 shows two possibilities of corresponding automaton graphs by using the graph
editor of the GIFT system:
Fig. 5. Mealy and Moore automata for the spindle control task
zamfira@unitbv.ro
GIFT - An Integrated Development and Training System 749
Because the design is typically made with user-defined variables (e.g. xl, yr) the
student must adapt them to the real sensor and actuator interface notations (e.g. xr must
be replaced by x00 of the 3-Axis Portal sensor signals, which means “X-Axis: Crane
position right”). This can be done interactively for each input and output variable within
the ECP (see Fig. 8).
zamfira@unitbv.ro
750 K. Henke et al.
In this chapter, further applications of the GIFT tool beside the standard GOLDi devel‐
opment process mentioned above will be explained. These applications are used along‐
side our teaching models at Ilmenau University of Technology and have been proven to
be successful in teaching in depth knowledge of techniques to achieve flawless designs
based of Finite State Machines.
Besides teaching the basics of FSM’s the GIFT system will be used in undergraduate
lessons to demonstrate different, unexpected side effects which can be occur when a
FSM based specification is incomplete and/or contains contradictions. The GIFT system
can demonstrate the resulting faulty behavior of the design task and gives information
how to avoid them.
Due to the intuitive approach for the design of automata graphs it could be possible
that students.
• forget to specify all next-state conditions which lead to an incomplete design or
• specify inconsistent next-state conditions which lead to a contradictory design.
In both cases it is not recommended to realize such faulty designs because the
behavior of the implemented design is normally completely different as the desired
behavior. Figure 9 (left) gives an example for a faulty Moore automaton of the spindle
design task with contradictions in states Z1 and Z3. The GIFT system automatically
indicates an incorrect design by highlighting the faulty state transitions in red and calcu‐
lates the behavior of the resulting design based on the faulty next-state functions (Fig. 9,
right). The correct Moore automaton graph is shown in Fig. 5, right.
zamfira@unitbv.ro
GIFT - An Integrated Development and Training System 751
The idea is to stop the motion during the “motion to left” (state Z1) or “motion to
right” (state Z3) to switch to the two “stop” states Z0 or Z2 by deactivating the xs variable.
The motion to left or right will be continued by activating xs again. Due to the intuitive
design, students often ignore, that it is only possible to change the motion direction
(between Z1 and Z3 or vice versa) or stay in these states when xs is activated.
This will otherwise result in the mentioned contradictions. Although the resulting
design can be activated to drive to the left or right direction from the two “stop” states
Z0 or Z2 it is never possible to stop the motion again (see Fig. 9, right). In the worst case
the motion will not be stopped in Z3 if xs is deactivated and the right end position is
reached – which finally could damage the electro-mechanical model (or the hardware
model in the f2f laboratory without any further protection mechanisms).
Fig. 9. GIFT: Incorrect FSM design (left) and resulting automaton graph (right) for the given
spindle control task
Fig. 10. ECP: Termination of the design execution due to design errors
zamfira@unitbv.ro
752 K. Henke et al.
If students decide to ignore the automatic design checks of the GIFT system and
export faulty next-state and output functions to the ECP of the GOLDi system and start
the execution, the implemented protection unit will finally terminate the execution to
avoid any damages of the electro-mechanical hardware models (which is shown in
Fig. 10).
In preparation of the hands-on laboratories students can enter and simulate their FSM
design for the given lab task. The GIFT assisted preparation process for hands-on lab
sessions is shown in Fig. 11. In contrast to the automated GIFT integration into the
GOLDi infrastructure and the execution of the design in the ECP, for hands-on lab
sessions the students must realize their design by manually connecting integrated circuits
(e.g. AND/NAND, OR/NOR Gates, D/JK flip- flops) by wires.
Fig. 12. GIFT: Export of minimized D Flip-Flop and JK Flip-Flop equations (left) for hands-on
lab session (right)
zamfira@unitbv.ro
GIFT - An Integrated Development and Training System 753
Equations for D and JK flip-flops (Fig. 12, left) as well as output equations can be
generated by the GIFT system. During the hands-on laboratory students will build up a
sequential circuit on the basis of these next-state and output functions to control simple
technical facilities. While the results achieved within the GOLDi infrastructure are self-
assessed, in hands-on lab sessions this is done by a teacher/tutor (Fig. 13, right).
Fig. 13. GIFT: Automaton graph for the water level control task (left) and the specification
scheme (right)
zamfira@unitbv.ro
754 K. Henke et al.
the pumps in a way that the water level remains between the lowest and the highest. If only one
pump is active, they should be alternatively be used.”
Fig. 14. GIFT: Decomposition into two parallel automata with 3 and 2 states
The number of active pumps as well as the alternation is described in the specification
scheme (Fig. 13, right).
One example for decomposition into two parallel automata is shown in Fig. 14. The
first parallel automaton (Fig. 14, left) is responsible for the number of active pumps:
• Z0: both pumps are inactive
• Z1: one pump is active – depending the actual state of the second parallel automaton
• Z2: both pumps are active
The second parallel automaton (Fig. 14, right) switches the states Z0 and state Z1
according to the specification scheme (shown in Fig. 13, right).
Fig. 15. ECP: Execution of parallel water level control design in virtual mode
zamfira@unitbv.ro
GIFT - An Integrated Development and Training System 755
Fig. 16. ECP: Execution of a control task for the “Production Cell” with 12 parallel automata
5 Conclusion
zamfira@unitbv.ro
756 K. Henke et al.
Acknowledgment. The authors would like to acknowledge the work of Tobias Vietzke, Andrey
Yelmanov, Lisa-Marie Schilling, Nicole Ponischil, Lennart Planz, Stephen Ahmad, Bastian
Hellweg and David Sukiennik for their work within the GIFT and GOLDi framework.
This work was supported in part by the European Commission within the program “Tempus”,
“ICo-op – Industrial Cooperation and Creative Engineering Education based on Remote
Engineering and Virtual Instrumentation”, Grant No. 530278-TEMPUS-1-2012-1-DE-TEMPUS-
JPHES [10] as well as “DesIRE - Development of Embedded System Courses with
implementation of Innovative Virtual approaches for integration of Research, Education and
Production in UA, GE, AM”, Grant No. 544091-TEMPUS-1-2013-1-BE-TEMPUS-JPCR [11].
References
1. Poliakov, M., Wuttke, H.-D., Larionova, T., Henke, K.: Automated testing physical models
in remote laboratories by control event streams. In: International Conference on Interactive
Mobile Communication, Technologies and Learning, San Diego, CA, USA, October 2016
2. Henke, K., Vietzke, T., Wuttke, H.-D., Ostendorff, S.: GOLDi – Grid of Online Lab Devices
Ilmenau. Int. J. Online Eng. (iJOE) 12(04), 11–13 (2016). Vienna, Austria, April 2016, ISSN
1861-2121
3. Henke, K., Vietzke, T., Wuttke, H.-D., Ostendorff, S.: GOLDi – Grid of Online Lab Devices
Ilmenau. In: Demonstration of Online Experimentation exp.at 2015 International Conference,
São Miguel Island, Azores, Portugal, June 2015
4. Henke, K., Wuttke, H.-D., Vietzke, T., Ostendorff, S.: Using interactive hybrid online labs
for rapid prototyping of digital systems. Int. J. Online Eng. (iJOE) 6, 57–62 (2014). Vienna,
October 2014
5. Henke, K., Ostendorff, S., Wuttke, H.-D., Vietzke, T., Lutze, C.: Fields of applications for
hybrid online labs. Int. J. Online Eng. (iJOE) 9, 20–30 (2013). Vienna, May 2013
6. Henke, K., Ostendorff, S., Vogel, S., Wuttke, H.-D.: A grid concept for reliable, flexible and
robust remote engineering laboratories. Int. J. Online Eng. (iJOE) 8, 42–49 (2012). Vienna,
December 2012
7. Henke, K., Vietzke, T., Hutschenreuter, R., Wuttke, H.-D.: The remote lab cloud “goldi-
labs.net”. In: 13th International Conference on Remote Engineering and Virtual
Instrumentation REV 2016, Madrid, February 2016
8. Atmel Corporation. http://www.atmel.com
zamfira@unitbv.ro
GIFT - An Integrated Development and Training System 757
zamfira@unitbv.ro
A Web-Based Tool for Biomedical Signal Management
Abstract. The paper deals with the implementation and development of web-
based platform, named WebSA 2.0, oriented to management of biomedical
signals within a database. It comes from the need to create a space which makes
easier sharing of biomedical signals from different sources whose are under digital
processing for supporting biomedical research. The use of web technology with
that purpose permits to enlarge the scopes of the system, as well as to add valuable
services of any I+D environment. Four types of biomedical signals are consid‐
ered: cry signal, electroencephalogram signal (EEG), electrocardiogram signal
(ECG) and electroculogram signal (EOG). The performance of the collaborative
web-based system was tested within the intranet and for several Windows stand‐
ards with satisfactory results. The WebSA 2.0 system could be useful for any
research situation in which the digital processing of different biomedical signals
be involved.
1 Introduction
The Neurosciences and Image and Signals Processing Study Center (CENPIS) of the
Universidad de Oriente has been working on projects related to the processing of
different biomedical signals to apply in the medical clinic (neonatal diagnosis, anesthesia
monitoring, devices detection in ECG signals, etc.). These projects require biomedical
signal samples which must be stored on servers arranged for that purpose to guarantee
the access to them. Nowadays the implementations of web sites that address this speci‐
alized subject are rare in the World Wide Web (WWW). This is evident in the absence
of any web site today which dedicates its efforts to the storage of biomedical signals of
different kind with possibilities of free access and exchange of experiences and infor‐
mation among specialists.
The CENPIS particularly works on the acquisition and analysis of biomedical
signals, such as: infant crying signal (linked to the extraction of relevant acoustic features
for determination of pathologic diseases in newborn babies), electrocardiogram signals
(ECG) and electroencephalogram signals (EEG) associated to the detection of artifacts
zamfira@unitbv.ro
A Web-Based Tool for Biomedical Signal Management 759
and robustness to noise, as well as electroculogram signals (EOG) with a pretty hard
impact on the diagnosis of hereditary ataxia. Based on the signal samples provided by
specialized hospitals for each signal or the recordings done by the center staff, several
software products have been developed to support the study and diagnosis of different
diseases. As this information doesn’t have a recollection protocol and a unified access
within the center, it avoids the possibility to manage the signals and information by the
researchers. This situation is worst when young researchers start working because they
lack of a unified tool where they can easily obtain biomedical signals, that are the basis
of the DSP (Digital Signal Processing) techniques that are carried out by the CENPIS
researchers [1–3]. Moreover this situation is applied to allied research groups that
support their work in DSP not only in our country but in the rest of the world. So while
the work is aimed to solve this absence in the CENPIS its usefulness extends beyond
our institution.
The aim of this work is to develop a web platform that integrates and facilitate access
and management of biomedical signals from various sources as well as the resources
sharing among DSP researchers around the world.
Four biomedical signals are held in the system which is used by the CENPIS researchers:
newborn cries, electroencephalograms (EEG), electrocardiograms (ECG) and ocular
movements (EOG).
Crying: crying signal consists of WAV file with 12 s. cry recording, with all the
newborns placed in a cubito supine position and a pinching in the calcaneus as a consis‐
tent pain stimulus. Associated with each signal a basic clinical information regarding
neurophysiologic status of baby at birth taken is properly collected from their clinic
profile [2].
Electroencephalogram and Electrocardiogram: These two signals in the system are
treated as one, since both are contained in one data record in which 19 channels corre‐
spond to electroencephalogram and one channel to electrocardiogram, thus counting the
overall record with 20 channels. They are obtained from a MEDICID device1, which
digitally emits a set of five files with different extensions and structure; and depending
on each extension specific information is provided.
The file extensions that are generated in each record are: .PAT, .INF, .PLG, .MRK
and .CDC.
.PAT: It contains clinical patient data, such as name, sex, age, etc. It is a text file.
.INF: It contains information on the characteristics of the record: frequency cuts of
filters, sampling period, etc.
.PLG: It contains signals recorded in units of A/D converter.
1
Medical equipment developed by the Cuban Institute for Digital Technologies, and located in
many Cuban hospitals.
zamfira@unitbv.ro
760 S.D. Cano-Ortiz et al.
.MRK: It contains states, trademarks and other information inserted in the record. It
is an ASCII file (with characters and integers (32 bits)).
.CDC: It contains calibration values and DC level for each channel recorded. It is a
file of floating-point values (single 8087) with the following structure:
Cal (1) DC (1) Cal (2) DC (2) … Cal (n) DC (n)
where each Cal (i) DC (i) is the calibration value and DC level of the i-th channel.
In order to obtain any value in microvolt, simply perform the following operation:
The criteria for the system design were based on several factors: type of code, degree
of development, market position, usability, accessibility, download speed, functionality.
The client/server architecture was chosen as architecture in which client and server can
act as a single entity or as separate entities, adding the possibility of belonging to the
same platform or to different platforms as well [4, 5]. For the implementation of the
system Apache was used as a web server, MySQL-Server was used for database
management and Joomla as content management system (CMS) [2, 5–11].
Figure 1 shows the layout of the client/server architecture used in this work:
The user’s web browser sends a request to the web server who retrieves the file and
passes it to the PHP engine for processing. The PHP engine analyzes the script, opens
a connection and sends the query to the MySQL server, which is responsible for
processing it and returns the result to the PHP engine that finishes executing the script,
depending on the case involves application HTML to format the results, and finally
zamfira@unitbv.ro
A Web-Based Tool for Biomedical Signal Management 761
returns it to the web server. The web server returns the HTML code to the user’s browser
for its viewing.
Given the problematical situation of the research, automation object and the concepts
presented in the foregoing sections, a 3-layer distributed system based on a web interface
was implemented, a system able to manage new signals, their own samples and the access
to them by users or customers (e.g. researchers from or related to the CENPIS). This system
was implemented with the CMS Joomla not only to keep informed the scientific community
about the current work of CENPIS but also to enable communication among their members
through the configuration of different services as: news, articles, forums, blog, private
messaging, etc. With the implementation and deployment of this system a solution to the
research problem is given in compliance with aims of work.
In this section the performance of the web system resulting from analysis and design
phases stated above is described. This system manages four biomedical signals needed
by the CENPIS research staff to carry out their investigations. For the design of the
interface all functionalities provided by the CMS were considered, using a template
designed to fix the style of the interface. In that sense some features from the design of
the previous platform (version 1.0) are included.
The system was evaluated in terms of CENPIS intranet and with the participation of
researchers working with the four biomedical signals described, All the aspects
concerned with the data management section within the corpus were successfully tested,
where the user was able to access by simply clicking on one of the four options of the
type of signal present in the Corpus. Moreover the user also has got the option to access
other services such as: online chat (e.g. instant communication with administrator or
other remote user), blogging to discuss ideas, new procedures, etc., e-mail for commu‐
nication with users within Intranet or Internet, repository of technical information, links
to other related sites, etc.
zamfira@unitbv.ro
762 S.D. Cano-Ortiz et al.
Finally all the proposed functionalities were tested with satisfactory results. Tests
showed compliance with requirements in the design stage as:
Usability: This system can be used by all researchers, doctors, students, interested in
analyzing the signals offered on site (showed consistency in its simple interface,
limited technical language, and guiding ease of getting around on the site).
Performance: Quick access to the pages, thanks to a right use of resources that are
available in the Client/Server model, and the speed of queries in the database.
Portability: The system is able to run on Microsoft Windows operating systems (e.g.
run successfully for Win 98, XP, Vista, 7 and 10).
Security: different roles were defined according to the activity that corresponds to the
user in the system, leaving well marked privileges corresponding to each of the role
user, and taking into account the vulnerability of data provided [12].
Appearance and Interface: The interface of the system is through a Web browser (with
a simple and explicit design that allows the user to interact with the system without
needing a deep training).
Help and online documentation: The system has a help page of the site, which is offered
to online user, with basic instruction on how to navigate the site (also has a site map
and a page which provides information about site developers).
Dependencies and relations with other software: The WebSA 2.0 platform is inserted
into collaborative environment developed software that integrates different DSP-based
tools developed by CENPIS researchers2. WebSA 2.0 also fixes some problems
presented in the initial version of WebSA on Cry3 [13].
4 Conclusions
In this research a system with web technology aimed to facilitate a remote access and
management of a Corpus data with biomedical signals of different nature (acoustic
newborn cry signal, ECG signal, EEG signal, and EOG signal) is developed, linked to
a research environment where digital processing of biomedical signals is used (e.g.
CENPIS). Incorporating web technology system adds useful features to disseminate,
share and manage information (through chat services, blog, e-mail). The system proved
all its functions successfully in the CENPIS intranet and for different configurations of
Windows. The use of open source (free software), object-oriented standards and porta‐
bility make it a useful tool for other users or potential users of other scientific or academic
institutions that manage biomedical signal processing. Next step in the work will be a
real-time performance in the Web which allows the participation of foreign academic
institutions in the evaluation of the WebSA system.
2
There is a link to MediCry 1.0 (MySQL database with all clinical and biomedical information
of newborn babies whose cries were previously recorded), and to the CryTrainer 1.0 (it is a
web-based trainer oriented to learn how to read spectrograms of infant cry signal with potentials
for newborn diagnosis).
3
WebSA on Cry (or WebSA 1.0) was developed in January 2007 by CENPIS researchers for cry
signals only.
zamfira@unitbv.ro
A Web-Based Tool for Biomedical Signal Management 763
Acknowledgements. Part of this research was made thanks to the financial support derived from
the Webbasierte FuE-Plattform Zur Signalanalyse Project (WebSA) in collaboration with the
University of Applied Sciences from Dusseldorf, Germany and the INAOE (México).
References
zamfira@unitbv.ro
Optimization of Practical Work for Programming Courses
in the Context of Distance Education
Amadou Dahirou Gueye1 ✉ , Pape Mamadou Djidiack Faye2, and Claude Lishou3
( )
1
Alioune Diop University, Bambey, Senegal
dahirou.gueye@uadb.edu.sn
2
Virtual University, Dakar, Senegal
papedjidiack.faye@uvs.edu.sn
3
Cheikh Anta Diop University, Dakar, Senegal
clishou@ucad.sn
1 Introduction
zamfira@unitbv.ro
Optimization of Practical Work for Programming Courses 765
ever the driving force behind this so-called knowledge societies and particular emphasis
globally on the development of STEM (Science, Technology, Engineering and Mathe‐
matics). Science education is all about handling, experimentation and practical work.
On the other hand, distance education has grown rapidly in recent years, offering a
massive number of students from all over the world the best courses in the best univer‐
sities.
However, distance learning designers and their platforms are more and more worrying
about learners focusing on the need for practical work indispensable in certain types of
education. Hence the obvious interest to anticipate the needs, expectations and potential
problems of learners in the design of online training in order to meet the challenges of
distance, isolation or remoteness between the different actors [1, 2]. To do so, many
distance learning platforms exist and incorporate self-assessment tools and online assess‐
ment [3]. This is the case of Moodle, which is used by many universities [4]. Moreover,
we find that the main platforms of MOOCs have already taken the trouble to provide
answers to these challenges: in a few months, Coursera [5], EDX [6] and Udacity [7] have
evolved in the direction better ergonomics of navigation and especially adding features to
fully participate in interactions which, even in academic MOOCs, are learning. On the
other hand, we also realize that Google and Openfire use WebRTC API to provide tools to
communicate in real time from browsers. These Google tools could be used to enable
learners and teachers to communicate in real time. The authors in [8] proposed a virtual
classroom solution easily integrated into platforms as distance learning and mathematics
allows students to have no limits for practical work by allowing them the mathematical
formulas entry.
However, we should think more to extend the functionality of learning platforms
with the aim of increasing the possibilities of collaboration focusing on achieving prac‐
tical work of scientific disciplines. The field of Science Technology Engineering and
Mathematics (STEM) retained in this paper mainly computer programming.
Currently, solutions of practical work as a plug-in are available to extend the func‐
tionality of remote training platforms [6, 8]. These solutions, while integrating video
features, audio, chat, screen sharing, audio, are generic while the requirements to achieve
practical work may differ depending on the specialty. For literary disciplines, learners
just need to see and hear the teacher while for others like computers, teachers and learners
need to implement computer programs. Despite the existence of the Screen Sharing
feature in the virtual classroom solutions, remote training platforms do not offer the
ability to properly carry programming during practical work. For the latter, it is not only
having visibility into the work of a participant but to create an interactive environment
between the participants. This interactivity cannot be managed with screen sharing
solutions that consume much bandwidth. In this paper, we propose an optimization
solution of practical work in programming, easily integrated in a distance education
platform. The rest of the paper is organized as follows: in Sect. 2, we first present the
state of the art on Node.js, WebRTC technology and architecture of a standard compiler.
Then we present in Sects. 3 and 4 the integration architecture of our solution in a training
platform and some scenarios of operation. Section 5 describes the presentation of results.
The conclusion summarizes the results and perspectives.
zamfira@unitbv.ro
766 A.D. Gueye et al.
2 Related Work
First, we present the Node.js and WebRTC technologies used in our approach and the
architecture of a standard compiler.
2.1 Node.js
Node.js is one recent framework to implement the event model through the entire stack.
Developed in 2009 by Ryan Dahl, Node.js (or just Node) is a single-threaded server-
side JavaScript environment implemented in C and C++ [9]. Node.js architecture makes
it easy to use as an expressive, functional language for server-side programming that’s
popular among developers [9]. Node.js use the JavaScript V8 engine, developed by
Google [10], a fast and powerful implementation of JavaScript [11] that helps Node
achieve top performance.
Unlike other modern environments, a Node process doesn’t rely on multithreading
to support concurrent execution; it’s based on an asynchronous, event driven I/O model
[11]. Its event driven, non-blocking I/O model makes it lightweight and efficient, ideal
for data intensive real time applications that run across distributed servers [12].
For better or worse, JavaScript is the world’s most popular programming language.
If you’ve done any programming for the web, it’s unavoidable. JavaScript, because of
the sheer reach of the web, has fulfilled the “write once, run anywhere” dream that Java
had back in the 1990s [13, 14]. Around the time of the Ajax revolution in 2005, Java‐
Script went from being a “toy” language to something people wrote real and significant
programs with. Some of the notable firsts were Google Maps and Gmail, but today there
are a host of web applications from Twitter to Facebook to GitHub. Since the release of
Google Chrome in late 2008, JavaScript performance has improved at an incredibly fast
rate due to heavy competition between browser vendors (Mozilla, Microsoft, Apple,
Opera, and Google). The performance of these modern JavaScript virtual machines is
literally changing the types of applications you can build on the web.2 A compelling,
and frankly mind-blowing, example of this is jslinux,3 a PC emulator running in Java‐
Script where you can load a Linux kernel, interact with the terminal session, and compile
a C program, all in your browser [13, 14].
Node uses V8, the virtual machine that powers Google Chrome, for server-side
programming. V8 gives Node a huge boost in performance because it cuts out the
middleman, preferring straight compilation into native machine code over executing
byte code or using an interpreter. Because Node uses JavaScript on the server, there are
also other benefits [9, 11]:
Its goal is to offer an easy and safe way to build high performance and scalable
network applications in JavaScript. Those goals are achieved thanks it’s architecture
[13, 14]:
Single Threaded:
Node use a single thread to run instead of other server like Apache HTTP who spawn
a thread per request, this approach result in avoiding CPU context switching and massive
execution stacks in memory. This is also the method used by Nginx and other servers
developed to counter the C10K problem [12].
zamfira@unitbv.ro
Optimization of Practical Work for Programming Courses 767
Event Loop: Written in C++ using the Marc Lehman’s libev library, the event loop
use epoll or kqueue for scalable event notification mechanism [13, 15].
Non-blocking I/O: Node avoid CPU time loss usually made by waiting for an input
or an output response (database, file system, web service) thanks to the full-featured
asynchronous I/O provided by Marc Lehmann’s libeio library [13].
These characteristics allow Node to handle a large amount of traffic by handling as
quickly as possible a request to free the thread for the next one.
Node has a built-in support for most important protocols like TCP, DNS, and HTTP
(the one that we will focus on). The design goal of a Node application is that any function
performing an I/O must use a callback. That’s why there is no blocking methods provided
in Node’s API [9].
The HTTP implementation offered by Node is very complete and natively support
chunked request and response (very useful since we are going to use the twitter streaming
API) and hanging request for comet applications. The Node’s footprint for each http
stream is only 36 bytes (source) [12, 13]. Figure 1 shows the Node.js architecture.
In our solution, Node.js acts a very important role. It allows us to manage commu‐
nications between the web client and the compiler. Thus the messages received by the
server Node.js are transformed into messages understandable by the compiler. Message
formats will therefore be defined between the client and the server Node.js.
WebRTC is an open source project introduced by Google in 2011 [16] which ensures
communications in real time via a JavaScript API. [17] The project aims to develop a
technology allowing web browsers to support interactive communications point to point
and offer data exchange synchronous mode [18, 19]. The WebRTC is intended to give
browsers the ability to offer audio, video or written, file transfer, screen sharing and
remote control of computers.
The main components of the WebRTC API defined by the W3C working groups
(World Wide Web Consortium) and IETF (Internet Engineering Task Force) are [20, 21]:
MediaStream: allows a browser to access the camera and microphone;
RTCPeerConnection: Enables audio and video calls;
RTCDataChannel: allows browsers to send data in a peer-to-peer connection.
zamfira@unitbv.ro
768 A.D. Gueye et al.
Figure 2 shows, first WebRTC C++ API implemented in some browsers and also
the web WebRTC API allowing web developers to integrate services offered by the
WebRTC into their applications.
In our solution the WebRTC will be used to enable audio communication between
participants.
zamfira@unitbv.ro
Optimization of Practical Work for Programming Courses 769
Our solution is to provide a compiler hosted in a remote server to allow multi-user access
and redirect input output (I/O) to the client and not on the computer hosting the compiler.
In other words, the proposed compiler will be able to run a received remote source code
and redirect I/O to the client. The client gets to compile his program remotely via the
server Node.js (Fig. 4).
3.1 Components
Client + browser: The client accesses the application through a web browser.
Node.js server and WebRTC: offer video conferencing functionality between clients
and also allows communication between clients and the compiler.
The compiler: it is the key element of our solution. It is composed of three entities:
1. Sockets: allow remote communication with clients.
2. The code replacer: replace the input and standard output functions in code, with
input redirection functions outputs.
zamfira@unitbv.ro
770 A.D. Gueye et al.
3. Executor hot code: This is the element that will be in charge to run the client source
code.
4. Database: It allows the compiler to temporarily store the last code compiled by a
client to avoid him to return the code to run. This is for optimizes bandwidth.
3.2 Operation
Clients communicate with each other via the Node.js. So they can chat by audio or chat.
To communicate with the compiler, the client passes by the Node.js server. Thus we
define compilation commands between the server Node.js and the compiler. Between
the client and the Node.js server, chat messages are used to send commands. Therefore,
the Node.js server that will be responsible for translating these messages understandable
format by the compiler then send them to the latter.
The compiler will compile and store the source code in a database with a unique
identifier. This identifier will be sent to it by the Node.js server during an execution
request to identify the program. Whenever the user makes a request to execute a program,
the Node.js server will send the identifier to the compiler and ask him to start the the
program execution. During the execution of the program the inputs and outputs will be
redirected to the client. That is, if the program has to display information on the screen,
this information will be transferred to the user interface and if the program has to wait
for a value, this value must be expected from the client.
At each execution, the compiler executes the last result of the compilation of the
program by receiving as a parameter the identifier of this compilation.
Customers who share the same salon will all stages of compilation and program
execution. The Node.js server will be responsible for running the run state for all other
clients.
With this solution, it’s just the code in text format that is shared in real time instead
of capturing the entire screen of the user that is likely to be too heavy, thus not facilitating
interactivity.
For the dialog between the Node.js server and the compiler there are 4 types of
commands:
ASK_COMP: This is a compilation request message.
ASK_EXEC: execution request.
ASK_INPUT: Request a value to enter.
ANSWER_INPUT: Message to send a value to the compiler following a request
message between ASK_INPUT.
END_PROG: end of program.
zamfira@unitbv.ro
Optimization of Practical Work for Programming Courses 771
1. The client sends a compilation request to the Node.js server. This request contains
the client’s source code.
2. The Node.js server is responsible for forwarding the message to the compilation
server in a format that can be understood by the server.
3. The compilation server uses code replacement functions to replace the standard I/O
functions with new functions that will redirect the input to the client.
4. The server uses the code executor to verify that the code does not contain an error.
5. The code is stored in the database with a unique identifier in case there is no error,
in order to prevent the client from returning the code during execution.
6. The result of the compilation is returned to the client via the server Node.js (Fig. 6).
zamfira@unitbv.ro
772 A.D. Gueye et al.
User NodeJS Server Socket server compiler IO replacer Executor and checker Database
compilation request
disencapsulation message
Source code ok
done
notification
response encapsulation
sending response
1. The client sends a run request that contains the identifier of the execute code.
2. The Node.js server sends the request to the compiler.
3. The compiler connects to the database to load the code based on the identifier.
4. The compiler uses the code executor to start the program.
5. The executor executes instructions 1, 2, 3 and 4.
6. The code executor executes instruction (5). As is an output instruction, it sends a
message to the server Node.js DISPLAY_OUTPUT This type of message displays
a message on the client browser.
7. The executor starts instruction (6). As this is an input instruction, it sends an
ASK_INPUT message to the Node.js server to let it know that it expects a value
from the client.
8. The client sends the value to the Node.js server, which in turn transfers it to the
compiler (Code Executor).
9. The executor launches instruction (7), which is a calculation instruction.
10. The executor executes instruction (9). As this is an output statement, it sends a
DISPLAY_OUTPUT message to the Node.js. This type of message is used to
display a message on the client browser.
zamfira@unitbv.ro
Optimization of Practical Work for Programming Courses 773
11. To show the client the end of the program, the compiler sends a message
END_EXEC (Fig. 7).
Compilateur
DiagrammeSequence_1
execution request
disencapsulation message
source code
launch program
execution instruction1,2 ,3 and 4
execution instruction 5
DISPLAY_OUTPUT
tranfer request to nodejs
execution instruction 6
Affichage sur le navigateur
execution instruction 7
execution instruction 8
DISPLAY_OUTPUT
tranfer message
END_EXEC
displaying on browser transfer message
end program
zamfira@unitbv.ro
774 A.D. Gueye et al.
5 Realization
zamfira@unitbv.ro
Optimization of Practical Work for Programming Courses 775
The user Dahirou, present on another post, continues the writing of the source code
initiated by the user Djidiack (Fig. 9).
The user Dahirou compiles the program which is visible to all (Fig. 10).
The user Djidiack executes the program which is also visible to all (Fig. 11).
6 Conclusion
In this paper we have proposed an online practical work solution which can be integrated
into a distance learning platform for computer programming courses. Access to a session
of practical work is done at a distance, via the web and several participants can
zamfira@unitbv.ro
776 A.D. Gueye et al.
simultaneously access the same session of lab. These participants share the same
compiler hosted in a remotely accessible server.
To illustrate our solution, we chose the programming language C by communicating
through a protocol that we defined our central compiler with the server Node.js which
ensures the link with the client. Thus the inputs/outputs of a program executed by the
central compiler are redirected to the client. This means that a student will no longer
need to install a compiler on his machine to participate in a practice session because it
must go through a centralized compiler that makes it possible to make visible the steps
of writing, compilation and the execution of a program to all participants in the session.
Our solution enables teachers and participants to feel at ease by providing a flexible
and interactive environment. On the one hand, it allows a learner to share the source
code or the result of a computer program with other learners in a single click; and on
the other hand, for a teacher, to manage the course of a lab with the possibility of giving
or withdrawing the hand to a learner or several learners at the same time. Our solution
also enables the teacher to ensure the team work of the learners and thus save time.
In the future, we intend to implement our solution in other programming languages
and test its integration in remote laboratories.
References
zamfira@unitbv.ro
Optimization of Practical Work for Programming Courses 777
14. Professional Node.js: Building javascript based scalable software, Pedro Teixeira (2012)
15. Node.js for PHP Developers, Daniel Howard (2013)
16. Elleuch, W.: Models for multimedia conference between browsers based on WebRTC. In:
2013 IEEE 9th International Conference on Wireless and Mobile Computing, Networking
and Communications (WiMob), pp. 279–284, 7–9 October 2013
17. Zeidan, A., Lehmann, A., Trick, U.: WebRTC enabled multimedia conferencing and
collaboration solution. In: Proceedings of World Telecommunications Congress 2014, WTC
2014, pp. 1–6, 1–3 June 2014
18. Vogt, C., Werner, M.J., Schmidt, T.C.: Leveraging WebRTC for P2P content distribution in
web browsers. In: 2013 21st IEEE International Conference on Network Protocols (ICNP),
pp. 1–2, 7–10 October 2013
19. Hinow, F., Veloso, P.P., Puyelo, C., Barrett, S., Nuallain, E.O.: P2P live video streaming in
WebRTC. In: 2014 World Congress on Computer Applications and Information Systems
(WCCAIS), pp. 1–6, 17–19 January 2014
20. Sredojev, B., Samardzija, D., Posarac, D.: WebRTC technology overview and signaling
solution design and implementation. In: 2015 38th International Convention on Information
and Communication Technology, Electronics and Microelectronics (MIPRO), pp. 1006–
1009, 25–29 May 2015
21. W3C Editor’sDraft. http://w3c.github.io/mediacapturemain/getusermedia
22. Waite, M.M., Goos, G.: Compiler construction, Karlsruhe, 22 February 1996
zamfira@unitbv.ro
Enabling the Automatic Generation of User
Interfaces for Remote Laboratories
1 Introduction
Remote laboratories (RLs) are an important component of distance and blended
learning for science and engineering education. They allow learners to experi-
ment in order to validate or refute a hypothesis, accept or reject a taught sub-
ject. By definition, they provide remote access to hands-on sessions, which are
c Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6 73
zamfira@unitbv.ro
Enabling the Automatic Generation of User Interfaces 779
essential for the process of learning and assimilating scientific concepts [2,5].
As some Web technologies emerged and died, many architectures for remote
laboratory systems have been proposed. These architectures range from being
case-specific to more generalised. The most adopted architectures, are Client-
Server based where typically the Server interfaces the physical equipment of
a lab, and the Client provides a software application through which users can
access the lab. With the birth of the ‘separation of concerns’ paradigm enabling
Service Oriented Architectures (SOA), lab providers started building their labo-
ratories in a more modular way [8,10]. With such architectures, the access to the
remote setup is done through web services or APIs where the laboratory server
is exposed as services [4,9,13]. The main aim of adopting a Service Oriented
Architecture for RLs is to separate the tiers of the remote laboratory system.
The Smart Device Specifications for remote labs describe the Server as services
through well-defined interfaces as proposed in [9]. This approach is motivated
by the complete separation of the Client from the Server encourages the broader
sharing of remote labs. The Smart Device Paradigm decouples the Client and
the Server further enabling the personalization of the client applications. When
the Smart Device Specifications are adopted for an RL, the Server is exposed
as services described by APIs enabling any person with programming skills to
create user applications to connect to the labs. This is possible by “talking” to
the APIs and understanding how client applications can access the services pro-
vided by a remote setup. With the Smart Device Specifications, the development
and deployment of remote laboratories is much easier, faster, and modular for
different stakeholders, namely the lab provider and the user client developer.
In this context, invoking a service is equivalent to controlling actuators or
retrieving data from sensors making up the laboratory setup. Provided the APIs,
it is possible to personalize the client application accessing the labs by enabling
the teachers to use the RLs in different ways, according to their educational
needs, by designing their own experiments. This is the case of remote laboratories
that are configurable and offer the flexibility of conducting different experiments,
corresponding to different scientific phenomena. In this work, we refer to the
activity which allows students to freely vary the parameters on lab equipment
as an experiment, and we refer to the combination of sensors and actuators used
in an experiment as a “configuration” from a lab owner point of view. Since
the information provided by the APIs based on the Smart Device Specifications
do not covey enough information that shows how the sensors and actuators are
connected and dependant, usually the teachers resort to the mediated contact
with a lab provider to have information about what kind of experiment(s) the lab
in question implements. On another note, creating User Interfaces (UIs) is still
largely reliant on disposing of a software developer, preventing teachers with no
such privileges from personalizing their own applications. We recognise that it is
important in such setups to give teachers the autonomy to select and create user
interfaces for remote labs that fulfil their own pedagogical objectives, without
the need to contact a lab provider through an application developer. Therefore,
the aim of this work is to support teachers in choosing the experiments and
zamfira@unitbv.ro
780 W. Halimi et al.
2 Related Work
In [7] the authors make their debut in defining Smart Devices (SDs) motivated
by the need to move away from adopting proprietary technologies for building
remote laboratories, and the need to converge towards common conventions for
designing and building remote laboratory systems. Accordingly, they re-engineer
the server side by implementing separate services for the different hardware
access which are possible for their example lab. In parallel, instead of creating
a complete web application or widget, they provide four separate ones for each
of the accessible services: a graph tool, a video feed, a control panel for the
system’s parameters, and a tool for saving the experimental data. The users of
a remote lab can choose any subset or all the provided widgets to use the lab in
a ‘metawidget’. While this effort is a move toward a personalization of the user
client, it is still proprietary for the embedding web-based environment.
Later in [8,9], the authors elaborate in more detail about the Smart Device
Paradigm and introduce the concept of LaaS (Lab as a Service). The Smart
Device Paradigm revisits the reputed Client-Server architecture for remote labs
by re-thinking the server side and equipping its component with some ‘intelli-
gence’. This is based on Thomson’s definition of smart devices, as devices which
have identity and kind, memory and status tracking, communication capabil-
ities, and more [11]. Accordingly, the Smart Device Specifications extend this
definition to support complex systems such as remote labs. Motivated by the
need to completely separate the server and client sides to further enable the
personalization of client applications, the mentioned specifications represent the
behaviour of the connected sensors and actuators as services exposed through
well-defined APIs. The services representing a sensor or actuator instance are
fully described through ‘metadata’. The ‘metadata’ provide a description of the
considered lab through the General Metadata which tells the name of the lab,
a short high-level description, a contact person, and licensing information. The
API Metadata defines the supported services by the lab, by specifying the cor-
responding sensor and actuator requests and responses. Moreover, the Smart
Device Specifications provide service descriptions for authorisation, which takes
care of user authentication. This metadata category is of no interest to this work.
zamfira@unitbv.ro
Enabling the Automatic Generation of User Interfaces 781
It is claimed that the ‘metadata’ is enough for building user applications with-
out the need for further information from a lab provider. While this might be
true for laboratories supporting one experiment, it is not true for configurable
labs which provide the possibility of conducting many experiments with the
same connected equipment. This is due to the absence of a description of how
the sensors and actuators are connected and which configurations are possible.
The Smart Device specifications provide a description of services as independent
units.
Other frameworks for the generation of remote lab user interfaces exist, such
as the tool based on EjsS in [12]. In this work, the authors bring to impor-
tance the need for user interfaces which can be well integrated in web-based
learning environments such as Moodle. Additionally they invoke the necessity to
support open web technologies and move away from Java applets which are no
longer supported by modern web browsers. While they provide a solution that
is reusable, and prevents application developers from building UIs from scratch
for each lab, this framework only supports the generation of UIs for labs which
are compatible with their implementation of the presented app builder.
zamfira@unitbv.ro
782 W. Halimi et al.
is worth mentioning that all exchanged requests and responses with the Smart
Device are JSON-encoded, further facilitating the parsing of the API calls, which
can be automatized. But what sensors to pick with which actuators? How are the
sensors and actuators working together? Are there several possible experiments
that can be conducted with the same setup, but using different sensor-actuator
configuration? All of this information is not included in the existing SD Specifica-
tions. What differentiates remote laboratories from other cyber-physical systems
is that they are built to fulfil an educational goal: conducting pre-defined experi-
ments to reflect on certain topics. With no knowledge about the interconnections
of the lab components, it is not possible to build a UI that interfaces ‘pedagog-
ically meaningful’ experiments. In this section, we extend the Smart Device
Specifications to describe the possible “configurations” or “experiments” of labs
supporting one or various experiments, further enabling the auto generation of
the user interface. We extend the detailed ‘metadata’ to add a service to the apis
which returns the configurations or experiments supported by the remote lab, in
addition to the requests and responses models. The extended SD Specifications
with the experiments service provides enough information to enable the auto-
matic generation of user interfaces without the need of the lab owner to confirm
the possibility of conducting a particular experiment. Our proposed extension is
two-fold:
1. Define the models for an Experiment, SendExperimentsRequest, SendExperi-
mentRequest, ExperiementsMetadataResponse, and ExperiementMetadataRe-
sponse
2. Define the new api calls: getExperiments and getExperiment
zamfira@unitbv.ro
Enabling the Automatic Generation of User Interfaces 783
zamfira@unitbv.ro
784 W. Halimi et al.
{"method": "Send",
"nickname": "getExperiments",
"summary": "Returns a list of possible experiments",
"notes": "Returns a JSON array with all the ids of possible experiments",
"type": "ExperimentMetadataResponse",
"parameters": [{
"name": "message",
"description": "The payload for the getExperiments service.",
"required": true,
"paramType": "message",
"type": "SimpleRequest",
"allowMultiple": false
}],
"authorizations": {},
"responseMessages": [{
"code": 402,
"message": "Too many users"},{
"code": 404,
"message": "Experiments not found"},{
"code": 405,
"message": "Method not allowed. The requested method is not allowed
by this server."},{
"code": 422,
"message": "The request body is unprocessable"
}]}
zamfira@unitbv.ro
Enabling the Automatic Generation of User Interfaces 785
"ExperimentMetadataResponse": {
"id": "ExperimentMetadataResponse",
"properties": {
"method": {
"type": "string"},
"experiments": {
"type": "array",
"items": {
"$ref": "Experiment"
}}}}
1
http://graasp.eu/.
2
https://www.edx.org/.
zamfira@unitbv.ro
786 W. Halimi et al.
The mind map in Fig. 2 shows two possible experiments that can be done with
the Mach-Zehnder interferometer, upon which we will base our explanation of
the implementation and function of the automatic UI generator in Sect. 5.
The first and second experiments are conducted in a high light intensity
setup, meaning that the density filter is not attenuating the intensity of the light
coming from the monochromatic light beam. The first experiment enables the
users to quantitatively understand light interference, by visualizing the resulting
fringes on the screen, and/or also the feed from the infrared camera, in addition
to depicting the direction in which the fringes move when the mirror mounted
with a piezo actuator manually controlled with a voltage which is increasing or
decreasing in value. In the second experiment, the students can quantitatively
study light interference by observing the emitted signal from the photo diode as
the piezo is controlled with a triangular signal causing a translation motion.
zamfira@unitbv.ro
Enabling the Automatic Generation of User Interfaces 787
When a getExperiments api call is sent to the Smart Device hosting this
laboratory, the following response is received:
{"method": "getExperiments",
"experiments": [{
"experimentId": "qualitative",
"fullName": "Qualitative Study",
"description": "Observing light interference on the screen",
"sensors": [{"sensorId": "Video"}, {"sensorId": "VideoIR"}],
"actuators": [{"actuatorId": "laser"},{"actuatorId": "piezo"},
{"actuatorId": "bs1"},{"actuatorId": "bs2"}]
},{
"experimentId": "quantitative",
"fullName": "Quantitative Study",
"description": "Studying the signal provided by the photodiode",
"sensors": [{"sensorId": "photodiode"}],
"actuators": [{"actuatorId": "laser"},{"actuatorId": "piezo"},
{"actuatorId": "bs1"},{"actuatorId": "bs2"}]
}]}
The response shows that there are two possible experiments with the exper-
imentId s “qualitative” and “quantitative”. Accordingly, the list of sensors and
actuators for each of the experiments can be either used from this response, or
retrieved by a separate call to getExperiment while passing the corresponding
experimentId.
zamfira@unitbv.ro
788 W. Halimi et al.
5.2 Implementation
The automatic UI generator is a tool that enables the creation of a fully func-
tional remote lab web client with a few clicks. The teacher needs to know the IP
address and the port number over which a Smart Device is serving the desired
remote lab. Using this information, the tool initiates a WebSocket connection
with the lab server, and subsequently call the getExperiments service, which
returns an array describing each experimental configuration supported by the
Smart Device. As mentioned in Sect. 3, each experiment is described by: the
experimentId that uniquely identifies each experiment, the fullname and descrip-
tion of the experiment, in addition to the sensors and actuators arrays that
contain the ids of all the respective sensors and actuators used by each exper-
imental configuration. These configurations are displayed as checkboxes having
the full name and the description of the experiment as their labels. The teachers
can then select one or more of the presented possible configurations according
to their educational goals. After performing this selection, the auto generator
knows the ids of all the different sensors and actuators required for each exper-
iment, and will thus send getAcutatorMedata and getSensorMetadata requests
to the lab server in order to acquire the necessary information about each (See
Fig. 3).
Fig. 3. How the automatic UI generator interacts with the Smart Device to build the UI
For actuator access, the auto generator makes use of some of the fields
obtained from the actuator metadata, in order to generate the necessary UI
components. It uses the actuatorId which uniquely identifies each actuator, to
populate the actuatorId field of the request packet which is sent to the Smart
Device whenever a user of the generated lab client alters the state of an actuator,
thus making a call to the sendActuatorData service. The auto generator also uses
the values field of the metadata, which is an array of all the measurement values
zamfira@unitbv.ro
Enabling the Automatic Generation of User Interfaces 789
zamfira@unitbv.ro
790 W. Halimi et al.
Fig. 4. The landing page of the automatic UI generator showing the two available
experiment configurations for the Mach-Zehnder lab
embedded in the platform. Graasp supports OpenSocial widgets through its own
implementation of the Shindig Apache server, enabling third party applications
to access its database for user information, and for saving and retrieving files,
actions, and other platform specific data [1]. Consequently, the OS widget will
have all the aforementioned features of the LTI-targeted application, in addition
to the following (See Fig. 5):
– Action logging: the generated graasp gadget will use the ActionLogger
library3 , which provides an easy mechanism for logging the activities of the
students. Interactions with the different UI components are saved as Activity
Streams that have the actor-verb-object format. The logged activities can
later be used to perform learning analytics.
Fig. 5. The generated remote lab client application on graasp for the Mach-Zehnder
lab
3
https://github.com/go-lab/ils/wiki/ActionLogger.
zamfira@unitbv.ro
Enabling the Automatic Generation of User Interfaces 791
– Saving experimental data: The lab application allows students to save the
actuator and sensor data that were acquired while conducting the experiment.
The data is saved in a specific format, that allows students to use it in other
applications on the platform. For example, the students can have a graphical
view of the experimental results using the Data Viewer application4 .
6 Teacher Customization
The automatic UI generator provides a basic and fully functional client applica-
tion for operating a remote lab. The UI components are very basic, and might not
be visually attractive. Using the generated code, the teachers can further person-
alize the UI appearance to fit their taste and needs. For example, a teacher in the
Gymnase de Morges in Switzerland, chose to customize the UI to be embedded
in graasp as shown in Fig. 6.
In this widget, there are two tabs to switch between two possible experiments.
In the Quantitative Study tab, there is a simulation diagram which allows stu-
dents to control the lab by clicking on the corresponding image of a component.
For example, to turn the laser beam ON/OFF it is enough to click on the box
representing the light source. On the diagram are also present the placements of
the IR camera and the normal camera allowing the student to know about the
perspective of the video feeds. In this widget, the teacher chose to only display
the video coming from Camera 2 showing the fringes on the screen. Next to it is
a graphing tool that shows the signal captured by the photo diode in real-time.
zamfira@unitbv.ro
792 W. Halimi et al.
Since the teacher doesn’t want the students to have to scroll, and since the sim-
ulation diagram conveys a real-time status of the lab, he decided that there are
enough UI components for the students to conduct the experiment while having
a good user experience.
Of course, the UI could have been customized otherwise to show the UI
components differently, or to resize them in a different way. For example, an
input box to control the piezo actuator could have been a replacement for the
slider control. Also, instead of only showing the feed of Camera 2, both feeds
from Camera 1 and Camera 2 could have been shown, in addition to the graphing
tool. All of this is possible by starting from the code provided by the automatic
UI generator. This alleviates the burden of establishing connections and parsing
the remote lab APIs, making it more easy to personalize the appearance of user
client according to a desired user experience.
References
1. Bogdanov, E., Ullrich, C., Isaksson, E., Palmér, M., Gillet, D.: From LMS to PLE:
a step forward through opensocial apps in moodle. In: International Conference on
Web-Based Learning, pp. 69–78. Springer (2012)
5
https://creativecommons.org/licenses/by-nc/2.0/.
zamfira@unitbv.ro
Enabling the Automatic Generation of User Interfaces 793
2. Dasarathy, B., Sullivan, K., Schmidt, D.C., Fisher, D.H., Porter, A.: The past,
present, and future of MOOCs and their relevance to software engineering. In:
Proceedings of the on Future of Software Engineering, pp. 212–224. ACM (2014)
3. Halimi, W., Salzmann, C., Gillet, D.: The Mach-Zehnder interferometer — a smart
remote experiment based on a software template. In: 2016 13th International Con-
ference on Remote Engineering and Virtual Instrumentation (REV), pp. 287–292.
IEEE (2016)
4. Harward, V.J., Del Alamo, J.A., Lerman, S.R., Bailey, P.H., Carpenter, J., DeLong,
K., Felknor, C., Hardison, J., Harrison, B., Jabbour, I., et al.: The iLab shared
architecture: a web services infrastructure to build communities of internet acces-
sible laboratories. Proc. IEEE 96(6), 931–950 (2008)
5. Lowe, D.: MOOLs: massive open online laboratories: an analysis of scale and feasi-
bility. In: 2014 11th International Conference on Remote Engineering and Virtual
Instrumentation (REV), pp. 1–6. IEEE (2014)
6. Pereira, A., Ostermann, F., Cavalcanti, C.: On the use of a virtual Mach-Zehnder
interferometer in the teaching of quantum mechanics. Phys. Educ. 44(3), 281
(2009)
7. Salzmann, C., Gillet, D.: Remote labs and social media: agile aggregation and
exploitation in higher engineering education. In: 2011 IEEE Global Engineering
Education Conference (EDUCON), pp. 307–311. IEEE (2011)
8. Salzmann, C., Gillet, D.: Smart device paradigm standardization for online labs.
In: 4th IEEE Global Engineering Education Conference (EDUCON) (2013)
9. Salzmann, C., Govaerts, S., Halimi, W., Gillet, D.: The smart device specification
for remote labs. In: 2015 12th International Conference on Remote Engineering
and Virtual Instrumentation (REV), pp. 199–208. IEEE (2015)
10. Tawfik, M., Salzmann, C., Gillet, D., Lowe, D., Saliah-Hassane, H., Sancristobal,
E., Castro, M.: Laboratory as a service (LAAS): a model for developing and imple-
menting remote laboratories as modular components. In: 2014 11th International
Conference on Remote Engineering and Virtual Instrumentation (REV), pp. 11–20.
IEEE (2014)
11. Thompson, C.W.: Smart devices and soft controllers. IEEE Internet Comput. 9(1),
82–85 (2005)
12. de la Torre, L., Sánchez, J., Andrade, T.F., Restivo, M.T.: Easy creation and
deployment of javascript remote labs with EjsS and moodle. In: 2016 13th Inter-
national Conference on Remote Engineering and Virtual Instrumentation (REV),
pp. 260–261. IEEE (2016)
13. Zutin, D.G., Auer, M., Ordu, P., Kreiter, C., et al.: Online lab infrastructure as
a service: a new paradigm to simplify the development and deployment of online
labs. In: 2016 13th International Conference on Remote Engineering and Virtual
Instrumentation (REV), pp. 208–214. IEEE (2016)
zamfira@unitbv.ro
A Practical Approach to Teaching Industry 4.0
Technologies
1 Introduction
There are pervasive disruptive forces of change around the world that necessitate the
Canadian manufacturing sector to undergo a transformation that is fuelled by the
Internet of Things (IoT), big data, and artificial intelligence in order to remain sus-
tainable. This imminent transformation is generally referred to in literature as Industry
4.0. Industry 4.0 is the fourth industrial revolution, following the previous three rev-
olutions, namely: the introduction of the steam engine, electricity, and information
technology. The main objective of this revolution is to develop new business models
that tap the potential optimization in production and logistics caused by increased and
integrated industrial automation, cloud computing, global databases, networked intel-
ligent system monitoring and control, and autonomous decision-making. While the
technological enablers of Industry 4.0 are all around us, a few manufacturers are truly
transforming. One of the main reasons being the general lack of trades-workers,
technicians, and engineers who are knowledgeable in Industry 4.0 concepts.
Industry 4.0 manufacturing paradigm is heavily dependent on machine to machine
communication (horizontal integration at the plant level), business and manufacturing
processes integration (vertical integration), and the value chain integration (horizontal
integration at the services level). All this integration has to be supported by a variety of
networking technologies. Tom Wanyama in his book titled “A Practical Approach to
System Integration” summarizes the Industry 4.0 networking technologies in Fig. 1.
The figure shows that Industry 4.0 depends on horizontal integration networking
technologies at the factory floor level, at the plant level where they create the Industrial
Internet of Things, at the enterprise level, and at the value chain level where they create
the Internet of Services. Moreover, the figure show that Industry 4.0 requires the use of
vertical integration technologies such as Ethernet and TCP/IP, Open Platform Com-
munications (OPC), Virtual Private Networks (VPN), and cloud computing which
enable integration of the factory floor systems with SCADA systems, as well as
business automation systems such as Manufacturing Execution Systems (MES),
Enterprise Recourse Planning (ERP) systems, and other enterprise wide manufacturing
optimization and data analytics systems [19].
zamfira@unitbv.ro
796 T. Wanyama et al.
networking technologies shown in Fig. 1. But this paper focuses on the practical aspect
of these courses which is taught through laboratories and laboratory based projects. For
this aspect, we have developed a set of laboratory infrastructure that can be used by
students to carry out laboratories and projects in the following areas:
• Machine-to-machine communication.
• Vertical integration of manufacturing systems from the factory floor to the enter-
prise and value chain level.
• Value chain integration and process optimization.
• SCADA and Human Machine Interface (HMI) development.
• Advanced system control using artificial intelligence methods such as fuzzy logic
and artificial neural networks.
Our laboratory equipment is accessible remotely through the Internet, meaning that
we use Industry 4.0 technologies to provide online hands on learning of the same
technologies. This is important because with the increased use of technology in every
aspect of human life, digital learning has become a very important form of teaching and
learning. But the future of digital learning lies in the hybridization of courses, incor-
porating in-person components into online classes. This learning framework is gen-
erally referred to in literature as blended learning. Research shows that blended
learning is exceptionally promising. Like purely digital learning, blended learning lacks
the time and space constraints imposed by in-person courses, thus much more con-
ducive for the expansion of learning time. But, unlike purely digital learning, blend
learning includes opportunities for reflection and interaction with peers and teachers.
Moreover, today’s university students already lead blended lives. They access news,
pay bills, search for vacation destinations online, and they communicate by email and
social media [7]. On the other hand, they go to movies, shop in malls, and visit friends
and family in-person. This makes blended learning an appropriate pedagogical para-
digm for today’s students. However, the need to carry out laboratory experiments
during the course of engineering and science programs is a great challenge to the online
component of blended learning.
Although Industry 4.0 technologies were developed to increase resource opti-
mization in the manufacturing industry, they can be used to provide improved remote
access to laboratory equipment. In fact many innovative hardware and software solu-
tions that are making massive changes to the industrial world in the context of Industry
4.0, are also being adopted to support teaching and learning. Common examples of
such innovations include the following:
• Cloud server based learning management systems such as Moodle that support the
offering of millions of courses to millions of students worldwide.
• Wikis which enables cooperative text production and different kinds of assessment
modes of quizzes that give teachers the chance to test students whenever they want
and as many times they want during the semester.
Digitally supported learning brings advantages to students in terms of awareness of
the course content, as well as increased collaboration. Therefore, it is imperative that we
use Industry 4.0 technologies in teaching; especially laboratory work as we prepare the
next generation of engineers to work in Industry 4.0 enabled environment. This gives
zamfira@unitbv.ro
A Practical Approach to Teaching Industry 4.0 Technologies 797
them the skills required to work in semi virtualized world that can be realized as in the
following examples: Analyzing a defective machine, monitoring and optimizing energy
consumption of multiple production sites, coming up with a logistics concept for a
virtual factory, or designing a virtual car [14].
In engineering education computer-supported cooperative and collaborative
learning have long been established as methods which support self-driven and
work-related learning processes. Introducing online (Internet) laboratories lifts such
common learning methods to a new level; and in this paper we present laboratory
equipment that we designed to be accessible online through the Internet. The equip-
ment uses Industry 4.0 remote data access technologies, and it is used onsite and
online. Onsite, the equipment is used to teach hardware configuration and integrated,
and online, it is used to teach PLC programming, PLC automation data access, software
applications integration, and Human Machine Interface (HMI) development. This paper
describes the following:
• The laboratory equipment architecture.
• The deployment of the laboratory in one of the course in the process automation
program at McMaster University.
• The feedback we on the educative effectiveness of the laboratory.
The rest of this paper is arranged as follows: Sect. 2 covers work in literature that is
related to online laboratories. In Sect. 3 we present the framework of our Industry 4.0
technologies laboratory equipment, while Sect. 4 deals with the deployment of the
laboratories in a system integration course at McMaster University. Section 5 deals
with the discussions, and the conclusion and future work are covered in Sect. 6.
2 Literature Review
Laboratories place a critical role in the teaching and learning of science based courses.
These laboratories fall under three general categories, namely: hands-on, simulated,
and remote laboratories. Each laboratory categories has strength and weakness, and
there is no consensus in literature on a category that is the most effective [10].
zamfira@unitbv.ro
798 T. Wanyama et al.
some in literature believe that excessive exposure to simulation may result in a dis-
connection between real and virtual worlds [9].
Remote laboratories used real infrastructure that requires space and management,
just like hands-on laboratories. In remote laboratories the experimenter is geographi-
cally detached from the laboratory equipment, as opposed to being collocated with the
equipment as in hands-on laboratories [10]. It is generally agreed in literature that
remote laboratories are increasingly becoming popular because of the following
reasons:
• They can be shared among different institutions, resulting into a shared pool of real
laboratory infrastructure; and hence reducing cost (note that remote laboratory
equipment can be designed to be used in both remote and hands-on modes) [15].
• They have the ability to extend the capabilities of real laboratory equipment,
making it accessible at any time, and from anywhere [2, 3].
• Some studies show that remote laboratories are at least as effective as hands-on
laboratories [3, 11].
Web-based simulated laboratories and remote laboratories can be used support
online teaching and learning. But, while the framework for supporting online course
content delivery is mature, the framework for supporting online laboratories is still
lacking [8]. Therefore, the need to carry out laboratory experiments in engineering
programs is a great challenge to the online component of blended learning. In their
paper titled “A LabVIEW-Based Remote Laboratory Experiments for Control Engi-
neering Education”, Stefanovic et al. state that: “The idea of having a remote
web-based laboratory corresponds to attempt to overcome different constraints and may
be the next step in distance learning” [17].
It is important to note that there are efforts to solve the challenge of integrating
remote laboratories into blended learning [18]. For example, in German, each of the
seven universities that make up the LearnNet network has to provide a remote lab to all
members [12]. In addition, Internet-based remote-access laboratory was developed,
implemented, and piloted at Stevens Institute of Technology in 2005 [4]. In their
implementation, the experimental equipment can be used in the traditional on-site
fashion or it can be accessed remotely through the Internet. Generally, it is possible to
use a combination of available technologies and specific methods to control, configure,
and acquire data from experimental setups over the internet [18].
zamfira@unitbv.ro
A Practical Approach to Teaching Industry 4.0 Technologies 799
Since one of the main focuses of the School of Engineering Practice and Technology at
McMaster University is hands-on learning, the McMaster iLabs (MiLabs) are based on
a framework that supports both onsite and (remote) offsite lab access. Onsite students
are able to configure, program and control laboratory equipment directly, while offsite
students access the laboratory equipment through Internet of Things (IoT) gateways to
remotely program and control it. Currently, the following laboratories are offered by
MiLabs:
• The first set of laboratories focuses on network design, configuration, and wiring.
These laboratories are primarily carried out onsite.
• The second set of laboratories can be offered onsite and offsite. They cover PLC
programming with a focus on programming machine to machine communication
integration of manufacturing and business automation software applications, and
PLC automation systems data access.
zamfira@unitbv.ro
800 T. Wanyama et al.
SmartWire network integrates basic automation devices such as switches, LEDs, and
relays with complex devices such as remote IOs and PLCs. We intend to use a
Smartwire to EtherNet IP gateway to integrate SmartWire with EtherNet IP. This will
enable the moving of process parameters at the basic technologies level of the IEC
automation hierarchy to the enterprise level.
zamfira@unitbv.ro
A Practical Approach to Teaching Industry 4.0 Technologies 801
Fig. 3. Ethernet IP, Modbus serial, and Modbus TCP remotely accessible laboratory setup
• The Human Machine Interface connected to the substation network in Fig. 4 can
read the power parameter from the CLICK micro PLC through the Schneider
TSXETG100 Modbus RTU to Modbus TCP gateway. In this case, each register is
read separately by the HMI. Since HMIs do not support logic instruction, the
registers are read periodically, causing a great amount of traffic on the Modbus RTU
network. This causes the HMI to flag a message timeout error from time to time.
This issue is addressed by using a Productivity 3000 PLC as an IED to read the
CLICK registers through the Modbus gateway, using a Modbus TCP read
instruction.
• The Productivity 3000 PLC can be configured to communicate with the CLICK
micro PLC, the SEL IEC61850 Relay, and the Power meter through the Modbus
serial network using Modbus RTU, or through the Ethernet network using Mod-
bus TCP. In addition, the PLC can communicate with the Eaton ELC-CAENET
remote I/O and the PowerXL DG1 or Powerflex40 VFD through Ethernet IP.
• Students configure and program the network devices using a laboratory computer
that functions as an engineering station shown in Fig. 3. The station has two net-
work cards. One that connects to the process, and the electrical substation networks;
and the other that connects to the plant (university) network.
The laboratory equipment is accessible remotely through Industry 4.0 compliant
eWON Cosy gateways as shown in Fig. 4. The gateway creates a Virtual Private
Network (VPN) via a cloud server called Talk2M to support the laboratories as follows:
zamfira@unitbv.ro
802 T. Wanyama et al.
VPN
Lab University
Equipment Lab Network Network
Machine Network
PLC HMI
VFD
4 Deployment of MiLabs
In the fall of 2015, we offered four related laboratories and a course project using the
MiLabs equipment. The first two laboratories were done in house to enable students to
know each other and develop working relationships. Thereafter the following labora-
tories were offered online:
zamfira@unitbv.ro
A Practical Approach to Teaching Industry 4.0 Technologies 803
zamfira@unitbv.ro
804 T. Wanyama et al.
Fig. 5. Sample student’s HMI solution for MiLabs based course project
zamfira@unitbv.ro
A Practical Approach to Teaching Industry 4.0 Technologies 805
Students were generally excited about doing laboratories and the course project by
accessing laboratory equipment online. Therefore, they gave us a lot of unsolicited
feedback. The following is some of the feedback we obtained that are representative of
the general view of the class about MiLabs:
1. “……Performing labs 3a and 3b remotely was an overall good experience. It
allowed me to focus more on the material as I was alone in a quiet area and could
not become distracted by others. Troubleshooting was also more enjoyable as I had
to rely solely on myself. I believe working remotely and alone would also stop other
students from giving up on a problem easily as they do not have easy and instant
access to the lab instructor and are forced to use critical thinking and problem
solving skills. An area that could be improved is restricting the amount of access
students have to certain tags in order to reduce the risk of shutting down or
damaging the system. Configuring and accessing systems remotely is something
that we will use when working in all industries so being able to experience it first
hand in the lab was valuable to me……”
2. “………. I am writing to you about OPC lab for 4AS3. Since I was a part of the lab
in ETB B111 and I also went home and did the lab again, I have a more informed
perspective of how the entire lab was run. The labs were very straight forward and
were not a struggle for many of the students. Only problems that I experienced were
due to miscommunication and lack of preparation from the other students. Make
sure that the VPNs are set up and ready to go before attempting the lab especially
from home. The learning environment was very good and the chat room that was
set up really helped communication of issues. Having all of the software ready to go
before hand really speeds things up and following the procedure was not difficult.
The project idea is great and I also feel like we should spend more time on OPC
and cover more of its abilities just due to how easy the remote access was from
home. This new type of learning was a very good experience and I thoroughly
enjoyed the labs. These labs should be continued and definitely added to, the remote
access experience was very unique and I learned a lot……..”
3. “………..Referring to labs 3a and 3b, I found that these were well structured. It
allowed the experimenter to get a bit of experience of using VPN to access a
machine and checking parameters, status etc. It also allowed the individual to get
experience creating an HMI to represent the information…….”
4. “……..I know many people now that are connected to their workplace 24/7, being
able to watch systems run and make changes as needed. It is part of our future of
big data, analysis, and optimizations. Next, it gives everyone the opportunity to
work simultaneously on a system, and be able to ask and answer questions to get a
better understanding of what is happening. For students that commute long dis-
tances, it can be a huge benefit. This term it benefited me extremely as I only had the
one lab on Fridays. I didn’t need to commute from Brantford to Hamilton in order
to do the lab. It saves me time, money, and lets me sleep a little longer. Finally, it
gives us a hands-on experience of the types of software we will be using in the field,
and gives us better insight into how communication protocols relate and connect to
each other. This is incredibly helpful. After the third lab I was able to much better
understand the systems in place and I could troubleshoot most problems much more
zamfira@unitbv.ro
806 T. Wanyama et al.
easily and quickly. Even though industry is evolving at a rapid pace, and the
software is changing to accommodate that, updating this lab annually wouldn’t be
a problem as the majority of the software would still support legacy devices.
……..”
Towards the end of the laboratory module, students did a post laboratory test,
followed with a laboratory based project. The students who did the laboratories
associated with MiLabs remotely performed slight better than those who did the lab-
oratories onsite in both the laboratory test and project. This can be attributed to the face
that remotely located students hard far more access time to the equipment than their
counterpart who had to come to the laboratory once a week.
5 Anticipated Outcomes
6 Discussion
From the students’ feedback as well as our own studying of the way MiLabs were
deployed in the course PROCTECH4AS3 – Advanced System Components and
Integration at McMaster University, it is clear that students appreciate online labora-
tories if they are offered within the following structure:
• Run one or two laboratories onsite to start off the class so that the students develop
working relationships and become acquaintances. This improves their collaboration
during online laboratories.
• Hold pre-online laboratory preparation session. During this session the students
should install and test the applications that support remote access to the laboratory
equipment. In addition, they should install and test applications that enable then to
zamfira@unitbv.ro
A Practical Approach to Teaching Industry 4.0 Technologies 807
collaborate in their groups. Finally, test run the major instructions of the laboratory
during this session. For example in our laboratories the main instruction requires
students to record their screen using an application called CamStudio. This
recording provides extra proof of participation in the laboratory session.
• Run a few scheduled synchronous online laboratories, and use a simple confer-
encing mechanism such as a chatroom. This allows students to post questions and
get support from the instructor and colleagues, just as they would in onsite
laboratories.
• It is good to add a small project which students do asynchronously without the
chatroom support. This give them the freedom to try out different thing and learn
from that experience.
There were two main issues that students identified about MiLabs, namely failing to
log on during high access volume, and students having to do extra work to prove that
they logged onto the system. The first issue has been addressed by developing three
new remote access laboratory stations; and the second issue has been address by
making the new laboratory stations capable of recoding user access.
References
1. Bunse, B., Kagermann, H., Wahlster, W.: Industry 4.0: Smart Manufacturing for the Future.
Germany Trade and Invest, Berlin, German, July 2014
2. Canfora, G., Daponte, P., Rapuano, S.: Remote accessible laboratory for electronic
measurement teaching. Comput. Stand. Interfaces 26(6), 489–499 (2004)
zamfira@unitbv.ro
808 T. Wanyama et al.
3. Cooper, M., Donnelly, A., Ferreira, J.M.: Remote controlled experiments for teaching over
the internet: a comparison of approaches developed in the PEARL project. In: Proceedings
of the ASCILITE Conference, Auckland, New Zealand (2002)
4. Del Alamo, J., Nikulin, V.: Engineering laboratory accessible via the Internet. In:
Proceedings of the 2006 ASEE Annual Conference and Exposition, Session 1526, Chicago,
USA, 18–21 June 2006
5. eWON: eWON Cosy 131 Installation Manual. https://ewon.biz/sites/default/files/ig-022-0-
en-ewon_cosy131.pdf. Accessed 15 May 2016
6. Farrington, P.A., Messeimer, S.L., Schroer, B.J.: Simulation and undergraduate engineering
education: the Technology Reinvestment Project (TRP). In: Proceedings of the 1994 Winter
Simulation Conference, Lake Buena Vista, Florida, USA (1994)
7. Glazer, F.S.: Blended Learning: Across the Disciplines, Across the Academy, Stylus
Publishing, LLC (2011)
8. Koller, D., Ng, A.: Coursera makes top college courses free online, Phys.org. http://phys.org/
news/2012-07-coursera-college-courses-free-online.html. Accessed 26 Feb 2014
9. Magin, D.J., Kanapathipillai, S.: Engineering students’ understanding of the role of
experiments. Eur. J. Eng. Educ. 25(4), 351–358 (2000). 2
10. Nickerson, J. V., Hands-on, simulated, and remote laboratories. ACM Comput. Surv. 38(3),
(2006). Article 7
11. Scanlon, E., Colwel, C., Cooper, M., Paola, T.D.: Remote experiments, re-versioning and
re-thinking science learning. Comput. Educ. 43(1–2), 153–163 (2004)
12. Esche, S.K.S., Prasad, M.G., Chassapis, C.: A remotely accessible laboratory approach to
undergraduate education. In: Proceedings of the 2004 ASEE Annual Conference and
Exposition, Session 3220, Salt Lake City, UT, United States
13. Schuh, G., Reuter, C., Hauptvogel, A., Dölle, C.: Hypotheses for a theory of production in
the context of industrie 4.0. advances in production technology. Lect. Notes Prod. Eng. 11
(2015)
14. Schuster, K., Groß, K., Vossen, R., Richert A., Jeschke, S.: Preparing for industry 4.0 –
collaborative virtual learning environments in engineering education. In: The International
Conference on E-Learning in the Workplace, New York, NY, USA, 10th–12th June 2015
15. Sell, R.: Remote laboratory portal for robotic and embedded system experiments. Int.
J. Online Eng. 9(8), pp. 23–26 (2013)
16. Singh, I., Al-Mutawaly, N., Wanyama, T.: Teaching network technologies that support
industry 4.0. In: Proceedings of the Canadian Engineering Education Association
Conference, Hamilton – Canada, June 2015
17. Stefanovic, M., Cvijetkovic, V., Matijevic, M., Simic, V.: A labview-based remote
laboratory experiments for control engineering education. Comput. Appl. Eng. Educ. 19(3),
538–549 (2009)
18. Tumkor, S., Esche, S.K., Chassapis, C.: Design of remote laboratory experiments using
LabVIEW web services. In: Proceedings of the International Mechanical Engineering
Congress and Exposition, Education and Globalization; General Topics, vol. 5 (2012)
19. Wanyama, T.: A practical approach to industrial systems integration, industry 4.0 and
industrial internet of things: case of manufacturing, energy, building, environment and
business data integration using ethernet and OPC technologies, Hamilton, Ontario-Canada,
September (2016)
20. Wanyama, T., Singh, I.: A training demonstration for experiential learning in OPC based
process automation data access. In: Proceedings of the Canadian Engineering Education
Association Conference, June 2013, Montreal, QC, Canada
zamfira@unitbv.ro
Design of WEB Laboratory for Programming
and Use of an FPGA Device
1 Introduction
Web-based remote laboratories are consisted of physical laboratory model and meas‐
urement and control apparatus that could be remotely controlled through Internet
connection via Internet browser. Brief history of web-based remote laboratories is given
in [1], and positive influences of web-based remote laboratories are described in [2].
Developing countries could benefit from web-based remote laboratories by increasing
teaching capacities without downgrading quality of lectures [3]. Some other benefits are
opportunity for lecturers to demonstrate real physical experiments during lecture [4],
higher availability of laboratory capacities to students in contrast to the fixed terms,
higher availability of laboratory models to students with special needs, and availability
zamfira@unitbv.ro
810 N. Jović and M. Matijević
of laboratory models to teachers and students on those faculties that do not have neces‐
sary equipment to perform experiments. Safety of students is greatly improved during
execution of risky experiments (high voltage, temperature, chemicals, etc.). Cost of
equipment to number of users’ ratio is well favored [1–5]. Conditions that web-based
remote laboratory should satisfy are:
• Availability of laboratory model 24/7
• Client side should be written using HTML5 and JavaScript, without third-party soft‐
ware (Adobe Flash, Java)
• Minimal administration (reservation-free, access is granted to the first connected
client)
• Maximal security (safety of hardware, difficult to hack)
• System should return to the initial state after execution of experiment
• Experiment duration should be short (a few minutes)
Outcomes of this work are:
• Enabling web-based laboratory lectures for programming and control of an FPGA
• Providing environment for testing and execution of student’s design on real hardware
via Internet connection
• Easy to use graphical interface
• Contributing to development of lecture material in Serbian language based on MIT
6.111 “Introductory Digital Systems Laboratory” course
• Contributing to the general software solution for web-based remote laboratories which
is in development stage within Faculty of Engineering, University of Kragujevac
First version of the web-based remote laboratory software was tested in Faculty of
Engineering, University of Kragujevac during 2015/2016 winter semester within the
course “Measurement and Control”. A hundred and ninety students were able to
complete four laboratory exercises via Internet connection on 2–4 physical models
available in the same time [6].
Use of a web-based remote laboratory helped in solving several problems:
• Limited resources regarding space, teaching staff and laboratory equipment
• Lectures were done with respect to the existing resources, but students were provided
with access to the laboratory resources via Internet connection anytime they want
• Students had an opportunity to do individual laboratory exercises despite “20 students
per group” limitation imposed for given level of study (BSc) within University of
Kragujevac
Software solution for a web-based remote laboratory possess an aggregator, which
enables adding unlimited number of new laboratory models into the existing content of
a web-based remote laboratory [7]. Similar solution for programming and control of an
FPGA is given in [8], so idea is not new. Several drawbacks in solution given in [8] are
expensive equipment and windows-only server side. One of the aims of this work is
overcoming these problems by making lightweight and portable server side that could
work on broad spectrum of devices (PC, RaspberryPi, BeagleBone, etc.). Laboratory
zamfira@unitbv.ro
Design of WEB Laboratory for Programming 811
models used in this work are based on Digilent Nexys2-500k FPGA development plat‐
form, which are chosen because of small price and lots of digital inputs and outputs.
Fig. 1. Structural diagram of web-based laboratory model. (1) Client (2) Internet (3) Server
(4) Arduino Leonardo for controlling digital inputs of the FPGA development board (5) USB
JTAG programmer for FPGA development board (6) FPGA development board (7) LCD
screen connected to the FPGA (8) Web cameras for real-time video streaming
Fig. 2. Laboratory model for programming and control of an FPGA device. (1) FPGA
development board, (2) Resistor-based voltage divider circuit (3) Arduino Leonardo (4) USB
JTAG programmer (5) VGA output from FPGA board
zamfira@unitbv.ro
812 N. Jović and M. Matijević
real-time video stream. Once programmed, user can control up to 12 inputs to the FPGA
development platform. Inputs to the FPGA development platform are generated with
Arduino Leonardo development board and/or Raspberry Pi single board computer
(Fig. 1).
Laboratory model incorporates Arduino Leonardo development platform with a
resistor-based voltage dividers for controlling inputs of the FPGA development platform
(Fig. 2).
There are two web cameras for monitoring outputs of the FPGA development plat‐
form and recording images from the LCD screen connected to the FPGA development
board.
Features of the used FPGA development platform are [9]:
• Xilinx Spartan 3E FPGA circuit (Fig. 3, 1)
• Built-in USB JTAG programmer (Fig. 3, 2)
• 16 MB of Micron PSDRAM and 16 MB of Intel StrataFlash ROM memory (Fig. 3, 3)
• Xilinx Platform Flash for non-volatile configuration storage (Fig. 3, 4)
• 50 MHz cyrstal oscillator (Fig. 3, 5)
• 60 general purpose input/output pins (3.3 V) (Fig. 3, 6)
• Eight LEDs (Fig. 3, 7), four eight-segment displays (Fig. 3, 8)
• Four pushbuttons (Fig. 3, 9) and eight switches (Fig. 3, 10)
• PS/2 (Fig. 3, 11), VGA (Fig. 3, 12) and RS-232 connectors (Fig. 3, 13)
• Power over USB, external adapter or battery (Fig. 3, 14)
In order to control physical inputs of the FPGA development board over Internet
connection, it is necessary to make an interface from FPGA board to the server which
zamfira@unitbv.ro
Design of WEB Laboratory for Programming 813
will pass digital signal to the inputs of the FPGA development board. For passing a
course successfully, twelve digital inputs are sufficient. This is achieved by sending a
digital signal from Arduino Leonardo to the Digilent Nexys2 development board.
Because of logic level mismatch between Arduino Leonardo (5 V) and Digilent Nexys2
FPGA development board (3.3 V), precautions must be taken in order to prevent damage
to the FPGA development board. This is achieved using resistor based voltage divider
circuit (Fig. 4).
Fig. 4. Resistor based voltage divider circuit for interfacing Arduino Leonardo and Digilent
Nexys2 FPGA development board
Digilent Nexys2 FPGA development platform has four PMod connectors for
connecting various peripherals to it. Each of the four PMod connectors has eight general
purpose input/output pins. Those connectors are used for connection with Arduino
Fig. 5. Schematic of the connection between Arduino Leonardo and Digilent Nexys2
zamfira@unitbv.ro
814 N. Jović and M. Matijević
Leonardo development board. Digital outputs from Arduino Leonardo (D1–D12) are
connected to the digital inputs on PMod connectors of the Digilent Nexys2 FPGA
development board (JA1–JA8 on first PMod connector, and JB1–JB4 on the second one).
Schematic is given in Fig. 5.
This configuration has its own limitations in terms that user is in control of only
twelve inputs to the FPGA development platform, but number of available inputs is
sufficient to successfully pass laboratory exercises based on material given in [10].
zamfira@unitbv.ro
Design of WEB Laboratory for Programming 815
and embeds it into its own page. Diagram of the whole web-based remote laboratory
architecture is presented in Fig. 6.
zamfira@unitbv.ro
816 N. Jović and M. Matijević
Fig. 7. Index page of the aggregator web page. Laboratory models are shown on pictures with
corresponding descriptions.
zamfira@unitbv.ro
Design of WEB Laboratory for Programming 817
• labList Shows all available laboratory models within web-based remote laboratory,
default view when user access central server. Thumbnail and short description is
shown for each laboratory model
• navbar Navigation panel
• webLabView Web page of laboratory model. This is the page where user can conduct
an experiment. It uses iFrame to embed content from micro server.
Each view has its own controller with all logic contained within, achieving Model-View-
Controller philosophy. Index page of the aggregator is shown in Fig. 7.
Fig. 8. Graphical interface of the client side for the remote experimentation on the FPGA
development board. (1) Real-time video stream of the FPGA development platform (2) Real-time
video stream of the LCD monitor connected to the FPGA development platform (3) Field for
uploading user design (4) Eight switches (5) Four pushbuttons
Web-based remote laboratory is concepted in a way that user can upload its own
program for controlling a laboratory model, monitoring outputs of the model in real-
time. That program should not invoke any harmful effect on the laboratory model. For
that reason, as well as abstracting low level implementation from the user, program is
scanned for any security threat, and then it is embedded into a larger main program
which will call subroutines from user program. Real time control is achieved with the
help of the Socket.IO communication library. In this particular case, instead of program
for controlling of the laboratory model, user would upload compiled Verilog design in
zamfira@unitbv.ro
818 N. Jović and M. Matijević
the form of the bitstream file. Once the process of upload is finished, server will start
Digilent Adept program which will download user’s design on the FPGA development
board. After that, user can control the FPGA development board with four pushbuttons
and eight switches available on the web page of the laboratory model, and monitor output
through real-time video streams. As previously described, interface between micro
server and FPGA development board is Arduino Leonardo. It is programmed with
Firmata protocol, which will communicate with Node.JS server through JohnnyFive
library. This library enables control of Arduino’s GPIO ports over USB from PC [11].
Micro server is developed as finite state machine, where triggers for states are made with
EventEmitter library for Node.JS. State transition diagram is shown on Fig. 9.
Fig. 9. State transition diagram of the web-based remote laboratory micro server
One of the aims of this work is enabling realization of the teaching methodology seen
in [10] in the case of users who don’t have access to the organized laboratory resources
zamfira@unitbv.ro
Design of WEB Laboratory for Programming 819
5 Conclusions
This work presents a web-based remote laboratory for programming and control of an
FPGA device designed for use within the course “Introduction to Design and Control
of Integrated Circuits for Communication, Sensors and Actuators” in Faculty of Engi‐
neering, University of Kragujevac. Guided by the prior experiences in web-based remote
laboratories for the course “Measurement and Control” on the same faculty, expected
results are:
• Deeper understanding on topic of programming an FPGA device by student
• Smaller financial overhead regarding laboratory equipment, compared to traditional
laboratory exercises
• Easier tracking of student’s progress during the course
• Availability of laboratory equipment to student 24/7
zamfira@unitbv.ro
820 N. Jović and M. Matijević
Acknowledgment. Work on this paper was partly funded by the SCOPES project
IZ74Z0_160454/1 “Enabling Web-based Remote Laboratory Community and Infrastructure” of
Swiss National Science Foundation.
References
1. Heradio, R., de la Torre, L., Galan, D., Cabrerizo, F.J., Herrara-Viedma, E., Dormido, S.:
Virtual and remote labs in education: a bibliometric analysis. Comput. Educ. 97, 14–38 (2016)
2. Farrokhnia, M.R., Esmailpour, A.: A study on the impact of real, virtual and comprehensive
experimenting on students’ conceptual understanding of DC electric circuits and their skills
in undergraduate electricity laboratory. Procedia Soc. Behav. Sci. 2(2), 5474–5482 (2010)
3. Garcia-Guzman, J., Villa-L-pez, F.H., Silva-Del-Rosario, F.H., Ramirez-Ramirez, A.,
Enriquez, J.V., Alvarez-Sanchez, E.J.: Virtual environment for remote access and automation
of an AC motor in a web-based laboratory. Procedia Technol. 3, 224–234 (2012)
4. Abdulwahed, M., Nagy, Z.K.: Systematic evaluation of the use of remote and virtual
laboratories in engineering education. In: 21st European Symposium on Computer Aided
Process engineering – ESCAPE, vol. 21 (2011)
5. Matijević, M.S., Cvjetković, V.M., Filipović, V.Ž., Jović, N.D.: Basic concepts of automation
and mechatronics with LEGO mindstorms NXT. Tehnika 69(4), 653–660 (2014)
6. Mitrović, R., Jović, N., Cvjetković, V., Nedeljković, M., Matijević, M.: Internet mediated
laboratories in engineering education. In: Proceedings of XXII Conference on Development
Trends: “New Technologies in Teaching”, TREND 2016 (2016)
7. Jović, N.: Design of web-based remote laboratory for teaching purposes of programming and
use of an FPGA device. Master’s thesis, Faculty of Engineering, University of Kragujevac,
October 2016
8. Tsai, J.: Design and implementation of an online laboratory for introductory digital systems.
Master’s thesis, Massachusetts Institute of Technology, August 2005
zamfira@unitbv.ro
Design of WEB Laboratory for Programming 821
zamfira@unitbv.ro
Remote Triggered Software Defined Radio
Using GNU Radio
1
Remotelabs.in, Pune, India
jasveer@remotelabs.in, pavs94@gmail.com
2
Department of Computer Engineering,
Sinhgad Institute of Technology and Science, Pune, India
kunalbidkar13@gmail.com
1 Introduction
zamfira@unitbv.ro
Remote Triggered Software Defined Radio Using GNU Radio 823
2 Remote Laboratories
2.1 Definition
Remote Laboratories are those laboratories that provide access to any hardware present
at a physical location such that the laboratory can be operated at any location different
from that of the physical location of the device [1]. This means that the user need not
be present physically to perform the experiments on the hardware.
2.2 Advantages
The architecture, as given in Fig. 1, provides details about the Remote Triggered Soft‐
ware Defined Radio System.
• The Laboratory Computer is installed with the VNC Server. This computer has GNU
Radio Software which is the software accessed by the remote user.
• The Cloud Server is configured with the Apache Tomcat and Guacamole is installed
onto the Tomcat Environment.
• The front-end web application is the interface for the end user to access the VNC
connection.
• The client simply has to access the web application from the HTML5 browser to
view the contents of the VNC Session.
• The client has the ability to upload his Data Flow Graph from the web interface.
• The client has to ability to run real time experiments on the Remote System.
zamfira@unitbv.ro
824 J.S.T. Jethra et al.
4 How It Works
There are mainly 3 components that need to be configured to ensure that the remote
connection to the apparatus is successful. They are:
• To configure the Laboratory Server with VNC.
• To configure the Cloud Server.
zamfira@unitbv.ro
Remote Triggered Software Defined Radio Using GNU Radio 825
• To configure the Client Application which connects to the Laboratory Server via a
Web Browser.
• The configuration file on VNC server helps deduce it show application specific details
to the end user, for example; if the user wants to display the gnome desktop in Ubuntu,
the configuration file contains the runtime executable of gnome such as gnome-
session, if the user wants to start an application at the connection setup, VNC server
zamfira@unitbv.ro
826 J.S.T. Jethra et al.
can specify the application in the configuration file which helps starting that
application at the connection setup.
This laboratory computer has the application for Software Defined Radio – GNU Radio,
which is to be access by the end user with the help of VNC.
zamfira@unitbv.ro
Remote Triggered Software Defined Radio Using GNU Radio 827
The cloud server must be of a good configuration to handle the requests of the VNC
connection to serve both the laboratory server as well as the client system.
Guacamole, does not in itself configure or spawn VNC connections to the laboratory
computer. The connection parameters must be setup by an administrator who manages
connections to the VNC server.
A new connection is created in Guacamole which has the following parameters [5]:
• Hostname/IP Address of the Laboratory Server.
• Port Number for VNC; starts with 5901 and continues until different VNC displays
are created to which VNC server is listening on, usually 5900 or 5900 + display
number. For example, if the VNC server is serving display number 2, the port number
would be 5902.
zamfira@unitbv.ro
828 J.S.T. Jethra et al.
• Autoretry parameter is set to give the number of attempts made to connect automat‐
ically before giving an error.
VNC sessions also require the following to authenticate the sessions:
• Username of the laboratory remote computer.
• Password of the laboratory remote computer.
The client cannot request any display size from the VNC Server and hence it depends
on the server to provide a suitable screen size. Although, if there is less bandwidth on
the client system, the server can reduce the color depth. In case of Guacamole, it can
automatically detect 256-color images. The following configurations must be set to
ensure suitable display settings:
• Color depth is used to specify the color in bits per pixel. This parameter is optional,
and the color depth is present in 8, 16, 24 and 32 bits.
• Cursor is a pointer on the VNC display, if set to a remote cursor, the mouse pointer
will be rendered remotely. Although, the motion of the remote cursor will be slow
as it will depend on the bandwidth present on the client system.
• Read only is to ensure if the connection is read only, no input will be accepted from
the user, and the user will be able to see whatever the other users who are using the
same desktop doing. This parameter is optional as well.
In order to ensure, that the user can upload his files on the remote system and execute
his program, Secure File Transfer Protocol (SFTP) is used to configure the connection,
so that the user has the ability to upload on the laboratory system on a VNC connection.
SFTP also has certain parameters that need to be setup to ensure that the system is capable
of upload functionality.
They are as follows:
• sftp-hostname is the hostname or IP address of the server that is hosting the SFTP
server.
• Port number is 22 which is the default port number for File Transfer Protocol (FTP).
• sftp-username which is required to authenticate the connection to the server.
• sftp-password is also required for authentication.
• sftp-directory is the default directory location that will be present on the laboratory
server where the files will be uploaded.
Since VNC is really versatile there exist a lot of options that can be set based on the
functionality of the system. In case of Remote Software Defined Radio, these parameters
are sufficient to ensure that the system works as expected.
zamfira@unitbv.ro
Remote Triggered Software Defined Radio Using GNU Radio 829
together form different software defined radio elements and signal processing opera‐
tions. GNU Radio can be connected to a RF hardware to run experiments in real time.
The external hardware that is provided has a wide range from 400 MHz to 4 GHz. The
hardware is able to configure itself based on the data flow graphs provided by the soft‐
ware – GNU Radio, and hence the name Software Defined Radio. Internal hardware
components need not be changed or reconfigured because the complete controls of the
components are done with the software (Fig. 4).
This allows the construction of radios where the actual waveform is transmitted and
received are defined by software. Since SDR’s require a lot of Digital Signal Processing,
the computer system executing the application must be of high configuration.
GNU Radio blocks are written in Python and it follows a graphical user interface
approach to use these signal blocks. Hence SDR used with GNU Radio is a great option
for writing any kind of Digital Signal Processing applications.
The laboratory system, that is present to be accessed is a high configuration modern
system, capable of handling any configurations of the SDR, this system is preloaded
with the GNU Radio Companion software.
The laboratory system is the access point for Remote SDR.
5 Results
Remote Triggered Software Defined Radio helps connect the hardware to the web
application and access via a domain. The results of the system is as follows:
zamfira@unitbv.ro
830 J.S.T. Jethra et al.
Figure 5 shows a GSM Receiver program run on a Remote Software Defined Radio
System. The Output is again displayed on the browser, enabling anytime, anywhere
access to the SDR system.
Fig. 5. Accessing remote SDR from web browser to show GSM receiver.
6 Conclusion
Given the context of the Software Defined Radio Systems and the cost that goes in
implementing such a system, the connection of the SDR Hardware on a remote system,
is highly beneficial in providing an on-demand access to the system.
References
1. Jethra, J.S.T., Patkar, S.B., Datta, S.: Remote triggered FPGA based automated system. IEEE,
February 2014
2. Sareen, P.: Cloud computing: types, architecture, applications, concerns, virtualization and
role of IT governance in cloud. Int. J. Adv. Res. Comput. Sci. Softw. Eng. 3(3), March 2013
3. tej Koganti, K., Patnala, E., Narasingu, S.S., Chaitanya, J.N.: Virtualization technology in
cloud computing environment. Int. J. Emerg. Technol. Adv. Eng. 3(3), March 2013
4. Selva, A.F.B., Reis, A.L.G., Lenzi, K.G., Meloni, L.G.P., Barbin, S.E.: Introduction to the
software-defined radio approach. IEEE Lat. Am. Trans. 10(1), January 2012
5. Configuring Guacamole Protocol Web Documentation. https://
guacamole.incubator.apache.org/doc/gug/configuring-guacamole.html
6. The guacd server and Guacamole Protocol Web Documentation. https://
guacamole.incubator.apache.org/doc/gug/guacamole-architecture.html#guacamole-protocol-
architecture
zamfira@unitbv.ro
Open Educational Resources
zamfira@unitbv.ro
MOOC in a School Environment: ODL Project
1
Deusto Foundation, Bilbao, Spain
olga.dziabenko@deusto.es
2
Ellinogermaniki Agogi, Pallini, Greece
eleftheria@ea.gr
Abstract. Unlike schools 15 years ago, contemporary schools use many ICT
tools in their classes, e.g. computers, tablets, smartphones, etc. accompanying by
open educational software, OER and apps. Teachers gradually turned to more
student-centred approaches like inquiry, game-based, project-building, flipped
learning, learning-by-teaching to name only but a few. Personalized teaching and
learning supported by all these approaches help schools offer more effective and
efficient education. Although Massive Open Online Courses (MOOCs) have
proved to be helpful in university and adult education, until now they has not been
yet deployed in school education. The ‘Open Discovery of STEM Laboratories’
project (ODL) exploits this potential and opens up MOOCs for it. In this paper,
we discuss the first results of this implementation.
1 Introduction
The fast-changing global economy acts like an engine that generates the demands of the
skills that school, college and university graduates are expected to have in order to be
competitive and have a capacity to drive innovation. Therefore, the future prosperity
and social stability depend on the optimal use of our human capital. About 70 million
Europeans [1] lack sufficient reading, writing and numeracy skills, and 40% of the EU
population lack a sufficient level of digital skills. This potentially is one of the main
sources of unemployment, poverty and social exclusion. On the other hand, 40% of
European employers have difficulty in finding people with the right skills to foster
growth and innovation [2]. At the same time, numerically high-qualified young people
work in job positions that do not match their talents and knowledge.
Based on the research “The Survey of Adult Skills”, in 2016 the European Commis‐
sion has adopted a new and comprehensive skills Agenda for Europe to improve the
teaching and recognition of skills - from basic to higher skills, as well as transversal and
civic skills - and ultimately to boost employability. As we see, the digital literacy, and
therefore, the education in applied sciences, engineering, and technologies is one of the
keys to contribute to the European Commission’s first political priority, “A New Boost
for Jobs, Growth and Investment”. It is a responsibility of all education players - schools,
zamfira@unitbv.ro
834 O. Dziabenko and E. Tsourlidaki
universities and policy-makers to ensure that no-one is left behind and that Europe
nurtures the high-end skills that drive innovation and competitiveness.
The Open Discovery of STEM Laboratories (ODL) project [3] was created in order
to introduce the use of MOOCs in school curricula and in conjunction with the STEM
laboratories available online. ODL offers to school teachers a methodology for building
micro-MOOCs for their students. Exploring the MOOC idea in a school context, the
consortium determined that based on the content-load, activity and time consuming, it
would be beneficial to chunk the MOOC course on few small learning lessons – micro-
MOOCs (activity for 20–40 min in the classroom). Usually, course consists of several
lectures. Each lecture in format of micro-MOOC includes laboratory work, theoretical
and practical content, assessment and discussion. The suggested structure allows to
easily embed micro-MOOCs in classroom time and, furthermore, to reduce a time
required for creating learning materials.
In this paper we present the first outputs of our project. The main aim is to introduce
the benefits and lessons learned through the integration of micro-MOOCs - an innovative
approach for deploying STEM labs into a school. During the project, the team has created
multidisciplinary MOOCs and trains teachers to design and implement the MOOC
approach in their school. In this paper we discuss the format of MOOCs proposed for
application in schools, how to introduce online labs in a MOOC environment and how
to organize the individual and collaborative learning using this instrument. In order to
create our MOOCs the ODL project partners use the edX open platform [4] where a
MOOC space area was created for the project.
In Sect. 2 we describe the ODL project as it is – partners, aims and objectives as well
as its structure. Section 3 is devoted to illustrate the inquiry-based learning approach
used for introducing online laboratories. An example micro-MOOC is presented in
Sect. 4. Section 5 summarizes our conclusions and introduces possible future work.
The ODL project aims to foster teacher collaboration in creating innovative STEM
school curricula by open discovery of remote and virtual laboratories and their appli‐
cation in education. The consortium offers schools a micro-MOOC methodology for
transforming separate education materials into coherent lessons. Micro-MOOCs
preserve the principles of open teacher collaboration in STEM curriculum development.
It is planned that teachers will work together on creating micro-MOOCs that will be
united under the theme of one MOOC. In this case diverse national practices will be
applied.
For this purposes the project proposes a MOOC methodology to be used on different
subjects of school curricula and it offers an MOOC platform designed to meet teachers’
needs. The project aspires to train at least 300 school teachers to develop micro-MOOCs
for STEM lessons. By end of the project at least fifty-five micro-MOOCs that include
the use of remote and virtual laboratories will be available on the platform. Students and
teachers from EU school communities will have access openly to all the learning mate‐
rials.
zamfira@unitbv.ro
MOOC in a School Environment: ODL Project 835
The methodology proposed will help educators find and organize digital learning
resources while designing and delivering personalized instructions in a school environ‐
ment. Our training will facilitate the teachers to find and evaluate content; collect and
organize OERs, remote and virtual laboratories according the curricula; build STEM
micro-MOOCs, manage lesson plans, content, student activities - laboratory and practice
work; engage students through the students-centered learning and personalized feed‐
back.
The project focuses on teachers, curriculum designers and administrators strength‐
ening their profile by supporting them to deliver high quality teaching practices and to
adopt new methods and tools. In particular, the project will extend teachers’ knowledge
and skills and support new teachers so that they have all necessary competences right
from the start.
The project is designed to enhance digitalization of learning, teaching, and training
by improving accessibility to high quality learning through micro-MOOC, use of OER,
and teacher and school collaboration in modernization of STEM school curricula.
3 Inquiry-Based Learning
zamfira@unitbv.ro
836 O. Dziabenko and E. Tsourlidaki
On the ‘Engage’ stage the teachers aim to capture the students’ imagination and moti‐
vation. Here students get the first introduction to the topic and understand the learning
environment and tools that are used to build the inquiry curiosity. The ‘Explore’ stage
allows to develop students’ critical thinking and to help them explore new things on the
subjects at hand. The ‘Explain’ stage requires from students to explain the involved
phenomena using scientifically correct arguments. At this stage students start to create
a model, discuss the data collected with their peers and the teacher and begin to commu‐
nicate what they have learned. ‘Extend’ is the stage in which students expand their
knowledge on the concept(s) they have studied, make connections to other related
concepts, and apply their understanding to the real world. Finally, through discussion
and disputes students make analyses and evaluate the knowledge they acquired during
the activity.
Depending on the teachers’ and/or students’ needs three scenarios or pedagogical
frameworks are suggested in the project:
• Traditional approach or Confirmation inquiry [9]
• Structured or guided inquiry approach [9]
• Elicited or Open inquiry approach [10].
zamfira@unitbv.ro
MOOC in a School Environment: ODL Project 837
The key outcome of the ODL project is a collection of micro-MOOCs available for
school teachers. These micro-MOOCs will support the engagement of schools in inno‐
vative approaches of blended education in the everyday practices. The Light Pollution
micro-MOOC [13] is one of the micro-MOOCs of our collection.
Every micro-MOOC begins by providing information on the activity (see Fig. 1).
For example, “Light pollution is a global problem that affects us all. In this micro-
MOOC you will have the opportunity to learn more about light pollution and its impact
on the planet”. In addition, requirements that allow the efficient performance of the tasks
in the micro-MOOC are presented. The traditional approach with 5E structure is used
to build this inquiry-based scenario (see Fig. 2).
At the introduction stage the research questions and tasks pick the interest of the
students to the light pollution problem in the cities. The introduction of videos and
animations helps to highlight the problem briefly and in an attractive visual way. Several
video presentations, discussions, multiple-choice self-assessments are the main tools
that keep students involved. In this scenario the STEM online labs used is an interactive
map of sky glow and light pollution simulator which assist students to explore the
phenomenon.
This example introduces the basic requirements of the well-designed micro-MOOC,
namely:
– affective engagement of the students;
– harmonize learning process for students with different knowledge and interest;
– generating curiosity and leading to questions;
– a cognitive conflict;
– scientific investigation and explanation within the competence of the students
involved;
– creating scientific knowledge;
zamfira@unitbv.ro
838 O. Dziabenko and E. Tsourlidaki
– requiring the students to use inquiry skills to explain the involved phenomena;
– limiting time of use (1–2 lessons for the presentation and applying of remote/virtual
labs).
5 Conclusion
Although the ODL project is still at the beginning stage, it is clear that teachers are
interested in such approach. They see micro-MOOC as a tool to open a horizon of STEM
subjects to their students; to embed the use online labs in the framework of their curric‐
ulum, which they never could use otherwise and to broad their collaboration with the
colleagues on multidisciplinary aspects. In this paper we presented the preliminary
outputs obtained from the implementation of the micro-MOOC approach in secondary
school classes. Unlike traditional MOOCs that could last several months, the ODL
micro-MOOCs are adjusted so as to meet the needs of in-class activities and last from
20 min to a few of class hours. The inquiry-based cycle based on the 5E stages
(ENGAGE, EXPLORE, EXPLAIN, EXTEND, EVALUATE) was introduced. The
reader can try the micro-MOOCs available on the ODL portal [12]. One of them is Light
Pollution that explains the influence of the light on the ecosystem and humans, and risks
caused by this influence. By gaining and understanding of light pollution students could
be encouraged to search for solutions to reduce the negative impact light pollution on
the environment.
In the near future, the project plans to organize a set of workshops which will offer
a discussion for designing new learning materials and will give valuable feedback on
the impact of the proposed methodology to school education. The consortium is planning
to create at least 50 micro-MOOCs offered to schools. The study targets to evaluate the
impact of the proposed methodology on students’ knowledge and increase of interest
towards STEM. The analytics system incorporated in edX platform will provide us with
the necessary independent data and results.
The results of our trials will be published on the project website (http://opendisco‐
verylabs.eu/) and on the facebook group discussion wall. The study results could be
helpful for secondary school sector representatives, education instructors, parents and
policy makers to respond to current and future education needs.
Acknowledgement. This work was partially funded by the European Union in the context of
the ODL project (Project Number: 2015-1-ES01-KA201-016090) under the ERASMUS+
programme. This paper does not represent the opinion of the European Union, and the European
Union is not responsible for any use that might be made of its content.
We want to thank all ODL partners who contributed to the discussion of the ideas of MOOC
and an inquiry-based learning at secondary school sector to support the work is performed.
References
1. The Survey of Adult Skills (PIAAC): Implications for education and training policies in
Europe. European Commission (2013)
zamfira@unitbv.ro
MOOC in a School Environment: ODL Project 839
zamfira@unitbv.ro
Survey and Analysis of the Application of Massive
Open Online Courses (MOOCs) in the Engineering
Education in China
Based on a Survey of XuetangX, the World’s Largest
MOOC Platform in the Chinese Language
1 Introduction
The rapid development of massive open online courses (MOOCs) has broadened the
application scope of online education and transformed classroom teaching from
“teaching centeredness” to “learning centeredness”. This shift offers a new way for
learners to acquire knowledge. At the same time, according to the research by many
researchers, integrating MOOCs in blended courses contributes to improved motivation
and school performance of students, especially to the cultivation of engineering capacity
of engineering students. At present, many Chinese colleges and universities have
reformed or are reforming their teaching model through the use of MOOCs. During the
cultivation of engineering talents, how well are MOOCs applied and, are there any
difficulties or puzzles encountered in this process?
Based on a survey and analysis of XuetangX (xuetangx.com), China’s largest MOOC
platform, this paper aims to introduce to domestic and foreign scholars the application
of MOOCs in the engineering education in China. In the meantime, based on the relevant
zamfira@unitbv.ro
Survey and Analysis of the Application of MOOCs 841
survey, this paper analyzes the problems encountered in the popularization and appli‐
cation of MOOCs in China. It also gives some suggestions on the further development
and application of MOOCs in the engineering education in China.
As a new learning and teaching method, MOOCs, by virtue of its large scale and open‐
ness, is impacting the global education landscape. The emergence of MOOC makes it
easier to share educational resources, realize Educational informationalization and to
pursue self-directed learning and lifelong learning. Since 2013, MOOCs have mush‐
roomed in China. The wide application of MOOCs means that it stands to create an
entirely new, and fairer educational model. Teaching and learning can take place
anywhere at any time, by dint of the Internet-enabled course and lecture video, embedded
course test and assessment, and student-teacher interactions provided by MOOC. This
is a very case where educational resources are perfectly shared.
In addition, educational informatization ranks among China’s national strategic
layouts. In his letter of congratulation to the International Congress on ICT in Education
(held on May 23, 2015), Chinese president Xi Jinping said that China would advance
educational reforms and innovations, structure a network-based, digital, personalized,
and lifelong educational system, and build China into a learning society in which
everyone learns anywhere at any time [1]. While presiding an executive meeting of the
State Council on November 15, 2014, Chinese premier Li Keqiang stressed that China
would foster new business formats and new industries, among which online education
is prioritized. The Chinese Ministry of Education (MOE) is also boosting MOOCs in
terms of policy, fund, and platform construction, etc. [2].
A large number of scholars have made empirical studies on the relationship
between online learning behavior and learning effect in different environments, and
it is found that the online behavior of learners has an important influence on the
learning effect [3–7]. As shown by the result of the relevant survey [8], the blended
learning integrating MOOC has not only motivated students, but more importantly
has improved students’ ability to learn independently. According to the study, most
students believe that the blended learning offered by MOOC gives them a lot of
freedom as they can flexibly arrange their learning time and develop their own effi‐
cient learning plan. Many students think that the discussion group of MOOC makes
it possible for them to discuss questions in learning with learners from around the
globe so that they can experience the collision of ideas under different thinking
models. As such, the study holds that, the MOOC-based blended teaching model
plays a certain role in improving the teaching quality of higher education, and in
cultivating students’ capacity in collaboration and self-directed learning.
zamfira@unitbv.ro
842 Y. Long et al.
This part is based on a data survey of XuetangX, the world’s largest MOOC platform
in the Chinese language. It aims to provide a look at the application of MOOCs in the
engineering education in China.
On October 10, 2013, Tsinghua University launched XuetangX, the first-ever MOOC
platform in the Chinese language. Following the foundation of the MOE Research
Center for Online Education in April 2014, XuetangX became the platform for research
exchange and result application of the Center. After three years’ efforts, XuetangX has
emerged as one of MOOC platforms in the Chinese language with the largest course
scales and most active learners. By August 31, 2016, XuetangX had recorded 6.39
million of online sign-ups for the courses, and 4.21 million of registered users coming
from 137 countries and regions across the world. As of that time, XuetangX had opened
875 courses, including 250 from domestic colleges, universities, and institutions. Also,
it had introduced 4 courses from Stanford University, 31 from edx.org, and linked to
575 courses from edx.org. By then, XuetangX had built SPOC (Small Private Online
Course) platforms for 184 schools and institutions, opening 3350 SPOC courses. On
March 1, 2016, guokr.com released the 2015 Global MOOC Ranking List after survey
and statistics involving 14103 MOOC learners. According to the list, the top three
MOOC platforms worldwide with the most high-quality courses were Coursera.org,
edx.org, and XuetangX [9]. According to the statistical report by Xuetangx, “The
construction, operation and application of MOOCs by XuetangX (August 2016)”, the
course construction, the distribution of courses according to disciplines and the person-
time of sign-up and completion of XuetangX MOOCs are as follows [10].
First, XuetangX brings together high-quality domestic and foreign online course
resources, through independent development and introduction from foreign countries.
In addition, XuetangX has managed to cover the specialties of university degree courses,
and is developing towards systematism.
Figure 1 is the distribution of course resources of the 285 opened courses whose
courseware resources are on XuetangX (as of August 2016). As shown by the figure, of
the courses opened at XuetangX, 76.18% are from domestic colleges and universities,
2.62% are from domestic enterprises, 1.31% are from domestic non-business institutions,
and 19.9% are from abroad (including courses from some foreign institutions, colleges
and universities, in addition to those from prominent universities in the United States
and Australia, among others).
Second, Fig. 2 is the distribution of the 285 opened courses according to disciplines
whose courseware resources are on XuetangX (as of August 2016). Engineering courses
take the biggest share, accounting for 38.22% of the total courses; science courses come
second, making up 17.8% of the total; economics courses come third, representing
10.21% of the total.
zamfira@unitbv.ro
Survey and Analysis of the Application of MOOCs 843
Third, in terms of sign-up and completion, as of August 31, 2016, XuetangX had
recorded 6.39 million of online sign-ups, and 4.21 million of registered users, with the
average sign-up of each user reaching 1.5. Figure 3 is the distribution of sign-ups for
various courses. Of the total sign-ups, 25% are concentrated in economic and manage‐
ment courses, 23% in computing courses, and 20% in language courses. These three
types of courses are top three at XuetangX by sign-up. The courses have been operated
by 973 rounds, and 71,375 person-time have completed the courses, with the completion
rate being 1.15%.
zamfira@unitbv.ro
844 Y. Long et al.
XuetangX has 45,008 users from Tsinghua University, the sign-ups are 243,595, and
completion are 15,538. It has imported 706 courses from the University. According to
a survey by XuetangX, the percentage of students who have not attended blended courses
dropped from 68% in the fall of 2015 to 38% in the spring of 2016. That means more
and more students have been involved in blended learning. And the percentage of
students who have not completed online courses also dropped from 48% in the fall of
2015 to 24% in the spring of 2016. This implies that online courses have already become
part of students’ learning activities, and that online learning is known to a growing
number of students [11].
Additionally, 52 total MOOC-based blended course-time (38 courses) have been
opened, covering all levels from basic to professional, and all disciplines including
zamfira@unitbv.ro
Survey and Analysis of the Application of MOOCs 845
science, engineering, arts, and medicine. There are 14 engineering blended courses,
including Electrotechnics, Circuit Principle, Cloud Computing and Software Engi‐
neering, etc. The foundation courses in engineering account for 18% of the total blended
course (as shown in Fig. 4); the percentage of engineering courses is 42% (as shown in
Fig. 5).
zamfira@unitbv.ro
846 Y. Long et al.
engineering course at each round reached 10000–30000 s (as shown in Fig. 6), while
the completion was up to more than 600 (as shown in Fig. 7).
zamfira@unitbv.ro
Survey and Analysis of the Application of MOOCs 847
only 5 min or so. Not only the opening, developing, changing, and concluding, but also
each and every detail are crucial, posing the greatest challenge to me [13]”. Therefore,
in addition to a lot of funds, it requires teachers to invest a lot of time and energy to
make good MOOCs.
Apart from those teachers recording MOOC videos, while interviewing some
teachers using MOOC for teaching, it has also been recognized that, traditionally, the
calculation of teachers’ workload by the university is based predominantly on teaching
hours. In addition to the teaching hours scheduled by the university, teachers have to
spend a lot of time answering students’ questions on the Internet. However, such input
does not count into their workload. Before using MOOCs, most teachers only needed
to impart to students the knowledge well-known to them and answer their questions once
a week. After the introduction of MOOCs, teachers’ workload doubled and tripled, yet
there is not a corresponding supporting incentive mechanism for the calculation of
workload and rewarding measures, etc.
zamfira@unitbv.ro
848 Y. Long et al.
way. At the same time, colleges and universities should optimize their hardware and
software facilities and make them compatible with each other. For instance, the form of
desks and chairs in the classroom can be changed to provide students with a cozy learning
environment. Tsinghua University has set an example in this regard. It has arranged in
the teaching building roundtable classrooms and “transformers”-like desks, so as to
facilitate in-class discussions.
zamfira@unitbv.ro
Survey and Analysis of the Application of MOOCs 849
6 Conclusion
In conclusion, based on the afore-said surveys and analysis, we hold that Chinese
colleges and universities should be rational in the face of MOOCs boom. It is advisable
that they draw on advanced foreign experiences and the practice outcomes in teaching
reform of Chinese engineering education, and based on this, these colleges and univer‐
sities are advised to explore a MOOCs system taking on Chinese characteristics and, at
the same time, make the best use of MOOCs to assist traditional engineering education.
Only in this way can Chinese colleges and universities grasp the initiative to cultivate
top-notch global engineering talents.
Acknowledgment. We thanks XuetangX for providing the data about MOOCs, especially thanks
Chairman Nie Fenghua and Associate Curriculum Director Shi Xuelin for their kindly help.
References
zamfira@unitbv.ro
850 Y. Long et al.
3. Kizilcec, R., Piech, C., Schneider, E.: Deconstructing disengagement: analyzing learner
subpopulations in massive open online courses. In: Proceedings of the Third International
Conference on Learning Analytics and Knowledge, LAK 2013, Leuven, Belgium (2013)
4. Zong, Y., Sun, H., Zhang, H., Zheng, Q., Chen, L.: A logistic regression analysis of learning
behaviors and learning outcomes in MOOCs. Distance Educ. China (5), 14–22 (2016)
5. Jiang, L., Han, X., Cheng, J.: Analysis of the characteristics and learning effects of MOOCS
learners. China Educ. Technol. (11), 54–59 (2013)
6. Hong, M., Liu, M., Yang, J.: A study of the correlation between learning behavior and learning
effect. Coll. Engl. 5(2), 145–148 (2008). Academic Edition
7. Lv, Y., Yi, Y., Deng, C., S.Y.: The influence of network behavior on college students’
academic performance and mental health. Chin. J. Sch. Health 25(2), 250–251 (2004)
8. Mou, Z., Dong, B.: Exploration of blended learning mode based on MOOC. Modern Educ.
Technol. 24(5), 73–80 (2014)
9. MOOC Institute released the 2015 global MOOC rankings. http://www.jiemodui.com/N/
42596.html
10. Data comes from the statistical report by xuetangx.com, “The construction, operation and
application of MOOCs by XuetangX,” August 2016
11. Data comes from the survey report of application of MOOCs in Tsinghua University by
xuetangx.com
12. MOOCs provided by Tsinghua University on xuetangx.com. https://www.xuetangx.com/
courses?org=29
13. Xinjie, Y.: “Circuit Principle”—the first global MOOC by Tsinghua. http://www.tsinghua.edu.cn/
publish/news/4205/2015/20150324091655301219541/20150324091655301219541_.html
14. “Several Thoughts on the Reform and Development of Engineering Education in China”, a
report made by the Chinese Vice Minister of Education Lin Huiqing at the 2015 International
Forum on Engineering Education
zamfira@unitbv.ro
Conversion of a Software Engineering
Technology Program to an Online Format:
A Work in Progress and Lessons Learned
Abstract. Institutions may have multiple reasons for converting courses and
programs to an online format. The W Booth School of Engineering Practice and
Technology in McMaster University’s Faculty of Engineering has recently begun
the implementation of an online Software Engineering Technology Program as
part of the Schools’ Degree Completion Programs (the final 2-years of a
3-year-plus-2-year Degree). The intent of this conversion is to attract students
from an unlimited geographical area. While the online conversion is ongoing,
there are a number of important observations worth sharing. This paper provides
an overview of the motivation, challenges, and opportunities related to the
conversion of an existing Software Engineering Technology curriculum to a fully
online format. The purpose of this study is to highlight student feedback that
rejects the notion of a ‘flipped-classroom’ in favor of a more traditional
delivery-model (simply converted to an online format). This study also outlines a
suggested implementation model for the conversion of a curriculum to an online
format, with specific suggestions for the increased use of digital media, inter-
active resources, and synchronous online collaboration. Observations regarding
the development of supplementary course material and the resources required to
develop these materials are also provided. Recommendations from this study will
include: a suggested format for online delivery of engineering/technical courses,
suggestions regarding student assessment, a suggested timeline for implemen-
tation, suggested resources (technical support, etc.), suggested technology that
provides the greatest ease-of-use for both instructors and students, suggested
supplementary course materials, and a word about cost.
1 Introduction
Of the seven undergraduate programs, four of them are “degree completion pro-
grams” (DCP). These are designed for graduates of post-secondary institutions called
“colleges of applied arts and technology” or colleges for short (the college concept was
introduced in the province of Ontario in the mid 1960’s - college programs are
designed to be applied in nature and offer a diploma upon graduation). McMaster
University’s degree completion programs offer a four-year Bachelor of Technology
degree upon the completion of 24 courses above and beyond the completion of a
three-year college diploma in a related field. The degree completion programs are
designed to be more conceptual and theoretical in nature, to complement the applied
nature of the college diploma. They are also designed to be completed in a part-time or
full-time capacity. Students are often working in industry and completing the program
simultaneously. Generally, more than 50% of the students in the program are working
either full-time or part-time. In 2006 a Computing and Information Technology (re-
named in 2013 to Software Engineering Technology) program was included as a
DCP. The program contains 17 technical courses – similar in nature to Software
Engineering programs. These include courses such as Data Structures and Algorithms,
Software Requirements and Specification, Software Design, Software Architecture, etc.
It also contains 7 management courses, covering areas such as Engineering Economics,
Management Principles, The Management of Technical Projects, etc. The program was
originally introduced as a fully face-to-face offering, with no intention to offer courses
online. This was simply due to the fact that the other three DCPs were also being
offered in the traditional modality.
In this paper, we will present the motivation for converting a face-to-face program
in Software Engineering Technology to one that is fully online. We will then describe
the challenges we faced in designing the program and how we addressed them. Fol-
lowing the challenges, we will detail the current implementation of the program,
including the technology, resources, and assessment approaches we used. Next, we will
present the results of a small survey designed to measure the efficacy of our imple-
mentation. Lastly, we will conclude with key findings from our work.
Of the four degree completion programs, the software program has lagged in enroll-
ment since its inception in 2006. In a study that our department conducted in 2012, we
summarized the total number of graduates by program stream from colleges in Ontario
that feed students into our programs and the results are summarized in Fig. 1.
While the number of potential students available for our software program was
roughly equal to our manufacturing program, the actual enrollment for software was
roughly half that of manufacturing.
In 2013 we undertook a study to attempt to identify the reasons for low enrollment
in the software program specifically. We arrived at a list of a number of possible
factors, that were addressed in ways that are not relevant to this paper. However, one
factor - the geographic constraints of our location in the Hamilton area - was deemed to
be very important. A great many jobs and companies in the computing field are located
zamfira@unitbv.ro
Conversion of a Software Engineering Technology Program 853
Fig. 1.
in the Greater Toronto Area (GTA). Toronto is Canada’s largest city and is a major
business center. Toronto is about 50 miles from Hamilton and the commute is generally
about an hour each way – at the best of times. This commute would be difficult for
students living and working in the GTA. This was directly supported by data from our
study which indicated that the majority of students (55% in 2012) in our software
program came from one local college in Hamilton. As a result, we felt that the single
most important initiative that we could undertake to increase enrollment would be to
make the program accessible to students living and working anywhere in the province.
The software program was deemed to be particularly amenable to an online
delivery modality for a number of reasons:
1. The program historically used computer-based labs with open source implemen-
tations. These can easily be made accessible to students either through local or
cloud-based installations.
2. The students in the program are already “computer savvy” and are very familiar
with the use of technology for communication purposes.
3. The students see computer-based learning as an extension of the same learning
modality they routinely use at the technology companies that employ them.
There were a number of challenges in designing and developing a fully online program
in a part-time, degree completion context. We had anticipated these challenges a priori
and attempted to address as many as possible at design time. Unfortunately, for one of
the challenges, we did not provide an adequate solution before the course was run
online for the first time. Herein we will describe the challenges we faced and the
approaches that we used to ameliorate problems.
Our first design challenge was that all students are in various stages of completion
of the program. Therefore, there is no single “date” where all students that signed up for
a “face-to-face” program (before the transition to online) will have finished their
studies. Entry into the program is continuous over three semesters – Fall, Winter and
Summer. In order to address the challenge of not having cohorts in our program, we
zamfira@unitbv.ro
854 J. Fortuna et al.
opted to gradually phase-in the deployment of online courses with primarily technical
content (SFWR TECH courses). This was effective for three reasons. First, a number of
courses in the management (GEN TECH courses) part of our program had been online
for a few years. Therefore, most students currently enrolled in the program had already
been exposed to these courses which eased their transition to a fully online program.
Second, as we gradually rolled out SFWR TECH online courses over a period of about
two years (the first technical courses were developed in 2015 and all courses are
scheduled to run online after the fall of 2017), we obtained as much feedback as
possible from students regarding our design choices. A sample of that feedback will be
shown in the results section of this work. In any case, by receiving copious amounts of
feedback on an ongoing basis, we could make better choices for upcoming courses.
Third, we advertised that the program was moving in an online direction during all of
our advertising and recruiting activities, so that students were aware of the type of
program that they were signing up for.
A second challenge was that students have a wide range of opinions about what
constitutes a positive educational experience [1]. This is the challenge of managing
student expectations. In order to manage expectations, we attempted to reassure stu-
dents that the courses would be of high quality. We showed the students evidence that
considerable time and effort had gone into the design and deployment of our online
courses. While this was not a guarantee that students would find the efforts satisfactory,
we found that it was critical to reassure all those involved (particularly those that were
skeptical of an online modality – see below) that we were doing our best to deliver a
quality educational product.
There are a wide range of options for the design and delivery of an online course
[2], which created a third design challenge. Unfortunately, in meeting this challenge,
our initial efforts were somewhat off-base. In our initial investigation of pedagogical
approaches to online courses, it seemed obvious that a flipped-classroom approach is
particularly well suited to online courses. The flipped-classroom relies heavily on
supplemental material – often in the form of instructional video – that is meant to be
viewed asynchronously by the student. This is precisely the environment that an online
environment fosters. In fact, the flipped-classroom has been strongly advocated for in a
face-to-face environment as well [3]. Additionally, the management courses that were
already online were making extensive use of this approach, with apparent success.
However, as soon as we presented two technical courses – Data Structures and
Algorithms and Computer Security - online using this pedagogical design, a large body
of the students rejected the approach. In a meeting with the program chair early in the
semester of first offering of the aforementioned courses, student concerns were sum-
marized as:
• Lack of “instruction”.
• Missing the enjoyment of my lectures.
• Video lectures not equivalent to live lectures.
• Balance between lecture time and tutorial time.
• Flipped classroom model is contentious.
• Volume of work outside of the classroom.
zamfira@unitbv.ro
Conversion of a Software Engineering Technology Program 855
zamfira@unitbv.ro
856 J. Fortuna et al.
Each online course in the program was designed according to the following template
given to each course designer:
Guidelines for Developing On-Line Modules/Course
1. List the Main Learning Outcome and sub learning outcomes or sub topics
2. Provide Learning Materials (Asynchronous Mode) - videos, reading material,
powerpoint, interactive learning material
3. Provide Support Material (two types: one for reinforcing prerequisite material and
additional relevant material) – links, powerpoint, papers, other work
4. Provide Self-Assessment Tools & Exercises
5. Provide Interaction, Engagement and Discussion Activities (Synchronous Mode)
6. Provide Learning Outcome Assessment Tools (must be linked with the learning
outcomes) – tests, quizzes, assignments, group work, presentations, discussions
7. Provide Feedback on the learning outcome and module – instructor and student
zamfira@unitbv.ro
Conversion of a Software Engineering Technology Program 857
Herein we will examine selected results from an ethics approved recent survey where
we received 28 responses from students that took online courses during the calendar
year 2016. In some cases a student may have completed the survey multiple times as
the survey was designed for responses for each course taken.
Significant overall findings include:
• 67.86% of responses were from students employed-full time. 7.14% were employed
part-time. The remainder were not employed.
• 60.71% of responses were from students that would have to travel more than 12
miles to attend classes at McMaster. 10.71% would have to travel more than 48
miles to attend classes.
• 48.15% of responses were from students in the Software Engineering Technology
program. The remainder of the responses were from other DCP students taking our
online management courses.
• When asked which method of synchronous delivery respondents prefer (and they
could check all that applied), 85% of the responses favored Audio and Lecture
Slides (with annotation) and 74.07% preferred Audio and a Virtual Whiteboard.
29.63% preferred Audio and Video of the Instructor Speaking (webcam).
• When asked which type of interaction was preferred, 78.57% of responses were
from students that preferred to use chat. 10.71% preferred to use voice. The
remainder preferred not to interact.
• 46.43% of responses indicated that students had less interaction with classmates in
the online course. 35.75% indicated that they had the same or more interaction. The
remainder indicated that they did not interact at all.
• When asked how satisfied students were of the live component of the online course,
78.57% of the responses were from students that were either very satisfied or
satisfied.
• 67.86% of responses indicated that students found the live portion of the class most
effective.
• 64.29% of responses indicated that students were either very satisfied or somewhat
satisfied with the course as an online offering.
• 53.57% of the responses indicated that students preferred to take the course online
rather than face-to-face.
6 Conclusions
In this paper we have described the motivation for an online Software Engineering
Technology course and some of the challenges we were presented with during the
design of the program. We then detailed our implementation – based on a model that
roughly emulates the traditional face-to-face lecture through the use of live online
lectures with the addition of videos that provide supplemental tutorial support. We
found that students roundly rejected the notion of a “flipped-classroom” because they
deemed actual live instruction time in the form of a lecture to be very valuable.
zamfira@unitbv.ro
858 J. Fortuna et al.
Students felt that it was more efficient to be directly “instructed” rather than to be
forced to spend too much time conducting independent research. This makes sense
given the time constraints of students that are also working in industry in a full-time or
part time capacity.
The survey results directly support our contention that our focus on the live online
lecture as a replacement for the traditional face-to-face in class experience has been
quite well received, as the vast majority of our students were satisfied with our online
courses and a majority expressed a preference for taking the course online over a
face-to-face delivery even though about 40% of the responses indicated they would
have had to travel less than 12 miles to attend the class. Therefore, we feel that the
design decisions that we made and the implementation (although ongoing) have met
the needs of our student body that must maintain a work/life balance while furthering
their education in the fast-paced field of software design and development.
References
1. Pashler, H., McDaniel, M., Rohrer, D., Bjork, R.: Learning styles: concepts and evidence.
Psycholog. Sci. Public Interest 9(3), 105–119 (2009)
2. Duncan, H.E., Young, S.: Online pedagogy and practice: challenges and strategies. The
Researcher 22(1), 17–32 (2009)
3. Bishop, J.L., Verleger, M.A.: The flipped classroom: a survey of the research. In: ASEE
National Conference Proceedings, Atlanta, GA, vol. 30, no. 9, June 2013
4. Kelly, H.F., Ponton, M.K., Rovai, A.P.: A comparison of student evaluations of teaching
between online and face-to-face courses. Internet High. Educ. 10(2), 89–101 (2007)
5. Redpath, L.: Confronting the bias against on-line learning in management education. Acad.
Manag. Learn. Educ. 11(1), 125–140 (2012)
6. Sathik, M.M., Sofia, G.J.: Effect of facial expressions on student’s comprehension recognition
in virtual educational environments. SpringerPlus. 2, 455 (2013)
7. Gressman, P.: Engaging students through technology symposium 2015 presentation by Phillip
Gressman, 01 October 2015. http://repository.upenn.edu/showcase_videos/97
zamfira@unitbv.ro
Increasing the Value of Remote Laboratory
Federations Through an Open Sharing
Platform: LabsLand
1 Introduction
An Educational Remote Laboratory is a software and hardware solution that
enables students to access real equipment located in their institution, as if they
were in a hands-on-lab session, using an standard web-browser. The laboratories
are typically deployed in universities or research centers.
A key factor of remote laboratories is that once they are available through the
Internet their usage can be scaled up and used by students of other institutions.
c Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6 80
zamfira@unitbv.ro
860 P. Orduña et al.
Thus, two or more institutions can share different equipment to reduce costs by
requiring less duplicated equipment: it is typically only used in certain hours of
the day and in certain days of the year. Furthermore, this empowers a sharing
economy where multiple providers provide access to their laboratories to each
other, freely or not.
In the literature there is a wide range of remote laboratories in many fields
(e.g., robotics, electronics, physics, chemistry). Software frameworks have been
developed to make the development of remote laboratories more affordable (e.g.,
Remote Laboratory Management Systems such as WebLab-Deusto1 [30], iLab
Shared Architecture2 , RemLabNet3 [37] or Labshare Sahara4 [22]) and tools
(e.g., gateway4labs5 [32]) to provide integrations with educational tools (such as
Moodle, Sakai or other LMS, both through ad hoc solutions and through stan-
dards such as IMS LTI) or repositories linking remote and virtual laboratories
(such as Go-Lab [13,21], LiLa [34] or iLabCentral).
However, while the number of remote laboratory initiatives is high, the over-
all impact of these laboratories is fairly limited beyond the scope of the host
institution or the scope (and duration) of projects in which the host institution
is involved. There are cases where the laboratories are regularly used by other
institutions, but these are still exceptions and remote laboratories are not yet
widely used. This is not the case for virtual laboratories (simulations), where
the maintenance costs and work required once developed tend to be low.
In the literature there are studies that identify key elements for this problem:
lack of a technical framework, pedagogic framework or proper strategy. Some
business initiatives have been created focusing also on the sustainability (such
as Labicom6 [39] or RemoteLabs.in), but also with a limited reported impact.
The focus of this contribution is to outline a novel initiative addressing this
scaling problem. After over 10 years working on the area our research group
has started a spin-off, called LabsLand7 focused on this topic. This contribution
outlines the key component developed by this company: a portal that acts as a
repository of remote laboratories supporting multiple providers (relying on exist-
ing interoperability efforts), but which provides a quality assurance mechanism
(not available at this moment in any of the repositories found in the literature),
and based on simple contracts (supporting both free sharing and paid sharing)
that aims to provide reliability to the final users and sustainability for the lab-
oratory providers.
1
http://weblab.deusto.es.
2
http://ilab.mit.edu.
3
http://www.remlabnet.eu.
4
https://remotelabs.eng.uts.edu.au.
5
http://gateway4labs.readthedocs.org.
6
http://labicom.net.
7
https://labsland.com.
zamfira@unitbv.ro
Increasing the Value of Remote Laboratory Federations 861
Fig. 1. Robot laboratory [9]. At the left, the mobile robot itself. At the right, the user
interface once the program has been submitted.
zamfira@unitbv.ro
862 P. Orduña et al.
– Leave the laboratories completely open, so whoever wants to use them can use
them. This may reduce the chances of providing proper Learning Analytics or
supporting proper accountability mechanisms, in addition to avoiding priori-
ties among students coming from different institutions, leading to a tradeoff
between accessibility and advanced features [32].
– Share accounts between the different RLMS: if University A want to use
laboratories of University B, then someone in University A will provide a list
of usernames to University B and students will go to this institution using
credentials in University B. Ideally, some federated authentication could be
used to avoid providing credentials in different domains (such as Shibboleth,
OAuth or similar), but it is not typically the case.
– Federate laboratories: if a RLMS supports federation, then if installed in
two different institutions (e.g., University A and University B ), students of
University A will go to the RLMS of University A and they will transparently
use laboratories in University B, working in a institution-to-institution basis
(so University B does not need to know the list of students of University A
and simply rely on an existing agreement with that university).
From the items in this list, the most advanced mechanism is the federation
of remote laboratories through proper protocols oriented to market-like situa-
tions. These federation protocols have been used for fostering interoperability
between RLMS [30]. These interoperable bridges between different systems can
be enhanced if properties such as transitivity or federated load balance are pro-
vided [28].
8
http://ilab.mit.edu.
9
http://weblab.deusto.es.
10
http://github.com/saharalabs.
zamfira@unitbv.ro
Increasing the Value of Remote Laboratory Federations 863
In the literature there are different portal solutions that provide listings of virtual
and remote laboratories. In [8] 13 repositories are analyzed, out of which 6 were
involving remote laboratories (the rest are virtual laboratories - simulations),
plus another one is presented (golabz). Most of these repositories provide a por-
tal with more or less features, including: social features (e.g., rating resources,
adding comments, tags), materials (users’ materials, students’ materials, teach-
ers’ materials) or supportive apps. In particular, the Go-Lab portal (golabz11 )
provides all these features, providing support for both remote and virtual labora-
tories, a pedagogic framework, tools for sharing and reusing pedagogic contents
and tools for publishing results.
The last years have seen the rise of the access economy platforms, very often also
known as sharing economy platforms. Those companies rely on a business model
where their services are traded on the basis of access, as opposed to ownership.
This relatively novel term arose as a correction to the sharing economy one
to emphasize the fact that many of the major players do not actually share
anything. Instead, they provide the means for a provider to temporarily lend
something to a consumer.
A representative example is Airbnb12 . Airbnb is an online apartment and
room rental platform. It differentiates two types of user roles: hosts—who provide
the room or house—and guests—who pay for it. The platform provides significant
value to both. To hosts, they provide the means to get their property known,
to manage their payments easily, to schedule stays, to review potential guests,
etc. To guests, they provide the means to find rooms or whole apartments and
rent them easily, compare them in a competitive environment, be able to asses
the quality and security of the host through the rating and host identification
system, etc.
Other interesting example is BlaBlaCar13 . It is a car ride-sharing platform
that enables drivers and passengers to organize rides. It is particularly popular
in Spain. Whenever the Blablacar drivers are going to ride to a particular place,
they may post the details in the platform, along with the number of available
seats and the amount they charge. Then, those interested can join them. It is
noteworthy that, unlike Airbnb, the focus of Blablacar is not for the drivers to
earn money, but, simply, to let them recover part of the expenses of the trip.
Thus, they actually limit the maximum amount to charge depending on the trip.
While in this contribution we refer to these phenomenom as an access econ-
omy or a sharing economy [24,25] (its most popular term nowadays), this phe-
nomenom is not new and has received different names in the literature [2].
11
http://www.golabz.eu.
12
https://www.airbnb.com.
13
http://blablacar.com.
zamfira@unitbv.ro
864 P. Orduña et al.
Regardless the term used, they all agree in that a key feature in all these plat-
forms is delivering verifiable trust. Airbnb co-founder Joe Gebbia [10] points out
trust as a key feature for Airbnb since both guests and hosts need to trust each
other to run the service, explaining how the Airbnb reputation system does not
change much with few opinions but after a threshold of around 10 (good) opin-
ions, the chances of one trusting that person increase considerably. Indeed, trust
is analyzed in studies [43] which compare the results of the reputation system of
Airbnb with hotels opinions in TripAdvisor (where the average punctuation is
lower than Airbnb), pointing out some potential reasons (from individual entre-
preneurs offering rooms in Airbnb would be more selective with which guests to
accept so as to avoid a bad opinion or resetting the page to a fresh property
page to avoid having past bad opinions).
4 Discussion
As mentioned in Sect. 2.4 Remote laboratory portals, some repositories (such
as golabz) for virtual and remote laboratories do provide social features (e.g.,
rating resources, adding comments, tags), materials (users’ materials, students’
materials, teachers’ materials), as well as supporting applications, a pedagogic
framework or tutoring platforms [3].
However, none of the existing portals provides trust mechanisms other than
user-based ratings, and typically those portals supporting user-based ratings do
not have many ratings, and with few context about who is the person providing
the rating, how much as used the tool, etc.
This might be a minor issue in certain environments, where trust on the
tool is not so critical, such as where the tool is reliable (e.g., it is a simulation)
and it is always freely available. However, in the field of remote laboratories,
sustainability remains as an open problem [23]. An important advantage is that
two institutions can share real laboratories, real equipment, reducing the costs
if these costs are somehow shared. However, regardless the federation protocols
built (see Sect. 2.3), it is not possible to engage different institutions in using
each others’ laboratories if there is no reliable mechanism to trust the reliability
of the laboratories and the ability of the host institution to fix the problems if
they appear.
However, there is usually no way for the existing remote laboratory portals
to know whether the laboratories are running or not, how often failures happen
(e.g., what was the percentage of uptime during the last 9 months?) or how
long does it usually take to the laboratory owners to fix it. The portals are
additionally unable to track the usage of the tools from third parties (e.g., they
are when it is about their own resources) and publicly display this information
in a way useful for teachers (e.g., what laboratories are more popular in the
repository?). Furthermore, the existing portals do not embrace the ability to
manage the usage of the laboratories and manage potential payments.
We consider that this is a key factor for the not spread usage of remote
laboratories among different institutions (as compared with their wide usage in
zamfira@unitbv.ro
Increasing the Value of Remote Laboratory Federations 865
the host institutions [36,42]). No RLMS and no portal provide the ability to
manage payments properly (not necessarily by paying per access, but through
fixed rates or virtual mechanisms where the more someone uses your labs, the
more you can use labs of other providers), and even if they did, no portal has
the ability to provide real data on how reliable and trustworthy a particular
laboratory is. Without such information, the trust relation completely relies on
direct relationship between the provider and the consumer (where the consumer
must trust on the provider because they know each other or other reputation
system).
With such a portal, and with the technologies, pedagogic frameworks and
tools already demonstrated in the literature, we consider that it will be possible
to engage different providers and different consumers to use remote laboratories.
Only if this adoption happens, it will be possible to foresee a sustainable and
maintainable model for a distributed network of remote laboratories.
5 LabsLand Portal
LabsLand is a spin-off of the WebLab-Deusto14 research group. As part of
WebLab-Deusto, the team has worked on:
– A set of remote laboratories, including physics, electronics or biology.
– An Open Source Remote Laboratory Management System called WebLab-
Deusto for the development of remote laboratories. This RLMS is used in
a number of universities in different countries (Spain15 , Slovakia16 , Brazil17 ,
Serbia [26] or Georgia).
– A federation model and protocol for sharing laboratories in a market-like
decentralized environment [28].
– A set of tools for interoperability with other RLMS: both ad hoc [30,33] and
through a collaboratively developed and Open Source system called gate-
way4labs [31,32].
Now as part of LabsLand, and on top of the experience and maintaining
the tools mentioned above as Open Source, a spin-off has been created to deal
with the pitfalls that will be pointed out in the paper. The spin-off, in addition
to explore business models around remote laboratories, provides a centralized
portal (see Fig. 2) that will act as a technology-agnostic marketplace for remote
laboratories consumers and providers. The portal does not rely only in own
technologies (such as WebLab-Deusto), but also supports external providers. To
this end, interoperability efforts have been placed in gateway4labs18 to support
external remote laboratory providers, including the iLab Shared Architecture,
RemLabNet, UNR-FCEIA or repositories including remote laboratories such
as ViSH.
14
http://weblab.deusto.es.
15
http://weblab.ieec.uned.es.
16
http://weblab.chtf.stuba.sk.
17
http://weblabduino.pucsp.br/weblab/.
18
https://github.com/gateway4labs/.
zamfira@unitbv.ro
866 P. Orduña et al.
Fig. 2. LabsLand portal. It will be publicly available for the beginning of 2017.
The portal tries to guarantee trust from the very beginning, by regularly and
automatically checking the existing resources to be able to provide trustworthy
information to the teachers on which laboratories are reliable and how much
(with several different types of automatic checks), and enforcing different policies
on the remote laboratory providers to be clear on when the laboratory is going
to be available. The portal will not penalize that a remote laboratory provider
is not working 24/7: it will just require remote laboratory providers to define
in which time ranges it should be available and penalize those not being online
during the defined time. In addition to this, public opinions (only of those having
used the platform) and all the D.R.E.A.M.S. framework values are considered
in its design.
The LabsLand portal emphasizes certain technical characteristics and fea-
tures to provide and guarantee high quality and standards.
zamfira@unitbv.ro
Increasing the Value of Remote Laboratory Federations 867
zamfira@unitbv.ro
868 P. Orduña et al.
Teachers often need to trust that the remote laboratories they intend to use will
indeed work. Traditionally, in the remote laboratory environment, there have
been difficulties at that respect. Maintaining a remote laboratory takes effort
and resources, and due to their nature and mechanical components, are often
prone to failures.
LabsLand has designed and developed a system to automatically run dis-
tributed tests against deployed laboratories, both against internal (LabsLand-
provided) laboratories and external ones. This way, it will be possible to obtain
reliability and quality data, that will let potential laboratory users make con-
scious and informed decisions about which labs they rely on, and to what extent.
One of the most interesting of these laboratories is the Arduino Robot Lab-
oratory, which is designed, developed and maintained by LabsLand. This lab
is oriented to robotics students of all ages, or even to young programming stu-
dents. The equipment is a line-following robot controlled with an Arduino. It has
several sensors, including those to follow the line, several proximity sensors, etc.
The students can program the robot (by programming the Arduino), and
then see how their program behaves in the real robot through a webcam. There
are also buttons, a serial monitor and other devices available, so the student can
actually interact with the robot in real-time.
Because, as explained, it is intended to be useful for a wide audience, there are
two different programming interfaces available. First, young students (or those
who simply are not particularly inclined to Arduino C-like programming) can use
Blockly20 . Blockly is an Open Source scratch-like21 visual programming language
created by Google. The blocks that the LabsLand Arduino Robot Laboratory
provides match the robot’s library quite closely, so even through the blocks it is
possible to obtain a relatively low-level understanding of the robotics involved.
20
https://developers.google.com/blockly/.
21
https://scratch.mit.edu/.
zamfira@unitbv.ro
Increasing the Value of Remote Laboratory Federations 869
Second, those with some programming knowledge can program the robot directly
in the Arduino C-like language, through a simplified online IDE.
Figure 3 shows the actual robot to the left, and the interaction user interface
of the Arduino robot laboratory to the right.
22
http://bifi.es.
23
http://relle.ufsc.br/.
zamfira@unitbv.ro
870 P. Orduña et al.
Brazil. Certain Relle labs will be offered through the LabsLand portal. This
serves two main goals. Firstly, it adds direct value to the portal users through
a wider selection of useful, quality labs. Secondly, it demonstrates the federated
and inclusive nature of the LabsLand portal by integrating from the beginning
laboratories from different providers and different frameworks.
Among others, some remarkable Relle laboratories are the DC Electrical
Panel and the AC Electrical Panel laboratories, which provide access to physi-
cal boards that let students carry out different basic electronic calculations over
basic electronic components. Other interesting examples are the Optical Labora-
tory which lets the students experiment with optical lenses; or the Microscope,
which lets students examine different samples with a microscope in real-time.
Figure 5 shows the Relle AC Electrical Panel, which lets students predict
the value for the voltage and intensity for several circuits and compare their
predictions with reality.
zamfira@unitbv.ro
Increasing the Value of Remote Laboratory Federations 871
7 Conclusions
The daily usage of remote laboratories has been reported in the literature. How-
ever, while the number of remote laboratory initiatives is high, the overall impact
of these laboratories is fairly limited beyond the scope of the host institution or
the scope (and duration) of projects in which the host institution is involved.
In this contribution, existing efforts have been described from a technical
perspective, and sharing economy platforms have been described, identifying
some factors such as trust as key for their success. Thus, the contribution suggests
that this factor must be included in any portal attempting to encourage adoption
on the usage of remote laboratories beyond the scope of the host institution and
related projects or direct relationships.
Finally, the contribution presents the LabsLand portal (as part of a spin-off
of the WebLab-Deusto research group), which attempts to provide these features
so as to foster adoption of remote laboratory uses for achieving sustainability and
maintainability of remote laboratories. The contribution describes the features
and design philosophy of this portal, which will be available in early 2017.
References
1. Aktan, B., Bohus, C., Crowl, L., Shor, M.: Distance learning applied to control
engineering laboratories. IEEE Trans. Educ. 39(3), 320–326 (1996)
2. Böckmann, M.: The shared economy: it is time to start caring about sharing; value
creating factors in the shared economy. In: 1st IBA BT Conference, vol. 1 (2013)
3. Cao, Y., Tsourlidaki, E., Edlin-White, R., Dikke, D., Faltin, N., Sotiriou, S., Gillet,
D.: Stem teachers community building through a social tutoring platform. In:
Advances in Web-Based Learning, ICWL 2015, pp. 238–244. Springer (2015)
4. Bohus, C., Crowl, L.A., Aktan, B., Shor, M.H.: Running control engineering exper-
iments over the Internet. IFAC Proc. Vol. 29(1), 2919–2927 (1996)
5. Cedazo, R., Sanchez, F., Sebastian, J., Martı́nez, A., Pinazo, A., Barros, B., Read,
T.: Ciclope chemical: a remote laboratory to control a spectrograph. In: Advances
in Control Education, ACE (2006)
6. Coble, A., Smallbone, A., Bhave, A., Watson, R., Braumann, A., Kraft, M.: Deliver-
ing authentic experiences for engineering students and professionals through e-labs.
In: 2010 IEEE Education Engineering (EDUCON), pp. 1085–1090. IEEE (2010)
7. Del Alamo, J., Brooks, L., McLean, C., Hardison, J., Mishuris, G., Chang, V.,
Hui, L.: The MIT microelectronics weblab: a web-enabled remote laboratory for
microelectronic device characterization. In: World Congress on Networked Learning
in a Global Environment, Berlin, Germany (2002)
8. Dikke, D., Tsourlidaki, E., Zervas, P., Cao, Y., Faltin, N., Sotiriou, S., Sampson,
D.G.: Golabz: towards a federation of online labs for inquiry-based science educa-
tion at school. In: 6th International Conference on Education and New Learning
Technologies (EDULEARN) (2014)
9. Dziabenko, O., Garcı́a-Zubia, J., Angulo, I.: Time to play with a microcontroller
managed mobile bot. In: 2012 IEEE Global Engineering Education Conference
(EDUCON), pp. 1–5. IEEE (2012)
10. Gebbia, J.: How Airbnb designs for trust. In: TED Talks, February 2016
11. Geer, D.: Taking steps to secure web services. Computer 36(10), 14–16 (2003)
zamfira@unitbv.ro
872 P. Orduña et al.
12. Gillet, D., Latchman, H., Salzmann, C., Crisalle, O.: Hands-on laboratory experi-
ments in flexible and distance learning. J. Eng. Educ. 90(2), 187–191 (2001)
13. Gillet, D., de Jong, T., Sotirou, S., Salzmann, C.: Personalised learning spaces
and federated online labs for stem education at school. In: 2013 IEEE Global
Engineering Education Conference (EDUCON), pp. 769–773. IEEE (2013)
14. Gomes, L., Bogosyan, S.: Current trends in remote laboratories. IEEE Trans. Ind.
Electron. 56(12), 4744–4756 (2009)
15. Gordillo Méndez, A., Barra Arias, E., Quemada Vives, J.: Enhancing k-12 science
education through a multi-device web tool to facilitate content integration and
e-infrastructure access. In: 7th International Technology, Education and Develop-
ment Conference (INTED2013), 04–05 March 2013, Valencia, Spain, pp. 5432–5440
(2013)
16. Gravier, C., Fayolle, J., Bayard, B., Ates, M., Lardon, J.: State of the art
about remote laboratories paradigms-foundations of ongoing mutations. iJOE 4(1)
(2008)
17. Gustavsson, I., Zackrisson, J., Håkansson, L., Claesson, I., Lagö, T.: The VISIR
project–an open source software initiative for distributed online laboratories. In:
Proceedings of the REV 2007 Conference, Porto, Portugal (2007)
18. Hardison, J., DeLong, K., Bailey, P., Harward, V.: Deploying interactive remote
labs using the ilab shared architecture. In: 38th Annual Conference on Frontiers
in Education Conference, FIE 2008, p. S2A–1. IEEE (2008)
19. Hardt, D.: The OAuth 2.0 authorization framework (2012)
20. Henry, J.: Running laboratory experiments via the world wide web. In: ASEE
Annual Conference (1996)
21. de Jong, T., Linn, M.C., Zacharia, Z.C.: Physical and virtual laboratories in science
and engineering education. Science 340(6130), 305–308 (2013)
22. Lowe, D., Machet, T., Kostulski, T.: UTS Remote Labs, Labshare, and the Sahara
architecture. Using Remote Labs in Education: Two Little Ducks in Remote Exper-
imentation, p. 403 (2012)
23. Lowe, D., de la Villefromoy, M., Jona, K., Yeoh, L.: Remote laboratories: uncover-
ing the true costs. In: 2012 9th International Conference on Remote Engineering
and Virtual Instrumentation (REV), pp. 1–6. IEEE (2012)
24. Martı́nez-Polo, J., Martı́nez-Sánchez, J.T., Vivó, J.M.N.: Participation and sharing
economy: The Spanish case of #compartirmola. In: Entrepreneurship, Business and
Economics, vol. 1, pp. 15–22. Springer (2016)
25. Matzner, M., Chasin, F., von Hoffen, M., Plenter, F., et al.: Designing a peer-to-
peer sharing service as fuel for the development of the electric vehicle charging
infrastructure. In: 2016 49th Hawaii International Conference on System Sciences
(HICSS), pp. 1587–1595. IEEE (2016)
26. Milošević, M., Milošević, D., Dimopoulos, C., Katzis, K.: Security challenges
in delivery of remote experiments. XXII Skup TRENDOVI RAZVOJA: NOVE
TEHNOLOGIJE U NASTAVI, Zlatibor, 16–19 Feb 2016
27. Nedic, Z., Machotka, J., Nafalski, A.: Remote laboratory netlab for effective inter-
action with real equipment over the internet. In: 2008 Conference on Human Sys-
tem Interactions, pp. 846–851. IEEE (2008)
28. Orduña, P.: Transitive and scalable federation model for remote laboratories. Ph.D.
thesis, Universidad de Deusto, Bilbao, Spain, May 2013. http://paginaspersonales.
deusto.es/porduna/phd/
29. Orduña, P., Almeida, A., Ros, S., Lpez-de Ipiña, D., Garcı́a-Zubia, J.: Leveraging
non-explicit social communities for learning analytics in mobile remote laborato-
ries. J. Univers. Comput. Sci. 20(15), 2043–2053 (2014)
zamfira@unitbv.ro
Increasing the Value of Remote Laboratory Federations 873
30. Orduña, P., Bailey, P., DeLong, K., López-de Ipiña, D., Garcı́a-Zubia, J.: Towards
federated interoperable bridges for sharing educational remote laboratories. Com-
put. Hum. Behav. 30, 389–395 (2014). http://www.sciencedirect.com/science/
article/pii/S0747563213001416
31. Orduña, P., Botero Uribe, S., Hock Isaza, N., Sancristobal, E., Emaldi, M., Pes-
quera Martin, A., DeLong, K., Bailey, P., López-de Ipiña, D., Castro, M., Garcı́a-
Zubia, J.: Generic integration of remote laboratories in learning and content man-
agement systems through federation protocols. In: 2013 IEEE Frontiers in Educa-
tion Conference, Oklahoma City, OK, USA, pp. 1372–1378, October 2013
32. Orduña, P., Garbi Zutin, D., Govaerts, S., Lequerica Zorrozua, I., Bailey, P.H.,
Sancristobal, E., Salzmann, C., Rodriguez-Gil, L., DeLong, K., Gillet, D., et al.:
An extensible architecture for the integration of remote and virtual laboratories in
public learning tools. Tecnologias del Aprendizaje, IEEE Revista Iberoamericana
de 10(4), 223–233 (2015)
33. Orduña, P., Lerro, F., Bailey, P., Marchisio, S., DeLong, K., Perreta, E., Dzi-
abenko, O., Angulo, I., López-de Ipiña, D., Garcia-Zubia, J.: Exploring complex
remote laboratory ecosystems through interoperable federation chains. In: 2013
IEEE Education Engineering (EDUCON). IEEE (2013)
34. Richter, T., Boehringer, D., Jeschke, S.: LiLa: a European project on networked
experiments. Autom. Commun. Cybernet. Sci. Eng. 2009(2010), 307–317 (2011)
35. Safaric, R., Truntič, M., Hercog, D., Pačnik, G.: Control and robotics remote lab-
oratory for engineering education. Int. J. Online Eng. (iJOE) 1(1) (2005)
36. Santana, I., Ferre, M., Izaguirre, E., Aracil, R., Hernandez, L.: Remote laboratories
for education and research purposes in automatic control systems. IEEE Trans. Ind.
Inf. 9(1), 547–556 (2013)
37. Schauer, F., Krbecek, M., Beno, P., Gerza, M., Palka, L., Spilakov, P., Tkac, L.:
Remlabnet iii – federated remote laboratory management system for university and
secondary schools. In: 2016 13th International Conference on Remote Engineering
and Virtual Instrumentation (REV), pp. 238–241. IEEE (2016)
38. Simão, J.P.S., et al.: Relle: Sistema de gerenciamento de experimentos remotos
(2016)
39. Titov, I.: Labicom.net-the on–line laboratories platform. In: 2013 IEEE Global
Engineering Education Conference (EDUCON), pp. 1137–1140. IEEE (2013)
40. Torres, F., Candelas, F., Puente, S., Pomares, J., Gil, P., Ortiz, F.: Experiences
with virtual environment and remote laboratory for teaching and learning robotics
at the University of Alicante. Int. J. Eng. Educ. 22(4), 766–776 (2006)
41. Zappatore, M., Longo, A., Bochicchio, M.A.: Enabling MOOL in acoustics by
mobile crowd-sensing paradigm. In: 2016 IEEE Global Engineering Education Con-
ference (EDUCON), pp. 733–740. IEEE (2016)
42. Zappatore, M., Longo, A., Bochicchio, M.A., Zappatore, D., Morrone, A.A., De
Mitri, G.: Mobile crowd sensing-based noise monitoring as a way to improve learn-
ing quality on acoustics. In: 2015 International Conference on Interactive Mobile
Communication Technologies and Learning (IMCL), pp. 96–100. IEEE (2015)
43. Zervas, G., Proserpio, D., Byers, J.: A first look at online reputation on Airbnb,
where every stay is above average. Where Every Stay is Above Average, 23 January
2015
zamfira@unitbv.ro
Standardization Layers for Remote Laboratories
as Services and Open Educational Resources
1 Introduction
Offering hands-on sessions are one of the main requirements for implementing
STEM Education (Science Technology Engineering and Mathematics) [3,8]. By
conducting laboratory work, pedagogical objectives such as learning by doing,
applying theory to practice, learning to manipulate the physical environment and
understanding its flaws and limitations can be attained. Remote experimentation
is one way to attain that goal. Broadly speaking, a remote lab is a real physical
lab which is accessible at distance through computer networks. More specifically,
c Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6 81
zamfira@unitbv.ro
Std. Layers for Remote Laboratories as Services and OER 875
1
http://www.golabz.eu/.
2
https://www.oercommons.org/.
zamfira@unitbv.ro
876 W. Halimi et al.
2 Standardization Layers
Remote laboratories are highly interactive open educational resources. The main
goal of remote laboratories is to make available a real physical lab setup at
distance. To use a remote lab, users act on the system by sending commands,
consequently the lab responds by changing its physical state, and returning its
sensor values. Accordingly, remote labs are a wealthy source of different types
of data. Through the use of the remote lab, the students produce two types of
data: interaction data related to their actions and inputs on the user interface,
and experimental data which are the parameters they applied on the system and
the collected results as a consequence. It is unarguably important to make sense
out of the activity traces and experimental data which are generated. Therefore,
in addition to having both types of data, it is necessary to dispose of the context
information. The context is composed of the user identity, the course or the
pedagogical scenario in which the resources are utilised and any other details
which will contribute to a better perspective of the educational activity.
zamfira@unitbv.ro
Std. Layers for Remote Laboratories as Services and OER 877
lab, in addition to configuration data that puts the lab in a certain operational
mode, if supported. Moreover, there are other requests that the UI can send, for
example writing data requests on actuators for controlling the lab. At this level
we assume that the lab is capable of accepting requests and sending responses
about the sensors and the actuators. The lab as a service is a self-contained layer
that is operational regardless of a hosting platform. The information provided
by this layer is available to any platform trying to interact with the lab. The
concept of LaaS is detailed in Sect. 2.1.
Next, the remote lab is personalized as an Open Educational Lab (OEL) by
the development of a UI integrating the pedagogical elements required by the
context, and which is augmented with the necessary functionalities to insure
proper communication with the hosting platform and the interfacing with the
remote lab. This is done by calling adequate services of the LaaS. The concept
of OEL extending the notion of OER is detailed in Sect. 2.2.
Last in Layer 3, the OEL is integrated in a hosting platform while insuring
the propagation of contextual information, user activity traces, as well as data
related to the experimentation itself.
zamfira@unitbv.ro
878 W. Halimi et al.
and protect their labs. They are not accessible to the Client Application. Further
details on LaaS and the Smart Device paradigm are in [12].
zamfira@unitbv.ro
Std. Layers for Remote Laboratories as Services and OER 879
Once a laboratory setup is abstracted from the physical world as a set of services
based on the Smart Device paradigm, it is ready to be personalised for use. This
is done by building an application which invokes the API calls of the LaaS to
gain access to it. At this stage, the remote laboratory can be exploited without
any context (i.e. without being part of a lesson). But if chosen to be used in
a pedagogical scenario, augmenting the LaaS with a UI (as a front-end), and
with user identity management, activity tracking, and experimental data man-
agement (as back-end) turns it into an OEL ready to be integrated in a hosting
platform. Needless to say that the interfaces used with the hosting platform for
the mentioned requirements will be specific and cannot be standardized. This is
the level where the user credentials are managed, the activities traces are con-
sumed, and the experimental data is saved. In the next section, we present two
implementations of two different remote labs integrated in two online learning
environments with different infrastructures.
zamfira@unitbv.ro
880 W. Halimi et al.
3 Use Cases
In this section, we present two examples of remote laboratories developed and
integrated in learning environments as per the proposed guidelines. We will first
present the example of a control system lab which is integrated in edX3 . Then
we will detail the Mach-Zehnder interferometer remote lab which is integrated
in an educational social media platform: graasp4 .
The lab we are considering for this example is a control systems lab designed
and implemented to service a large number of users. It is integrated as part of
a control systems course, designed and deployed on a local copy of edX5 for
EPFL. The complete infrastructure of the lab is made of multiple replicas of the
same lab setup serviced on the Internet by Smart Devices, an HTML UI to be
integrated in edX, a .cgi interface for LTI authentication, database and other
services, and an edX server.
zamfira@unitbv.ro
Std. Layers for Remote Laboratories as Services and OER 881
lab UIs, there is a need to identify each tab for activity tracking and data
archiving.
– experiment duration: the teacher can set the time of the experiment through
edX in the LTI parameters. This will limit the amount of time students can
spend doing an experiment if others are waiting for turns.
At the time of the integration of the remote lab, edX didn’t implement an LTI
version which supports saving and retrieving data to the platform. Hence, a cgi
interface was put in place between edX and the external tools. The cgi interface
validates the LTI encoded request containing the edX user ID and other context
related information. Once the request is validated, the LTI module content is
integrated as an iFrame in the edX page (Fig. 2).
6
http://www.ni.com/myrio/.
zamfira@unitbv.ro
882 W. Halimi et al.
zamfira@unitbv.ro
Std. Layers for Remote Laboratories as Services and OER 883
abstracted as a set of web services accessible through APIs. This allows the per-
sonalisation of the OEL with a UI augmented with full-integration requirements
(context awareness, activity tracking, and experimental data storage) becomes
an Open Educational Lab. The resulting OEL is expected to interoperate with
the different services and other Open Educational Resources used in a learn-
ing scenario, hence providing an online learner what is a good experience. We
later present two remote laboratories integrated in two different online plat-
forms, with pedagogically sound resources and interaction features. The control
system lab is integrated in an LTI consuming platform. The lab is integrated
as an OEL by implementing the communication with the LTI container of the
platform (single-sign on for both platform and lab) and extending its properties
to support further user needs (data saving and retrieval). The other example,
is an interferometer integrated in an educational social media platform which
supports the requirements for integration. The lab is integrated as an OEL by
using the services provided by Graasp for user identity, activity tracking, and
data storage and retrieval. It is worth mentioning that the proposed solution for
edX could also reused with little effort in other environments supporting LTI
such as Moodle10 .
References
1. Learning Tools Interoperability - Implementation Guide (2016). https://www.
imsglobal.org/specs/ltiv1p2/implementation-guide Accessed 15 Jan 2017
2. Bogdanov, E., Limpens, F., Li, N., El Helou, S., Salzmann, C., Gillet, D.: A social
media platform in higher education. In: Global Engineering Education Conference
(EDUCON), 2012 IEEE, pp. 1–8. IEEE (2012)
3. Dasarathy, B., Sullivan, K., Schmidt, D.C., Fisher, D.H., Porter, A.: The past,
present, and future of moocs and their relevance to software engineering. In: FOSE
2014 Proceedings of the on Future of Software Engineering, pp. 212–224 (2014)
4. Duan, Y., Fu, G., Zhou, N., Sun, X., Narendra, N.C., Hu, B.: Everything as a
service (xaas) on the cloud: origins, current and future trends. In: 2015 IEEE 8th
International Conference on Cloud Computing, pp. 621–628. IEEE (2015)
5. Duval, E.: Attention please!: Learning analytics for visualization and recommen-
dation. In: Proceedings of the 1st International Conference on Learning Analytics
and Knowledge, pp. 9–17. ACM (2011)
6. Gillet, D., De Jong, T., Sotirou, S., Salzmann, C.: Personalised learning spaces
and federated online labs for stem education at school. In: Global Engineering
Education Conference (EDUCON), 2013 IEEE, pp. 769–773. IEEE (2013)
10
https://ieee-a.imeetcentral.com/1876public/.
zamfira@unitbv.ro
884 W. Halimi et al.
7. Halimi, W., Jamkojian, H., Salzmann, C., Gillet, D.: Enabling the automatic gen-
eration of user interfaces for remote laboratories. In: 2017 14th International Con-
ference on Remote Engineering and Virtual Instrumentation (rev). IEEE (2017)
8. Lowe, D.: Mools: massive open online laboratories: An analysis of scale and feasi-
bility. In: 2014 11th International Conference on Remote Engineering and Virtual
Instrumentation (REV), pp. 1–6 (2014)
9. Rodrı́guez-Triana, M.J., Govaerts, S., Halimi, W., Holzer, A., Salzmann, C.,
Vozniuk, A., de Jong, T., Sotirou, S., Gillet, D.: Rich open educational resources
for personal and inquiry learning: agile creation, sharing and reuse in educational
social media platforms. In: 2014 International Conference on Web and Open Access
to Learning (ICWOAL), pp. 1–6. IEEE (2014)
10. Saliah-Hassane, H., Reuzeau, A.: Mobile open online laboratories: a way towards
connectionist massive online laboratories with x-API (c-MOOLS). In: 2014 IEEE
Frontiers in Education Conference (FIE) Proceedings, pp. 1–7. IEEE (2014)
11. Salzmann, C., Govaerts, S., Halimi, W., Gillet, D.: The smart device specification
for remote labs. In: 2015 12th international conference on Remote engineering and
virtual instrumentation, pp. 199–208. IEEE (2015)
12. Salzmann, C., Halimi, W., Gillet, D., Govaerts, S.: Deploying large scale online labs
with smart devices. In: Cyber-Physical Laboratories in Engineering and Science
Education (under review) (2017)
13. Siemens, G., d Baker, R.S.: Learning analytics and educational data mining:
towards communication and collaboration. In: Proceedings of the 2nd International
Conference on Learning Analytics and Knowledge, pp. 252–254. ACM (2012)
14. Tawfik, M., Salzmann, C., Gillet, D., Lowe, D., Saliah-Hassane, H., Sancristobal,
E., Castro, M.: Laboratory as a service (LaaS): a model for developing and imple-
menting remote laboratories as modular components. In: 2014 11th International
Conference on Remote Engineering and Virtual Instrumentation (REV), pp. 11–20.
IEEE (2014)
15. UNESCO: What are Open Educational Resources (OER) (2016). http://
www.unesco.org/new/en/communication-and-information/access-to-knowledge/
open-educational-resources/what-are-open-educational-resources-oers/. Accessed
30 Dec 2016
zamfira@unitbv.ro
Present and Future Trends Including
Social and Educational Aspects
zamfira@unitbv.ro
Innovative Didactic Laboratories and School Dropouts
A Case Study
Carole Salis ✉ , Marie Florence Wilson, Fabrizio Murgia, and Stefano Leone Monni
( )
CRS4 – Centre for Advanced Studies, Research and Development in Sardinia, Pula, Italy
{Calis,marieflorence.wilson,fmurgia,stefano.monni}@crs4.it
1 Introduction
Although Sardinia is the 4th Italian Region in terms of high quality education, it also
has the worst record of early school leavers with a rate of 24% [1]. The country’s average
rate is 17.6%. The European target to be reached by 2020 is 10% [2]; Italy’s is 16% [3].
The “Tutti a Iscol@” project draws inspiration from the “Diritti a Scuola” project of the
Apulia Region (Italy), which obtained the Inclusive Growth RegioStar Award in 2015
[4], but the originality of the Sardinian project is the attention paid to the introduction
of Technology in the Innovative Didactic Laboratories (IDLs) used in extracurricular
activities (ECA) [5]. The IDLs were developed based on pupil’s interest for technology.
Emphasis was placed on teamwork, cooperative learning, learning by doing, and social
learning. The project gives great importance to the interaction of schools, local techno‐
logical economic operators (SMEs, Cultural Associations, University Departments) and
the Institutions (RAS: the Autonomous Region of Sardinia, SR: the Regional R&TD
Agency and CRS4, a Multidisciplinary Research Centre).
Among the causes of school dropouts we find: disengagement towards school, absen‐
teeism, lack of interest on learning, reduced level of active participation. The IDLs want
zamfira@unitbv.ro
888 C. Salis et al.
pupils to live a positive experience at school, and develop a positive image of staying
at school. The underlying pedagogical approaches are constructivism and experiential
learning, pragmatic approaches in which learners actively construct their own knowl‐
edge from experience. Since learning requires to think, discuss and to confront others,
attention is also paid to learning by thinking, cooperative learning and problem
solving [6].
The teacher-centred approach can be ineffective in involving pupils in school activ‐
ities. A smart use of ICT can have a positive impact on pupils motivation, turn them to
active learners, and have a positive effect on interaction with peers [7]. The educational
world recognizes the potential of using digital technologies for school dropout reduction
and increasing the education levels. Since the ICT tools are used in all spheres of life,
the European Economic and Social Committee (EESC) believes that a digital approach
within education systems can help to improve the quality of education provided to the
community, particularly if used with common sense [8].
Over the last years, ICT has been one of the most powerful job creation tools. Digital
skills have become a must. This is why schools and technological enterprises should
collaborate to form students that are familiar with the skills required by the digital
economy. Since the Sardinian productive fabric has many technological start-ups and
SMEs we felt important to involve them in this program as this seems to be a significant
international trend. In this paper, they will be referred to as economic operators or
economic actors. See Fig. 1.
The interaction of schools, economic actors and Institutions is crucial as well as the
choice to act through extracurricular activities. Pupils engaged, under supervision, in
enriched learning experiences have less behaviour problems, build self-esteem, develop
positive attitude towards school, and are less likely to dropout [9]. The project requires
teachers to be present during the Lab activities, as a reference point for pupils and to
ensure transfer of technology. Funding for purchasing ICT tools is available to schools,
to enable them to carrying on activities beyond the life of the Labs.
zamfira@unitbv.ro
Innovative Didactic Laboratories and School Dropouts 889
3 Methodology
Seven technological areas, promising for their educational potential and technological
interest were selected (see Table 1). The economic actors were invited to conceive orig‐
inal scenarios for all school levels to be used in the ECA. Scenarios are pupil-centred
technological activities developed based on the pedagogical and technological guide‐
lines given by the Educational Technology team of CRS4. Prior to being included into
the online catalogue for schools to choose from, all scenarios were assessed by a panel
of experts. Ninety original proposals, conceived by 64 economic actors passed the
selection. Since some scenarios could be duplicated, schools had a total of 173 Labs to
choose from. Schools indicated their preferences and an algorithm made the match,
giving a higher priority to schools with higher ranking on the list issued by the Region’s
Department of Education. Ranking was made based on dropout rates and Pisa-tests in
Maths and Italian language and literature.
Pupils participating to the IDLs were chosen based on their special educational needs
that are not necessarily linked to having a handicap, such as lack of motivation, poor
school attendance, behaviour difficulties, poor sustain from their families [5].
The most successful technological areas were those using Augmented Reality
(33.1%) followed by mobile learning with QR codes and NFC technologies (19.7%).
The least successful was that of Treasure Hunt Augmented (4.7%).
4 Anticipated Outcomes
Through the IDLs the school offer will be enlarged, its ICT based tools modernized,
schools and representatives of the local economic market will interact. Pupils should
gradually realize that the technologies used in the IDLs are useful to understand how
new technologies work and they will become able to apply critical thinking to the main
zamfira@unitbv.ro
890 C. Salis et al.
functions of the innovative tools offered by the market. At the end of the 3-year
experience, the identified good practices will be transferred to curricular activities.
5 Monitoring Activities
We monitored the impact of the IDLs on participating students by checking their attend‐
ance, satisfaction and intention to participate to further similar activities. Face-to-face
interviews were carried out in 22 schools (~25% of schools) selected to cover all 7
technological areas, all stages of education, and the whole Sardinian territory. In each
school we interviewed 3 pupils, 1 teacher and 1 technological tutor (representative of
the economic actors who conducted Lab activities). Participation to the online, self-
administered questionnaire was done on a voluntary basis. 53/98 participating schools
replied to the questionnaire (618 pupils, 53 teachers, 42 technological tutors).
Pupil’s feedback: interviewed pupils (56.1% male, 43.9% female) reported to have
enjoyed the IDLs: they had fun, found activities interesting and potentially useful for
their future. Most appreciated socialising with peers and the hands-on approach.
• Appreciation and participation: 100%
• They enjoyed teamwork: >92%
• Felt they improved in solving/resolving problems: >80%
• Reported to have learned new concepts: >92%
• Reported to have learned to plan/design/address: >86%
• Would repeat the experience: >95%
• Would suggest others to participate to similar activities: 100%
Teachers gave a positive feedback on the quality and relevance of IDLs. They
recognized the technological tutors competences to manage the Labs. They noted that
at times, students absent during curricular activities attended the Labs. A significant
number of schools used the opportunity to update their ICT tools, all claimed to be
willing to participate to similar new activities. - See Tables 2 and 3.
Table 2. Teachers feedback on quality of Labs and competences of the technological tutors
Relevance and quality of the Labs Competences of
technological tutors
Excellent Medium Poor No answer High Poor
21/22 0/22 0/22 1/22 22/22 0/22
zamfira@unitbv.ro
Innovative Didactic Laboratories and School Dropouts 891
Table 3. Teachers feedback on pupils participation and on the purchase of ICT tools
Any increase in pupils participation to Lab Did your institute purchase ICT Tools?
activities with respect to curricular activities?
Yes No Yes No
21/22 1/22 17/22 5/22
Tutors reported that pupils were most interested by the technology used in the activities
and that they generally displayed proper behaviour. – See Table 4.
Pupils: More than 50% claim that prior to the ECA, they had heard about the technology
used in the Lab they attended, but that they had not had the opportunity to use it. More
than 1 answer was accepted as pupils could find one activity to be both interesting and
easy to understand or fun and useful, etc.- See Fig. 2. The feedback is positive,
confirming the results of the face-to-face interview.
Tutors: 42/94 tutors filled out the questionnaire. We investigated their evaluation of
both the competences of the teachers (see Table 5) and pupils (see Table 6). Tutors
reported that the presence of teachers during ECA was an asset, and that all pupils
managed to acquire basic skills in the technology used in the Labs. Despite pupils’ varied
zamfira@unitbv.ro
892 C. Salis et al.
educational needs, tutors managed to catch their attention and to establish a good
communication and information flow with them.
Teachers: the most represented school stages of the sample are: High schools, (21/53
teachers), followed by the elementary school (17/53) and last, the junior high school
(15/53). This data underlines teachers interest for introducing technology in education
at an early age. We investigated teachers evaluation of the competences of the techno‐
logical tutors in their interaction with pupils (see Table 7). The feedback is positive, as
the guidelines given for the conception of Lab activities from both the pedagogical and
technological points of view were respected. We also enquired about teachers expect‐
ations (see Fig. 3).
zamfira@unitbv.ro
Innovative Didactic Laboratories and School Dropouts 893
The IDLs are part of a comprehensive strategy to reduce Sardinian pupils school drop‐
outs. Based on the partial positive results, which lead us to believe that the activities
proposed met the needs of all involved parts, the 2nd edition was enriched with 4 addi‐
tional areas: programming humanoid robots, exploring one’s environment with the help
of drones, drawing circuits through the use of conductive ink, fabbing (laser cut, 3D
printing,…) for problem solving.
We also are planning to give the economic actors the opportunity to develop Labs
that can be carried out in remote, mainly for the humanoid robot Laboratories where
schools and pupils can benefit the advantages of a remote Lab in terms of costs, the
absence of time and place restrictions, equipment sharing, and the possibility to share
not only the equipment but ideas and pedagogical scenarios developed by fellow students
at school, share the experience and scenarios developed by other institutions.
References
1. Save the children: Liberare i bambini dalla povertà educativa: a che punto siamo?, May 2016.
https://www.savethechildren.it/sites/default/files/files/uploads/pubblicazioni/liberare-i-
bambini-dalla-poverta-educativa-che-punto-siamo.pdf
2. European Commission: Directorate-General for Communication: Europe 2020: A strategy
for smart, sustainable and inclusive growth. Communication from the Commission,
Bruxelles. http://ec.europa.eu/europe2020/europe-2020-in-a-nutshell/index_en.htm
3. Indire. http://www.indire.it/2016/03/25/dispersione-scolastica-in-italia-abbandono-precoce-
scende-al-15/
4. RegioStars Awards 2015 Category 3: Inclusive Growth - Diritti a scuola (Puglia, IT) (2015).
http://ec.europa.eu/regional_policy/en/projects/italy/tackling-school-drop-out-rates-and-
improving-results
5. Salis, C., et al.: First monitoring results of the “Tutti a Iscol@” Project - a Technology Based
Intervention to Keep Difficult-to-Motivate Pupils in the School Mainstream, system. In:
Proceedings of E-learn: World Conference on E-learning in Corporate, Government,
Healthcare, and Higher Education, pp. 726–731 (2016)
6. Jonassen, D.H.: Learning to Solve Problems: A Handbook for Designing Problem-Solving
Learning Environments. Routledge, New York (2010)
7. Passey, D., Rogers, C.: The Motivational Effect of ICT on Pupils, Research Report 523, with
Joan Machell and Gilly McHugh Department of Educational Research Lancaster University
(2004). http://portaldoprofessor.mec.gov.br/storage/materiais/0000012854.pdf
8. EESC: OPINION of the European Economic and Social Committee on the Communication
from the Commission to the European Parliament, the Council, the European Economic and
Social Committee and the Committee of the Regions on Opening up Education: Innovative
teaching and learning for all through new Technologies and Open Educational Resources
COM, 654 fina - GU C451/26 del 16.12.14 (2013)
zamfira@unitbv.ro
894 C. Salis et al.
9. Massoni, E.: Positive effects of extra curricular activities on students. ESSAI 9 (2011). Article
27. http://dc.cod.edu/essai/vol9/iss1/27
10. Radu, I.: Why should my students use AR? A comparative review of the educational impacts
of augmented-reality. In: Proceedings of IEEE International Symposium on Mixed and
Augmented Reality (ISMAR), pp. 313–314 (2012)
11. Radu, I.: Augmented reality in education: a meta-review and cross-media analysis. Pers.
Ubiquitous Comput. 18(6), 1–11 (2012). doi:10.1007/s00779-013-0747-7
12. Watson, C., Ogle, J.: The pedagogy of things: emerging models of experiential learning. Bull.
IEEE Techn. Committee Learn. Technol. 15(1), 3–6 (2013)
13. Salis, C., Murgia, F., Wilson, M.F., Mameli, A.: IoT-DESIR: a case study on a cooperative
learning experiment in Sardinia. In: Proceedings of International Conference on Interactive
Collaborative Learning (ICL), pp. 785–792 (2015)
14. Pegrum, M.: Mobile Learning: What Is It and What Are Its Possibilities? Teaching and Digital
Technologies: Big Issues and Critical Questions, p. 142 (2015)
15. Kohen-Vacs, D., Ronen, M., Cohen, S.: Mobile treasure hunt games for outdoor learning.
Bull. IEEE Tech. Committee Learn. Technol. 14(4), 24–26 (2012)
16. Lai, C., Salis, C., Murgia, F., Atzori, F., Wilson, M.F.: ANDASA, a web platform for
enhancing network of knowledge and innovation. In: Proceedings of 19th International
Conference on Computer Supported Cooperative Work in Design (CSCWD), pp. 36–41.
IEEE (2015)
17. Fee, S.B., Holland-Minkley, A.M: Teaching computer science through problems, not
solutions. Comput. Sci. Educ. 20(2), 129–144 (2010)
18. Canessa, E., Fonda, C., Zennaro, M.: Low-Cost 3D Printing For Science, Education and
Sustainable Development, free ICTP eBook (2013). ISBN 92-95003-48-9. http://
sdu.ictp.it/3D/book.html
zamfira@unitbv.ro
Intellectual Flexible Platform for Smart Beacons
1
Zaporizhzhya National Technical University, Zaporizhia, Ukraine
galina.tabunshchik@gmail.com
2
Thomas More Mechelen-Antwerpen, Antwerp, Belgium
dirk.vanmerode@thomasmore.be
Abstract. The Smart Beacon System, is a system with Bluetooth Low Energy
devices, a back-end database with dedicated content management system (CMS),
provides a low-entry, easy to use solution for all creative people to dedicate
specific information to whatever object they desire, being paintings, statues,
shopping windows, garbage bins,… themselves. The Internet of Everything for
Everyone. A solution for dynamical maps storage, for usage in a mobile appli‐
cation, was investigated and an implementation suggested. There is also the idea
of smart interfaces implemented, which displaying content based on the user’s
preferences.
1 Introduction
There is always the need to deliver information to the different visitors, teachers or
students of a university campus and the city these campuses reside in, both for day-to-
day use and for specific events [1]. What can be very helpful when you want to provide
additional information about different objects. In the university’s museum this would
help to add multimedia content and background information about the pieces on display.
For tourism this is useful for the highlights of the village or city of the university location,
in the university campus it shows visitors the way and directs them of the points of
interest, like interesting university labs, exhibitions or simply to the canteen.
The development of an attractive user-friendly mobile application with dedicated
back-end server with a content management system (CMS) is extremely relevant in this
view as it allows to bring an effective solution to the market at limited cost and entry
difficulties [2, 3]. The app-user is the client of information related to a certain beacon at
a certain location and our solution allow user to get this information in an attractive way
on his or her smart phone through a dedicated application. The app itself fetches the
information from the server, related to the unique user ID (UUID) the beacon broadcasts
on regular basis. On this server the information is added and edited by the beacon owners
through the developed CMS. The users can decide on groups of beacons which are
allowed to display their information. The total system, CMS-server-app, we refer to as
the Smart Beacon System.
zamfira@unitbv.ro
896 G. Tabunshchyk and D. Van Merode
After implementation of the Smart Campus solution [4], and analysing user feedback
our task was to add functionality to the Smart Beacon system which will allow adding
navigation functions, media content and an intellectual interface.
Improvement of the system and research were conducted in four domains:
1. The CMS should provide managing of maps development and storage it in various
ways.
2. There should be possibility to support diversity of media content attributed to one
beacon.
3. The mobile application should provide the search option to find the optimal path to
the selected beacon location.
4. The mobile application should provide an intellectual interface, which allows
selecting information based on user preferences.
One of the tasks after implementation of the Smart Beacon System was the development
of a navigation system, which can be used for indoor locations. The idea is to find the
location from one beacon to the others, for an interactive tour around the campus or to
guide visitors to their specific location of interest. To provide navigation, first a map of
zamfira@unitbv.ro
Intellectual Flexible Platform for Smart Beacons 897
the building should be provided or developed. Next is showing the appropriate path to
another beacon location. This is why the newly developed solution consists of two parts:
a map editor and path detection.
The map editor allows creating a map of a floor. You can use a background picture
of a known area or develop it from scratch with the easy-to-use editor. Next beacons
can be located at the necessary places (Fig. 1). Figure 1 shows an edited floorplan, with
beacons placed at specific rooms on this floor.
Four editors were created and tested on efficiency. Follow the architecture of built
editors with time for rendering/rendering in conjunction:
• React + Redux + HTML table [7] (5300–6400 ms/1000–1500 ms);
• React + Redux + SVG element [8] (5100–6200 ms/1000–1400 ms);
• React + Redux + Konva.js [9] (700–900 ms/15–20 ms);
• React + Redux + Pixi.js [10] (300–500 ms/5–15 ms).
All these options were tested on response time between demand for the map and
display on the app. In these tests, only the last solution “React + Redux + Pixi.js” showed
the requested delay time of 5–15 ms for feedback.
The maps belonging to a range of beacons are stored as a picture and as an array
[100,100], which is used for the mobile application. In the application they are used for
path detecting tasks.
When a beacon is detected in the mobile application, the client-user gets the infor‐
mation belonging to this beacon. He is notified with a buzz or a ping. This information
consists of a text, a picture and a link, which can house any additional multimedia infor‐
mation which might be needed (Fig. 2).
After you got buzzing from a certain beacon you can ask to show it on the map
(Fig. 3). This map gives the user a sense of where he is located and where other beacons
zamfira@unitbv.ro
898 G. Tabunshchyk and D. Van Merode
can be found. The other ones are given another colour on this map. The user can select
one and ask for the direction to this one.
3 Smart Interfaces
The next improvement was aimed to make user experience better. To goal is to keep the
app-user interested in what a beacon has to say. A good way of doing this is to change
the displayed information over time, to avoid beacon-boredom. In a later development
the display of information would become dependent on external parameters. The amount
of beacon registrations by a certain client-user is obviously a possible parameter, this
way the users get different content, each time he passes a certain location. This way,
certain locations could have much to say. Another parameter could be the date. This will
be very handy to provide “story-telling marketing” to the user. It could present the user
with the desire to pass the location each day of the story-week. Another application is
the week menu displayed at the cafeteria, which is also dependant on the date. In exten‐
sion the beacon could say something different each day of the year. Other parameters
could be user-specific, such as gender or frequented location, although privacy consid‐
erations are in place here. A last set of parameters could come from online resources
such as the weather. And of course a set of the above mentioned parameters could
determine the eventual information available. A woman, that passed the cafeteria for
the 5th time on a cold day, maybe cares for a cup of soup (Fig. 4).
If users do get bored by one beacon, they can put a beacon in a blacklist, which
prevents it from displaying any other information. Of course this operation can be
undone.
Another feature for augment intelligence in the presentation is the “favourites”
option. The user can decide to “like” a beacon, when pushing the heart symbol. This
puts the UUID of this beacon in a local database on the smartphone. This adds the
zamfira@unitbv.ro
Intellectual Flexible Platform for Smart Beacons 899
Fig. 4. Multi-content
possibility to receive updated information even when the user is not in the range of the
liked beacon (Fig. 5).
Fig. 5. Favourites
The options for multi-content, maps, blacklist and liking of beacons are imple‐
mented. The different contents of a beacon can be selected from within the application,
same goes for finding the next beacon within range.
Future work will consist of automating the process of displaying certain information
to certain users, adding artificial intelligence. The liked beacons should also be stored
in a local database. Other improvements could be to allow and automatic display multi‐
media content in the CMS and app. For now the app starts multimedia content auto‐
matically if the link points to this type of content. A last feature would be to allow adding
and editing beacons in the CMS through an additional application.
4 Conclusions
Suggested platform allows different application domains depending from its preferred
functions. It can be used as simple advertisement application, which differs from the
zamfira@unitbv.ro
900 G. Tabunshchyk and D. Van Merode
References
1. Tabunshchyk, G.: Flexible technologies for smart campus. In: Van Merode, D., Tabunshchyk,
G., Patrakhalko, K., Goncharov, Y. (eds.) Proceedings of XIII International Conference on
Remote Engineering and Virtual Instrumentation (REV 2016), Madrid, Spain, 24–26
February 2016, pp. 58–62. UNED (2016)
2. Tabunshchyk, G., Van Merode, D., Goncharov, Y., Patrakhalko, K.: Smart-campus
infrastructure development based on BLE4.0. J. Electrotechn. Comput. Syst. 18(94), 17–20
(2015)
3. Van Merode, D., Tabunshchyk, G., Goncharov, Y., Patrakhalko, K.: Staroverov VInteractive
university platform. Cyчacнi пpoблeми i дocягнeння в гaлyзi paдioтexнiки,
тeлeкoмyнiкaцiй тa iнфopмaцiйниx тexнoлoгiй: тeзи дoпoвiдeй VIII Miжнapoднoї
нayкoвo-пpaктичнoї кoнфepeнцiї (21–23 вepecня 2016 p., м. Зaпopiжжя). – Зaпopiжжя:
ЗHTУ (2016). 344 c
4. Van Merode, D., Tabunshchyck, G.: Multipurpose smart beacon solution. In: Proceedings of
International Symposium on Ambient Intelligence and Embedded Systems (2017)
5. Cochran, D.: Twitter Bootstrap Web Development [Teкcт]. Packt Publishing, Birmingham
(2012). 68 pages
6. Osmani, A.: Learning JavaScript Design Patterns. O’Reilly, Sebastopol (2012). 188 pages
7. Redux website. http://redux.js.org/
8. SVG introduction. https://www.w3.org/TR/SVG/intro.html
9. Konva.js - 2d html5 canvas library for desktop and mobile applications. https://
konvajs.github.io/
10. The HTML5 Creation Engine. http://www.pixijs.com/
11. Desire project Website. http://www.tempus-desire.eu/
zamfira@unitbv.ro
An Approach for Implementation of Artificial Intelligence
in Automatic Network Management and Analysis
Avishek Datta ✉ , Aashi Rastogi, Oindrila Ray Barman, Reynold D’Mello, and
( )
Omar Abuzaghleh
Department of Computer Science, University of Bridgeport, Bridgeport, CT 06604, USA
adatta@my.bridgeport.edu, oabuzagh@bridgeport.edu
Abstract. The fast development of the Internet and the huge number of gadgets
connected to it has immerged with the challenge of focusing on the requirement
for further computerization and automation in the system administration and
architecture of network field. It is becoming increasingly difficult for businesses
to afford a downtime in their Internet system and it is also very important to make
configuration management and error reduction easy [1]. Some of the tasks which
are involved in the management of network configuration include maintaining
configuration files, building a standard for all the device maintenance, repair,
replacement and upgrades, issue proper rollback commands and have a proper
backup archive [2].
There are certain assumptions, rather, wrong assumptions for networks, as
this leads to complicate structure for managing the overloaded content with a very
high time consuming nature. Thus, automated networks control the data or
packets around the heterogenous networks, providing higher response time for
better communication services [3].
Automatic management control permits to proficiently utilize the connection
and hub limits. A second point of interest is adaptable control when burden
condition changed quickly amid times of over-burden and hardware failures.
Automatic management network has been developed and has been improvised
for maintaining the flow control, proper memory utilization and improving the
routing capacity [6].
1 Introduction
According to Ren and Li, “A Network Management System (NMS) refers to a collection
of applications that enable network components to be monitored and controlled” [7].
In some cases, it becomes very difficult to manage and control a particular network
when it gets very large in size. In these cases, the need for an Automated Network
Management (ANM) System is required. This system helps in handling and management
of complex Internet Systems. The ANM system provides for certain tools consisting of
diverse network entities. It can help in the reduction of maintenance and operation
costs [7].
zamfira@unitbv.ro
902 A. Datta et al.
The features of ANM defined among the ISO standards community include Fault
Management, Configuration Management, Accounting, Performance Analysis, Security
and Resource Management/User Directory. The goals include managing diverse
network entities, providing assistance in analysis of the network to the user/operator,
provide a user-friendly interface to the operator, minimization of internet traffic, allow
progressive growth of network and prevent faults from occurring in the networks [7].
2 Related Works
Automated Network Management Systems has been a subject of extensive research and
previous work. A new reach in the automatic network management is the Semantic Web
Services based systems along with the formation OWL-S apart from OWL+SWRL in
addition to the integration with policy-based network management [8].
A definitive long term objective is to make a useful foundation that empowers the
system to end up self-arranging so that fundamental system wide re-configuration is
started by the change inside of the framework itself so, not by the organized activities
of groups of human administrators [5].
An Automatic Traffic Management Control System has been recently implemented
in networks with Stored Program Exchanged (SPC). The implementation of traffic
control systems allows one to efficiently use link and node capacities in an efficient
manner. A second advantage is dynamic control when peak conditions change rapidly
during the periods of overload and equipment failure [3].
There is a huge amount of work related to policy based management. There are a lot
of space restrictions; hence, we can only provide a general outline of the differences
between our work and those of others. The efforts based on the policy execution displays
confusion between goals and means; it is defined in terms of configuration parameters
of network elements such as rules of firewalls. The rules should not break the policy but
a single implementation of intent [9]. The usage of phrase, ‘Security Policy’ conveys
the intent, but it almost means to execute and enforce the intent, turning this on its head
by having our security policy conveying intent and nothing more. As an essential step
in policy specification is what concerns itself when generating the routing filters based
on lisp-like specification language in a logical framework [9].
A large subset of policy based management research is focused on linguistic issues
of policy. Policy is synonymous with rule sets. Rule based solution is often too simplistic
because network phenomena are highly correlated and implications of remote change
have to be derived by the composition of the other entire configuration that will form
the context. For example, whether a particular Telnet may not be explicitly forbidden
by the policy, but such a Telnet might allow access to an application on that machine
that is forbidden by the policy [5].
Also, there is additionally proposed, the procedure to prepare the framework with
programmed and versatile management capacities. Learned example information coop‐
erates with area information to perform versatile administration undertakings-forecast,
conclusion and control activity. This example learning catches the fundamental system
designs and refines the pre-specified space learning.
zamfira@unitbv.ro
An Approach for Implementation of Artificial Intelligence 903
The advancement of MIB and CMIP is advised. One fundamental issue in admin‐
istration data and framework is to give a window through which administration appli‐
cations can get to the worldwide administration data. Our answer for this is to develop
by rationale programming on the conveyed physical MIBs. Accesses to perspectives are
changed into CMIP instructions to MIBs [10].
3 Proposed Model
The management and analysis of any kind of network management system is very
difficult. It is primarily due to the fact that a lot of decision taking and rerouting is
required in order to efficiently reroute and send any data packet towards its destination.
Digital Logistic is one of the most primary and prominent filed of research in today’s
world where the data flow is in Terabytes every instant of time.
Looking from a Digital Network point of view, when looking at a Production Plan‐
ning and Control assignments from a market/logistic-oriented view, the primary task is
the dynamic coordination if the available resources is with another control over resources
[11]. This means that rerouting of the data from the routers based on the decision by the
tools is very important and time consuming. It is of our opinion that this is generally due
to the excessive flow of data and that the individual nodes are overflowing and are
overloaded with data flowing through it. Logically, distributing the data through all the
network would typically free up the nodes and therefore even out the load over a partic‐
ular region.
Another issue which we found to be very compelling is the use of tools and the
manual configuration of these tools in order to perform dynamic controlling of data.
Configuring each node to control data dynamically is a tedious job for any network
monitor/administrator. In this case, he would have to personally be in charge of config‐
uration and maintenance every time something happens to that particular node or he
would have to program certain protocols/subroutines into the nodes each time there is
any kind of incident such as a power surge, power failure etc. Such a system is very
difficult to manage and administer because of the volume of connected devices/nodes.
Considering the above-mentioned problems, we have to come up with our own
design in the Automated Network Management System. Our system involves the pres‐
ence of an Artificial Intelligence System in order to control the efficient routing [12]. In
our system, we use the tools of dynamic control systems already present within the
systems to learn the flow patterns and then use that kind of learning and adaptive nature
of the artificial intelligence system to find out he nature of data flow and also control the
propagation space of the entire dataset. We also use the concept of Field Programmable
Gate Arrays (FPGAs) in order to divide the data in small arrays which would virtually
make each bit of small enough to be transmitted in real time and without any kind of
delays or lags. This in turn would also be effectively transmitted to the network space
upon which the Artificial Intelligence System would be able to work on. The advantage
of using the concept of FPGA at the data level instead of using it at the signaling level
is that the data can be easily manipulated whereas, it is difficult to divide the exiting data
signals.
zamfira@unitbv.ro
904 A. Datta et al.
zamfira@unitbv.ro
An Approach for Implementation of Artificial Intelligence 905
This picture essentially describes the basic architecture of our proposed system. As
mentioned above in the system description, the source transmits the data to the network
cloud. The cloud (Internet-space) contains several AI components in its various
switching devices which are trained on the different data sets and traffic flow patterns.
The AI modules take propagation flow decisions based on the existing traffic patterns
in the different regions of the Network-space. The different packets which had been
divided like FPGAs are usually sent into the Network Cloud and then are re-routed
according to the existing AI algorithm decision.
zamfira@unitbv.ro
906 A. Datta et al.
process can be interrupted by a master controller for a small period of time till the traffic
subsides considerably beyond a threshold, following which the system might, by then
collect some more data from the ongoing network traffic, which is also used to train the
AAIS further. This would make sure that the data is always collected irrespective of the
learning/training status of the AAIS.
3.2 Testing and Analysis of the Active Traffic Data Through Artificial Intelligence
System
The Artificial Intelligence System which will be present in the nodes within the Network
Cloud (Internet Space), once trained, must be able to perform real time traffic analysis
of the active flowing traffic. The data packets once dumped by the source into the
Network Cloud passes through the AI nodes which calculate the instantaneous data
traffic load patterns within its regional or immediate network. Then, it will be able to
detect the path through which there is minimum data traffic at that instant. The AI then
reroutes that fragment of the data through that path. This process is followed till the
packet reaches its destination node. The advantage of having a lightweight, but heavy
processing capability like the KNN gives the AI the advantage of implement ‘n’ number
of hidden layers which would successfully parallelize the job of analyzing the traffic
pattern at an instant.
The instantaneous data traffic analysis by the nodes allows the AI module to instantly
decide the path to be used by the data packet for its next level of the journey. It is true
that data traffic might not be high at every single node which is the reason the entire
bandwidth is utilized, a much faster network can be achieved for the purpose of data
exchange. Another reason for the achievement of speed for this system is the breaking
down of the entire data packet into packets of small size of creating Field Programmable
Gate Arrays (FPGAs).
Once the data is classified by the KNN classifier, the AI module now has the decision
from the AI module about the path to be used by the packet for its travel to its next node.
Once the decision is made and has been confirmed, the node locks down that node for
the data transfer and dispatches the data into that node for the data transfer and dispatches
the data into that node for its travel to the next node. Then the AI module of the previous
node performs its analysis for the transfer of the next data packet. Essentially, the entire
analysis system of the AI undergoes a refresh operation before preparing for the next
data packet. The subsequent data packet undergoes the same operation for its travel.
zamfira@unitbv.ro
An Approach for Implementation of Artificial Intelligence 907
For the sake of experimentation, we consider the system to contain 100 layers of
neurons with each layer containing 50 neurons. For the purpose of this system, we are
avoiding the presence of hidden layers (although that can be implemented as future
work). Since this system is designed to give an answer as to which path to choose there‐
fore, we have to have the same number of outputs as the number of inputs in the first
layer.
According to our proposed model, the learning is done by the ‘K-Nearest-Neighbors’
(KNN) Algorithm which helps the number of comparisons or testing or learning to be
done. This process essentially computes the nodes which are hypothetically close to the
ideal solution/result. This way, the paths in this case, which are not close to the ideal
solution/result are not iterated or computed for that particular node. This does not mean
that that distant node is eliminated permanently from the system. It only means for that
particular iteration/problem, that distant node/path is not valid as it contains more traffic
than the rest of the possible candidates. As mentioned in the previous section, this
process would be running according to an Adaptive algorithm or an Interruptive Algo‐
rithm which would allow and accommodate both the processes of analyzing any new
data along with collecting new data and the process of learning/training the system with
the collected data. The output of the data contains the path to be chosen for transmission
of data in the binary format with the path to be used for transmission being marked as
‘1’ and the other as ‘0’.
The decision-making capability of the system works in tandem with its learning
counterpart. The testing is basically performed at a time on one single packet of data.
Each packet of data is essentially a FPGA of the original complete data which was
dumped into the Internet space from the source. Once the weights of the entire system
subject the data packets to the algorithm for testing purposes and comes up with the
most appropriate path for the data packet that instant. Along with learning within its
learning algorithm so that it can learn from each and every outcome of its decisions to
make better decisions in future.
5 Result
The proposed system is completely theoretical at this point as such a network is not yet
available for testing the proposed model. However, based on the previous work by
various authors, we can say that the proposed system is perfectly capable of working
the way we propose.
Instead of using one particular path to transmit the data from one node to another, it
is better to distribute the data throughout the network and then converge on the particular
node. This not only saves time but also increases efficiency of the entire network. This
process also inadvertently controls and monitors the traffic throughout the network.
6 Conclusion
The above proposed system introduces a new method of effective data exchange, giving
improved speed of propagation thereby reducing propagation time. Proper planning for
zamfira@unitbv.ro
908 A. Datta et al.
the regular capacity of each node must be kept in mind while designing the system. This
system dissolves the role of the administrator having to manually write protocols/
subroutines to deal with any kind of data transactions. While the roles of the network
administrator are reduced, it is not however dissolved. The administrator will still be
needed to service, troubleshoot and debug the system as well as conducting update
operations on the AI without which the system will not be able to identify future styles
of packets. If such a system is truly developed it would certainly automate the entire
process of Network Management and thereby really define Automated Network
Management and Analysis.
Future work in this system can be done in the process of improving the data rate by
the use of frames instead of individual data packets. However, if the frames have packets
going to different destinations, it will create a problem. Another method of optimizing
this procedure would be to parallelize this operation thereby reducing learning costs and
time. A third approach to this problem is to have a Master-Slave configuration where
the major learning and training is done in the Master node while the Slave nodes will
only conduct testing services.
Therefore, to conclude, this paper postulates a method of automating the entire process
of Automatic Network Management System with the help of an Adaptive Artificial Intel‐
ligence System which reroutes the data packets received by it towards the destination
nodes without the intervention of a human entity or a protocol every single time.
References1
1. Qu, Z., Deng, J., Keeney, J., van der Meer, S., Wang, X., McArdle, C.: Pattern mining model
for automatic network monitoring in heterogeneous wireless communication networks. In:
25th IET Irish Signals and Systems Conference 2014 and 2014 China-Ireland International
Conference on Information and Communications Technologies, ISSC 2014/CIICT 2014.
IET, pp. 286–291, June 2014
2. Klie, T., Gebhard, F., Fischer, S.: Towards automatic composition of network management
web services. In: 10th IFIP/IEEE International Symposium on Integrated Network
Management, 2007. IM 2007, pp. 769–772. IEEE, May 2007
3. Filipiak, J.: Analysis of automatic network management controls. IEEE Trans. Commun.
39(12), 1776–1786 (1991). K. Elissa―Title of paper if known,‖unpublished
4. Wang, Y.L., Yuan, A.S., Wu, Q.: Automatic event driven system for network management.
In: 2012 14th International Conference on Advanced Communication Technology (ICACT),
pp. 838–843. IEEE, February 2012
5. Burns, J., Cheng, A., Gurung, P., Rajagopalan, S., Rao, P., Rosenbluth, D., Martin Jr., D.M.:
Automatic management of network security policy. In: Proceedings of DARPA Information
Survivability Conference and Exposition II, 2001, DISCEX 2001, vol. 2, pp. 12–26. IEEE
(2001)
6. Wong, S.Y., Kochen, M.: Automatic network analysis with a digital computation system.
Trans. Am. Inst. Elect. Eng. Part I Commun. Electron. 75(2), 172176 (1956)
1
(A paper of this length is never possible without the help of the contributing authors. It was an
honor to read your ideas and your effort. This is the list of papers which helped us to reach our
goal of making this idea a success).
zamfira@unitbv.ro
An Approach for Implementation of Artificial Intelligence 909
7. Feridun, M., Leib, M., Nodine, M., Ong, J.: ANM: automated network management. IEEE
Netw. 2, 2 (1988)
8. Xu, H., Xiao, D.: Applying semantic web services to automate network management. In: 2nd
IEEE Conference on Industrial Electronics and Applications, ICIEA 2007, pp. 461–466.
IEEE, May 2007
9. King, I.P.: An automatic reordering scheme for simultaneous equations derived from network
systems. Int. J. Numer. Meth. Eng. 2(4), 523–533 (1970)
10. Lin, Y.D., Gerla, M.: A framework for learning and inference in network management. In:
Global Telecommunications Conference, Conference Record, GLOBECOM 1992.
Communication for Global Users. IEEE, pp. 560–564. IEEE, December 1992
11. Nyhuis, F., Pereira Filho, N.A.: Methods and tools for dynamic capacity planning and control.
Gestão Produção 9(3), 245260 (2002)
12. Bai, H.: A Survey on Artificial Intelligence for Network Routing Problems
13. Ng, A.: Sparse autoencoder. CS294A Lecture notes, vol. 72, pp. 1–19 (2011)
14. McCarthy, J.: Artificial intelligence, logic and formalizing common sense. In: Thomason,
R.H. (ed.) Philosophical Logic and Artificial Intelligence, pp. 161–190. Springer, Netherlands
(1989)
15. Acuña, E., Rodriguez, C.: The treatment of missing values and its effect on classifier accuracy.
In: Banks, D., McMorris, F.R., Arabie, P., Gaul, W. (eds.) Classification, Clustering, and Data
Mining Applications. Studies in Classification, Data Analysis, and Knowledge Organisation,
pp. 639–647. Springer, Heidelberg (2004)
zamfira@unitbv.ro
Investigation of Music and Colours Influences
on the Levels of Emotion and Concentration
1 Introduction
From 1875, when was first discovered the existence of electrical currents by Richard
Caton, up to now, when the Mindwave device headset is used to estimate the brain
activity in terms of eSense meter values, there were years of research and studies.
A short overview notes that, in 1924, Hans Berger was first to record the electrical
brain activity and the changes from relaxation to alertness. He called that term EEG
(electroencephalogram). Ten years later, Adrian and Matthew talked about human brain
waves. In time, EEG technology was used in numerous medical sciences, mainly after
Grey Walter discovered the utility of electrodes to record brain activity (in 1964).
Another important step was the discovery of a system determining the eye-gaze
© Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6_85
zamfira@unitbv.ro
Investigation of Music and Colours Influences 911
direction to a certain point by Vidal 1973. And, since then, the neuroscience research
developed increasing understanding of brain activity.
The new developed technology is used to improve the brain activity, to improve
attention/concentration or to induce relaxation. Few examples are given below. Miller
already described EEG as the most common type of electrophysiological indicator used
for workload studies indicating various states or activity levels (Miller 2001).
Robolledo-Mendez proved that the levels of attention are associated to brain activity
mainly to the performance of a learner (Robolledo-Mendez 2008). Peters, based on
user’s behavior analysis, correlated visual to cognitive attention (Peters et al. 2009).
Crowley used psychological computer-based tests to measure meditation and attention
levels explaining the importance of stress in measurements (Crowley et al. 2010).
Furthermore, it was investigated how behavioral suggestions increase the level of
relaxation (Moslow et al. 2011), increase the driver’s brain signals regarding its level of
drowsiness/arousal using EEG (Khalilardali et al. 2012), and are improving players’
control (Gudmundsdottir 2011). A group of researchers from Mexico and Chile worked
on an experiment to determine if a low-cost BCI (Brain Computer Interface) device is
able to measuring the level of concentrating of a programmer during its tasks (Gonzalez
et al. 2015). Girase and Deshmukh found out that through a communication system a
person could use the device command to accomplish some intents. They stated, for
example, that a wheelchair can be totally controlled by human thinking (Girase and
Deshmukh 2016).
During time, there were developed several recording and measuring equipments with
results in investigating the level of emotion, attention and concentration. For example,
(i) BioRadio1 - a wearable medical device with programmable recording and transmission
of different combinations of signals, (ii) EEG Crystal and Crystal-Sleep2 medical devices
still in testing and accreditation that record the heart rate during sleep, (iii) MindWave3 - a
professional set used to measure the levels concentration and relaxation, and (iv) Muse4 - a
brain sensing headband that elevates the meditation practice. No doubt, these devices
connected to computers have become major working tool for professionals. They are
available in the market, easy to use and low-cost. Through BCIs (Brain Computer
Interfaces) we are capable of classifying the levels of attention and concentration, or
relaxation and meditation of individuals during diverse activities.
The Neurosky MindWave (Fig. 1) is a device used “to monitoring electrical signals
generated by neural activity in the brain” (Robbins and Stonehill 2014). It measures the
raw signals, the EEG power spectrum, precisely data regarding the user’s brain waves,
1
https://glneurotech.com/bioradio/physiological-signal-monitoring/wireless-eeg-research-analysis-
teaching/.
2
https://clevemed.com/.
3
http://store.neurosky.com/pages/mindwave.
4
http://www.choosemuse.com/.
zamfira@unitbv.ro
912 D. Ursuţiu et al.
and the eSense meters for attention (concentration) and meditation (relaxation). The
device consists of adjustable head band, sensor tip, ear clip, flexible ear arm, battery
area, power switch, sensor arm and inside ThinkGear chipset. The interface of the
MindWave headset as it appears on the screen can be seen below (Fig. 2).
zamfira@unitbv.ro
Investigation of Music and Colours Influences 913
The research focuses on improving the level of attention and concentration of the
students enrolled in distance education programmes. In order to obtain the best out-
come we use external factors such as music and colours noting the weak and the strong
points of the approach, and ways of dealing with the weak ones. It is structured as
follows:
• Preliminary analyses to encounter that the MindWave device could be used in
measuring the levels of emotion and concentration.
• Indentifying data, both data used in measurements and data about the individuals -
subjects of the research.
• Measuring results and LabVIEW model developing.
• Concluding over the results of the experiment.
The experiment started from understanding the brainwaves and the methods of
manipulating those according to the needs. Dr. Jeffrey Fannin explained the benefits of
increasing the brain waves, the unhealthy ways of increasing gamma brainwaves and
the problems associated with excessive increasing of those.
The analysed data is the result of several sessions of music listening involving a
number of ten persons of different ages, backgrounds and health statuses. In a working
environment, proper for mental activity, but also relaxing and free of environmental
stressors, the subjects/users were listening to diverse genres music from classical to
rock, from dance music to listening music, both instrumental and vocal, knowing that
zamfira@unitbv.ro
914 D. Ursuţiu et al.
the brain is organized to handle the aspects of music such as melody, harmony, rhythm
and timbre in the perception and cognitive processes. Music (a number of ten songs)
had been chosen carefully so it not only be preferred by the user and listened in a daily
basis. It was used the sample of music on a sample group, repeating the experiment
twice. The results encourage us to pursue to the actual experiment.
The measurements were done carefully, once at the time. To each song were
allowed 3 to 4 min. The user sat down working or/and listening. During the music
therapy session the user showed an increase in the level of concentration or quite
contraire in the level of relaxation, depending on the music genre he listened to. Once
the song ended, on the screen appeared the brainwave frequency type describing ranges
of activity, due to eSense calculation (Girase and Deshmukh 2016). The results were
saved for a further comparison. Note that there is measured the mental activity not the
physical one. Psychological factors such as mood or tiredness are being considered.
A first comparison between users showed a variety of data. The interpretation of the
results focused on the outcome resulted from the audition. For example, listening to the
beginning of the first movement of Beethoven’s Symphony no 5 presented diversity,
while De Lucia “Concierto Aranjuez” and Laura McKennit’s “Tango to Evora” showed
consistency, the same answer for all listeners.
In Table 2 we give few examples of the results, mentioning that we had chose the
most similar results appearing among listeners.
In the second experiment the listeners used BT headphones, and the third time we
repeated the same auditions, but this time a coloured panel was lightning the room. The
results showed that while orange and yellow do nothing to change the brain activity,
and blue and purple show a slightly change in the direction of relaxation, bright green
shows an improvement of about 20% increase of attention to the most subjects. This
translates in a higher level of concentration; therefore, we obtained higher work and
educational activity from just placing a coloured panel nearby, fact useful for the
students’ productivity and efficiency in studying from home. With respect to the use of
headphones, we can observe a slightly change in the level of concentration.
zamfira@unitbv.ro
Investigation of Music and Colours Influences 915
Table 3 presents the example of the same person’s measurements in all three
situations while was reading in the room: listening to music on speakers, listening to
music on headphones, and listening to music while the external factor “bright colour”
was exposed. We mention that between the experiments we had pauses, and the
subjects were exposed to natural light. Also, we keep the same list of songs played in
the same order. Could be seen that the brainwave frequency is changed by the use of
headphones, and so it is by being exposed to the green bright colour.
Table 3. Example of frequency ranges variation to one listener in all three situations
Person VII Person VII with Person VII
headphones (headphones + colour)
Yellow = Low Alpha Orange = Theta Yellow = Low Alpha
Red = Delta Red = Delta Orange = Theta
Orange = Theta Orange = Theta Orange = Theta
Orange = Theta Orange = Theta Yellow = Low Alpha
Yellow = Low Alpha Bright purple = High Orange = Theta
Gamma
Red = Delta Light Blue = Low Beta Orange = Theta
Orange = Theta Green = High Alpha Yellow = Low Alpha
Light Blue = Low Beta Green = High Alpha Orange = Theta
Yellow = Low Alpha Green = High Alpha Light Blue = Low Beta
Bright purple = High Gamma Orange = Theta Orange = Theta
The experiment was done in the University’s lab, and although the lab it has phonic
isolation, it is not free of noise. Thus, before repeating the experiment introducing
headphones and light, we were interested in listeners’ reaction to noise.
In Table 4 could be seen the associations between a certain noise and the type of
frequency and associated colour. We briefly explain the difference between the pink
and the white noise, both mentioned in the table above. The pink like noise (of 1/f
frequency) is the frequency considered to be the most relaxing frequency and occurs
widely in nature (e.g. the sound of the sea). On the other hand, the white noise is
created by a range of frequencies uniformly distributed, used to increase sensitivity to
regular surrounding sounds, or to cover background noises.
Laboratory Virtual Instrument Engineering Workbench (LabVIEW) is a system-
design platform and development environment developed by National Instruments that
integrates the creation of user interfaces (Panel) into the development cycle. LabVIEW
programs-subroutines are termed virtual instruments (VIs). Each VI has three com-
ponents: a block diagram (Diagram), a front panel (Panel), and a connector panel.
More LabVIEW developed not only to bring together data acquisition, analysis,
and logical operations, but also to understand how the gathered data has being mod-
ified, being offered for a variety of operating systems. Because its concurrent language,
it could be easily programming multiple tasks of grand benefit for test sequencing and
zamfira@unitbv.ro
916 D. Ursuţiu et al.
Table 4. The association between the noise and the type of frequency
Noise Type of frequency and associated colour
1/f noise Green = High Alpha
1/f noise + green light Bright purple = High Gama
White Noise Yellow = Low Alpha
White Noise + green light Yellow = Low Alpha
data recording. We used the application to gather data from different sensors (followed
by a statistical calculus) and to customize the analysis of the recorded signals.
Therefore, in the presented work, based on the literature references5,6 we developed
one special LabVIEW interface able to analyze the MindWave Mobile device Brain-
Wave signals on selected people exposed in different conditions of music immersion:
selected type of musical compositions, different type of noise and special active music
therapy sounds (composed music).
The LabVIEW application Panel and Diagram can be seen in Fig. 3. It presents the
statistics data on BrainsWave signals instantaneously (on BrainWave monitor) and
final statistics after selected time of listening (on BrainWave statistics) to the actual
selection of signals (music, noise, etc.).
5
MindWaveLabWiew driver: http://forums.ni.com/t5/NI-Labs-Toolkits/NeuroSky-LabVIEW-Driver/
ta-p/3520085.
6
EEG – Electroencephalography with LabVIEW and MindWave Mobile: http://cerescontrols.com/
projects/eeg-electroencephalography-with-labview-and-mindwave-mobile.
zamfira@unitbv.ro
Investigation of Music and Colours Influences 917
This preliminary LabVIEW application will be the base of one future multimedia
development able to: select and start the music and record and analyse the effects of
music therapy. We intend to add also to selected specific music some special lighting
condition (controlled from the application) and record and improve the students
capacity to learn on their homes based on e-LEARNING technologies (courses and
remote and virtual laboratories).
The main advantages offered by this LabVIEW application reside in the fact that it
facilitates the fast addition of new type of calculations, statistics, graphical presenta-
tions of signal and results. This flexibility it is necessary when we investigate a new
field, such as the field of music and colour therapy, in connection with the increased
learning capacity. In our case we plot the BrainWaves medium values (Delta, Theta,
Low Alpha etc., see Fig. 1) with reported percent relative to the highest level (con-
sidered 100%) and in the same times the maxim and medium values of attention and
meditation index. All this parameters were calculated and plotted with the intention to
see and monitor the effects of music therapy.
4 Conclusion
The anticipated results describe ranges of activity for each person in a comparison to
one another for the same stimulus, and the response of each person to different stimuli
(in our case music). The identified range of frequencies relates to particular mental
states and varies from delta to high gamma.
The report of the MindWave headset indicates through its eSense algorithms the
wearer’s mental state, and information about the brainwave frequency bands outputting
the EEG power spectrum. The influence of colour is visible, therefore, due to the
MindWave device measurements and LabVIEW analysis we conclude that brain sig-
nals of the users vary and could be controlled, being a function of the user’s health, his
taste in music, and/or his perception of colours, specific stimuli that determine
emotions.
The level of attention (concentration) could be increased for the benefit of intel-
lectual activity and we think will be one great addition in distance learning education
(eLEARNING programme).
Although our research is an evidence for positive results regarding the effects of
music and colours on the level of concentration and emotions, we continue the research
for further stronger findings. Though, it will be necessary to improve the developed
LabVIEW application, meaning to integrate it in one multimedia platform and to
perform a deeper statistical analysis together with a higher number of recorded signals
from other various types of sensors.
References
Crowley, K., Sliney, A., Pitt, I., Murphy, D.: Evaluating a brain computer interface to categorize
human emotional response. In: The 10th IEEE International Conference in Advances
Learning Technologies (2010)
zamfira@unitbv.ro
918 D. Ursuţiu et al.
zamfira@unitbv.ro
Framework for the Development of a Cyber-Physical
Systems Learning Centre
1 Introduction
zamfira@unitbv.ro
920 D. Centea et al.
Recent concepts such as the Internet of Things (IoT), cloud-based manufacturing and
smart manufacturing address a vision of digitally enabled production commonly
subsumed by the concept of Industry 4.0 [4]. In Industry 4.0, dynamic engineering
processes enable fast changes to production and respond flexibly to disruptions and
failures. In addition, Industry 4.0 is expected to address challenges such as resource and
energy efficiency.
Although many technologies included in the Industry 4.0 concept, like information
and communication technology, are available for use in industry, the employees are
generally not prepared for a successful implementation of Industry 4.0. Prinz et al.
demonstrate that learning factories can make a substantial contribution toward the
understanding of Industry 4.0 [5]. They present a variety of learning modules that enable
participants to transfer learned knowledge directly to their own workplace.
A key element of Industry 4.0 is Additive Manufacturing (AM) which “has evolved
over the past three decades and has been nothing less than extraordinary. AM has expe‐
rienced double digit growth for 18 of the past 27 years … to a market that was worth
over $4 billion in 2014. The AM market is expected to grow to more than $21 billion
by 2020” [6].
zamfira@unitbv.ro
Development of a Cyber-Physical Systems Learning Centre 921
Wank et al. noted that the complexity and effort for developing, implementing and
managing production systems that implement new technological trends will increase
continuously. They observed that many companies in the mechanical engineering and
plant engineering field “view Industry 4.0 with caution and skepticism. Therefore, it is
crucial that the benefits of these developments are demonstrated and evaluated. This
situation causes an urgent demand for research and learning facilities to offer new work‐
shops, trainings and other events to target the specific needs and production environ‐
ments.” The authors present the project “Effiziente Fabrik 4.0”, started in 2014, that
involve 12 company partners and two research institutes. The results of the project were
the design and implementation of Industry 4.0 concepts in the process-learning factory
“CIP” at TU Darmstadt [8].
In 1994 the U.S. National Science Foundation awarded a grant to develop a “learning
factory”. This term, initially used in this grant, was referred to interdisciplinary hands-
on senior engineering design projects with strong links and interactions with industry.
A partnership of Universities in the USA collaborated in the development of practice-
based curriculum able to provide an improved educational experience [9].
Fulfilling individual customer demands with affordable products requires flexible
and adaptable production processes [10]. These forms of production control and flexible
manufacturing increase the complexity of production systems. Current automation solu‐
tions cannot face these challenges. To meet these challenges and prepare future engi‐
neers for related issues, several universities have developed learning factories that deal
with Cyber-Physical Production Systems. Gräßler et al. [10] list some of the learning
factories developed in Germany and briefly describe the most advanced ones.
Industry needs graduates who have enough flexibility to become ideal employees.
According to Schreiber et al. a learning factory is a learning environment that can provide
this flexibility. Students can make practical experiences with the industrial a reality in
an experimental environment, and are trained to successfully handle unanticipated
zamfira@unitbv.ro
922 D. Centea et al.
complexity. They are encouraged to experiment and learn from their mistakes without
penalty. This helps them to keep their curiosity and flexibility [11].
Abele et al. observed that the use of learning factories has recently increased, partic‐
ularly in Europe, and has taken many forms of facilities varying in size and sophistication
aiming to enhance the learning experience of trainees in one or more areas of knowledge
[12]. However, the term learning factory covers a variety of learning environments.
Although no learning factory resembles another nor are they used in the same way,
several of the implementations developed in the last 10 years are generally used for
research, education, and training of industrial employees.
A typical learning factory includes a series of learning instruments used for training
students and people from industry. A question that often arises is the selection of the
processes and instruments that provide the best training for various stakeholders. Plorin
et al. define a conceptual framework termed “advanced Learning Factory (aLF)”. They
identify the major modules, the module configuration, major interaction modes and
transfer mechanisms in a generic learning factory. The aLF-framework provides the
processes necessary to define and run the training instruments installed in their learning
factory according to the demands. After delivering their training several times, they have
found that hands-on interactions foster a memorable learning process [13].
Tisch et al. consider that a “learning factory must be based on a didactic-technolog‐
ical approach, which supports the development of self-organized acting”. They propose
a Learning Factory Curriculum Guide that “offers a systematic approach to design
action-oriented, competency-based Learning Factories” [14].
A learning factory does not have to be a unique lab. The learning factory developed
in Germany at TU Braunschweig [2] includes three parts: Research Lab, Experience
Lab and Education Lab. The Research Lab is mainly focused on research and industrial
projects, the Education Lab focuses on occupational training, while the Experience Lab
is utilized as a learning and training area for university student and professional training
and education.
The achievement of Industry 4.0 requires the implementation of three key features [1]
• Development of inter-company value chains and networks through horizontal inte‐
gration
• Digital end-to-end engineering across the entire value chain of both the product and
the associated manufacturing system
• Development, implementation and vertical integration of flexible and reconfigurable
manufacturing systems
The CPS development presented in this paper includes these three key features as
fundamental elements of an Industry 4.0 implementation, as shown in the centre of
Fig. 1. This implementation is achieved through several foundation technologies
presented on the circle on Fig. 1.
zamfira@unitbv.ro
Development of a Cyber-Physical Systems Learning Centre 923
zamfira@unitbv.ro
924 D. Centea et al.
zamfira@unitbv.ro
Development of a Cyber-Physical Systems Learning Centre 925
4 Learning Facilities
zamfira@unitbv.ro
926 D. Centea et al.
zamfira@unitbv.ro
Development of a Cyber-Physical Systems Learning Centre 927
A crucial feature of the learning factory will be the use of tools based on web tech‐
nologies such as web sockets and HTML5 for system integration and remote access of
physical resources, design tools, data for data analytics, alerts and alarms, and so on.
The IoT data and components are accessible using the internet and mobile devices.
A physical layout of the learning factory is given in Fig. 6. The major components
are; metal 3D printer, 3D printer for plastics, 5-axis CNC machine, collaborative robots,
AGVs, various specialised stations that include post-processing, joining, marking/
labeling and assembly, and a packaging station.
SEPT offers seven undergraduate programs. The CPS Learning Centre will offer
each program the possibility to develop applications related to smart systems, as follows:
• Automation: IoT, smart factory, building automation, electronics
• Automotive: vehicle to vehicle technology (V2V), vehicle to everything technology
(V2X), electric vehicles, autonomous and connected vehicles
• Biotechnology: bioinformatics for health
• Manufacturing: additive manufacturing, smart factory
• Civil & Infrastructure: smart transportation, smart buildings and structures
• Software: networking and infrastructure, Big Data, Data Analytics, security, web
programming
• Energy: alternative energy, smart grid
The Learning Centre will include the following facilities and resources: Learning
Factory; Mechatronics lab; Robotics lab; Power Systems and Networking Lab; Auto‐
mation and IoT Lab; Software Development Lab; Design Lab; Smart Vehicles Lab; and
Manufacturing Lab.
The Learning Centre will be supported by a series of resources. They will provide
students with a skill set that will prepare them for future jobs. Some of these resources
are listed in Fig. 7.
zamfira@unitbv.ro
928 D. Centea et al.
It is expected that all SEPT undergraduate programs will include in their curriculum the
design and development of smart devices and system components that can be manu‐
factured in the Cyber-Physical Systems Learning Centre. Every year, each program, will
add new curriculum elements to continue to bolster the stated educational goals.
The Centre will enrich the program by adding smart systems components to existing
laboratory experiments. There will be no increase in the number of units or teaching
hours per week for any program. However, grouping the labs into one entity will allow
a much easier integration of the facilities for the intended outcome. For instance, Auto‐
motive students currently use mechanical/automotive labs to build their automotive-
related applications. Meanwhile, most of the capstone project designs include electronic
equipment integration. Labs that are currently used only by Automotive students (with
significant mechanical emphasis) or by the Automation students (with significant auto‐
mation and electronic emphasis) will be used for both Automotive and Automation
students to build the mechanical and the electronic components of their projects.
The CPS Learning Centre will allow graduates to be more marketable and have
employable skills related to multiple engineering disciplines, and will relate directly to
today’s industry needs. The process of implementing Industry 4.0 concepts in the work‐
place is complex and unclear. The graduates will be attuned to these obstacles and
opportunities related to Industry 4.0 and will be sought after by managers in industry.
zamfira@unitbv.ro
Development of a Cyber-Physical Systems Learning Centre 929
The CPS Learning Centre will enrich the undergraduate program by replacing some
existing laboratory experiments with developments of smart systems. Each undergrad‐
uate program is expected to develop smart applications related to their specialization in
the learning factory (e.g. smart vehicle for the Automotive program, smart home and
smart transportation for the Civil program, smart health for the Biotechnology program,
smart manufacturing for the Manufacturing program, and so on).
7 Summary
This paper presents the framework for developing a CPS Learning Centre used to teach
undergraduate and graduate students and to train specialists from industry in Industry
4.0 concepts. It is also expected to be a hands-on training facility that will allow small,
medium and large companies to be competitive in the current market by using modern
manufacturing approaches that implement various levels of digitization and cyber-
physical systems.
The Learning Institute is expected to allow the development of applications that
include elements of smart production, smart vehicles, smart energy, smart connectivity,
smart home and city, and smart health.
References
1. Monostori, L.: Cyber-physical production systems: roots, expectations and R&D challenges.
Procedia CIRP 17, 9–13 (2014)
2. Thiede, S., Juraschek, M., Herrmann, C.: Implementing cyber-physical production systems
in learning factories. Procedia CIRP 54, 7–12 (2016). Tisch, C.H., Cachay, J., Abele, E.,
Metternich, J., Tenberg, R.: A systematic approach on developing action-oriented,
competency based learning factories. Procedia CIRP 7, 580–585 (2013)
3. Berger, C., Hees, A., Braunreuther, S., Reinhart, G.: Characterization of cyber-physical sensor
systems. Procedia CIRP 41, 638–643 (2016)
4. Erol, S., Jäger, A., Hold, P., Ott, K., Sihn, W.: Tangible industry 4.0: a scenario-based
approach to learning for the future of production. Procedia CIRP 54, 13–18 (2016)
5. Prinz, C., Morlock, F., Freith, S., Kreggenfeld, N., Kreimeier, D., Kuhlenkötter, B.: Learning
factory modules for smart factories in industrie 4.0. Procedia CIRP 54, 113–118 (2016)
6. Thompson, M.K., Moroni, G., Vaneker, T., Fadel, G., Campbell, R.I., Gibson, I., Bernard,
A., Schulz, J., Graf, P., Ahuja, B., Martina, F.: Design for additive manufacturing: trends,
opportunities, considerations, and constraints. CIRP Ann. Manufact. Technol. 65, 737–760
(2016)
7. Collaborative Robotics: State of the Market/State of the Art, ABIresearch (2015). https://
www.abiresearch.com/market-research/product/1022012-collaborative-robotics-state-of-
the-market/ABIresearch. Accessed 11 Oct 2016
8. Wank, A., Adolph, S., Anokhin, O., Arndt, A., Anderl, R., Metternich, J.: Using a learning
factory approach to transfer industrie 4.0 approaches to small- and medium-sized enterprises.
Procedia CIRP 54, 89–94 (2016)
zamfira@unitbv.ro
930 D. Centea et al.
9. Hadlock, H., Wells, S., Hall, J., Clifford, L., Winowich, N., Burns, J.: From practice to
entrepreneurship: rethinking the learning factory approach. In: Proceedings of the 2008 IAJC-
IJME International Conference, Paper 081, ENT P 401 (2008)
10. Gräßler, I., Pöhler, A., Pottebaum, J.: Creation of a learning factory for cyber physical
production systems. Procedia CIRP 54, 107–112 (2016)
11. Schreiber, S., Funke, L., Trachta, K.: BERTHA - a flexible learning factory for manual
assembly. Procedia CIRP 54, 119–123 (2016)
12. Abele, E., Metternich, J., Tisch, M., Chryssolouris, G., Sihn, W., ElMaraghy, H., Hummel,
V., Ranz, F.: Learning Factories for research, education, and training. Procedia CIRP 32, 1–
6 (2015)
13. Plorin, D., Jentsch, D., Hopf, H., Mueller, E.: Advanced Learning Factory (aLF) – method,
implementation and evaluation. Procedia CIRP 32, 13–18 (2015)
14. Tisch, M., Hertle, C., Cachay, J., Abele, E., Metternich, J., Tenberg, R.: A systematic approach
on developing action-oriented, competency-based learning factories. Procedia CIRP 7, 580–
585 (2013)
zamfira@unitbv.ro
Applications and Experiences
zamfira@unitbv.ro
The Use of eLearning in Medical Education
and Healthcare Practice – A Review Study
Blanka Klimova(&)
1 Introduction
accessed at any place (provided that learners have access to the Internet), at any time,
and on one’s own pace. It also seems more cost-effective and supports distance edu-
cation [6]. No wonder that it found its place in medical education and healthcare
practice.
There are other eLearning modalities, which are also used in medical education and
healthcare [7]: asynchronous/synchronous audio or video; chat/video conferences;
computer-aided learning; computer based testing; educational online games; electronic
problem-based e-learning; electronic portfolio; online collaboration; online discussions
forums; repository and hypertext; virtual laboratories; and virtual learning environ-
ments. L’Engle, Raney, D’Adamo [8] suggest that eLearning may improve health
service delivery and the reach of health promotion activities all over the world, but
especially in developing countries. In fact, the main reason of using eLearning in
healthcare are to train geographically dispersed workforce, lower costs and higher
learning retention. Moreover, 95% of the respondents said that they exploited custom
designed online modules and 80% of the respondents reported that they used eLearning
courses as part of a blended learning program [9].
The aim of this study is to examine the use of eLearning in medical education and
healthcare practice and discuss its advantages and disadvantages to help deliver better
care for patients and populations.
2 Methods
The methodology of this study is based on the study by Moher, Liberati, Tetzlaff,
Altman [10]. The main method included a systematic review whose goal was to identify
the research studies on the basis of the key words in four databases Web of Science,
ScienceDirect, Scopus, and MEDLINE. This review was performed in the period from
2013 to October 2016 for the following key words: elearning AND medical education
AND healthcare practice. Most of the studies were found in ScienceDirect – 338
studies. In the Web of Science only 3 studies were detected, in Scopus 6 studies and in
MEDLINE 5 studies were identified. Thus, altogether 352 publications were detected in
the databases. The titles of all studies as well as their duplicity were then checked in
order to discover whether they focus on the research topic. 94 studies remained for
further analysis. After that, the author checked the content of the abstracts whether the
study examined the research topic. 31 studies/articles were selected for the full-text
analysis, out of which the findings of 17 research studies were then used in the
manuscript for the comparison of the findings in the part of Discussion, as well as in the
Introductory part to discuss the topic, and only 14 studies could have been then used for
the detailed analysis of the research topic.
The study was included if it matched the corresponding period, i.e., from 2013 up
to October 2016; if it included medical students, doctors, nurses, or other healthcare
practitioners; if it focused on the use of eLearning; and if the study was written in
English. The selection period starts with the year of 2013 since several reviews (e.g.,
[11, 12]) were published on this topic before this period.
zamfira@unitbv.ro
The Use of eLearning in Medical Education and Healthcare Practice 935
3 Results
This review study altogether comprises 14 original research studies. Twelve studies
were descriptive studies, one was prospective educational study and only one was a
randomized trial. Table 1 below describes these studies which were identified on the
basis of the review and their main findings and limitations. The table comprises seven
studies examining the use of eLearning in medical education, six studies exploring its
use in healthcare practice and one study investigating the cost effectiveness of
eLearning. The studies are arranged according to the alphabetical order of their first
author.
Table 1. Overview of the selected original research studies on the use of eLearning in medical
education and healthcare practice
Study Findings Limitations
Blake and Healthcare professionals, healthcare A relatively small sample of the
Gartshore [13] educators and pre-registered healthcare overall population to whom the
descriptive study students had a positive attitude towards online learning was offered;
online learning tools which appeared to geographically limited study
be engaging and improved their
knowledge of important public issues in
workplace
Chong et al. [14] Organizational support seems to be A geographically limited study
descriptive study important to promote accessibility of
information and communications
technology facilities for Malaysian
nurses to motivate their involvement in
e-learning
Corner et al. [15] eLearning appears to be an effective A few methodological
prospective study mode of delivering education in a large limitations
geographical area on the consistent and
reliable use of the Chelsea Critical Care
Physical Assessment (CPAx) functional
assessment tool. In addition, eLearning
modules may have utility as evaluation
tools
de Lazzari et al. The implementation of a software More in-depth evaluation from
[16] descriptive simulator into the e-learning the participants; no sufficient
study environment means opportunities for participants’ confidence with a
accelerated learning; lower costs in simulator
comparison with in vivo experiments;
no medical related accidents; and
increased attention
Elmore [17] eLearning seems to be an effective tool A geographically limited study;
descriptive study with a minimal investment a small number of participants
(continued)
zamfira@unitbv.ro
936 B. Klimova
Table 1. (continued)
Study Findings Limitations
Lahti et al. [18] eLearning helps to transfer adopted A small sample of respondents;
knowledge successfully into practice a short time span
Munch-Harrach Eight podcasts were produced on the A lack of in-depth evaluation
et al. [19] eLearning platform at little expenses.
descriptive study They also contributed to the external
presentation of the faculty
Murphy et al. [20] eLearning platform helped healthcare More in-depth evaluation from
descriptive study professionals to improve the provision the participants; maintenance of
of nutrition and lifestyle advice for the sustainability of current
cancer survivors eLearning platform
Polly et al. [21] eLearning intervention using virtual A lack of objective assessment
descriptive study laboratories has a big potential for the of diagnostic skills
improvement of students’ diagnostic
skills; it is close to reality; and there is a
high level of interactivity and feedback
Reid et al. [22] Medical students may face three key A geographically limited study;
descriptive study obstacle when using eLearning: only qualitative survey
injustice, passivity and lost at sea
Rider et al. [23] eLearning as a tool in online trainings A lack of in-depth evaluation.
descriptive study that model key prevention strategies
may play an important role in translating
policy into improved outcomes
Sissine et al. [24] When using a blended eLearning Different infrastructure between
descriptive study approach, there were significant cost countries; insufficient access to
savings (67%) in comparison with a the Internet in developing
traditional didactic method countries; costs inputs may
vary in a different setting
Thorne et al. [25] eALS (Advanced Life Support) course A geographically limited study
descriptive study is equivalent to the traditional,
face-to-face learning. Moreover, it
increases candidate autonomy; it is cost
effective; it decreases instructor’s
burden and standardization of course
material
van de Steeg et al. elearning course on delirium aimed at A potential delay in the
[26] randomized nursing staff had a positive influence on intervention effect
trial improved delirium care provided by
nurses; it also decreased the number of
older patients diagnosed with delirium
and broaden nurses’ knowledge on
delirium
Source: authors’ own processing
zamfira@unitbv.ro
The Use of eLearning in Medical Education and Healthcare Practice 937
4 Discussion
The findings in Table 1 show that eLearning significantly contributes to the faster
dissemination of knowledge into practice [13–16, 18, 21, 23, 26]. Furthermore, it
enhances inter-professional care, [13, 15, 19, 23, 26] interactivity and autonomy of
students’ learning, [13, 14, 21, 25] cost-effectiveness, [16, 17, 25] and may also serve
as a good evaluation tool [15]. However, as the study by Reid et al. [22] state, students
may face three key obstacles when using eLearning. These include injustice (a sense of
resentment: the idea that they were somehow being ‘done out of’ the education that
they deserved), passivity (students may experience a lack of control – that they were
‘passive recipients’ of eLearning material), and lost at sea feeling (unfamiliarity with
the eLearning approach).
Nevertheless, majority of the findings from Table 1 indicate that eLearning is a
significant tool both for medical education, for which it has been in use since 1990s,
and healthcare practice. George et al. [27] claim that eLearning can be equivalent,
possibly superior to traditional learning for healthcare professionals’ education.
eLearing courses might also improve healthcare practice in terms of better communi-
cation between healthcare team members, enhanced quality of care, and better out-
comes for patients [23].
Although there is an increasing trend to use eLearning in different branches of
medical education, [23, 28] the findings show the prevalence of the use of eLearning
particularly in the studies on nursing education and profession [14, 20, 26]. Further-
more, the research studies emphasize the importance of eLearning for the dissemination
of knowledge, [14, 15, 18, 23] understanding particular health issues, [13, 29] con-
tinuous education, [14] and training of busy healthcare professionals who wish to
access educational programs to maintain or extend their knowledge in response to
service needs. [23, 30].
The most critical drawback of the reviewed studies is that only one study was
randomized controlled trial, other studies were descriptive, which might result in the
overestimated effects of eLearning and biases in these publications and have a negative
impact on the validity of these research studies. Therefore future research should
concentrate on a longer time span randomized controlled studies to verify the efficacy
of eLearning and compare against standard teaching.
5 Conclusion
Overall, the findings of this review study indicate that eLearning seems to be an
effective mode for medical education and healthcare practice. However, as Walsh [31]
suggests, eLearning technologies should be used purposefully and wisely in order to
help deliver better care for patients and populations.
Acknowledgments. This review study is supported by SPEV project 2017, Faculty of Infor-
matics and Management, University of Hradec Kralove, Czech Republic. The author thanks the
SPEV students, especially Josef Toman for his help with the collection of the data.
zamfira@unitbv.ro
938 B. Klimova
References
1. Klimova, B., Poulova, P.: Learning technologies and their impact on an educational process
in the Czech Republic. In: Proceedings of the International Conference on Computer Science
and Information Engineering (CSIE 2015), pp. 429–434. Destech Publications, Lancaster
(2015)
2. eLearning Action Plan (2001). http://www.aic.lv/bolona/Bologna/contrib/EU/e-learn_
ACPL.pdf
3. Klimova, B.: Developing ESP study materials for engineering students. In: Proceedings of
2015 IEEE Global Engineering Education Conference (EDUCON 2015), pp. 52–57. Tallinn
University of Technology, Tallinn (2015)
4. Klimova, B., Poulova, P.: Forms of instructions and students’ preferences – a comparative
study. In: Proceedings of the 7th International Conference (ICHL 2014), pp. 220–231.
Springer, Heidelberg (2014)
5. Frydrychova Klimova, B.: Blended learning. In: Mendez Vilas, A., et al. (eds.) Research,
Reflections and Innovations in Integrating ICT in Education, pp. 705–708. FORMATEX,
Spain (2009)
6. Graham, C.R., Allen, S., Ure, D.: Benefits and challenges of blended learning environments.
In: Khosrow-Pour, M. (ed.) Encyclopedia of Information Science and Technology I-V. Idea
Group Inc., Hershey (2003)
7. Jawaid, M., Aly, S.M.: E-learning modalities in the current era of medical education in
Pakistan. Pak. J. Med. Sci. 30(5), 1156–1158 (2014)
8. L’Engle, K., Raney, L., D’Adamo, M.: mHealth resources to strengthen health programs.
Glob. Health Sci. Pract. 2(1), 130–131 (2014)
9. Nine Lanterns. http://elearninginfographics.com/elearning-in-healthcare-infographic-2/
10. Moher, D., Liberati, A., Tetzlaff, J., Altman, D.G.: The PRISMA Group. Preferred reporting
items for systematic review and meta-analysis: the PRISMA statement. PLoS Med. 6(6),
e1000097 (2009)
11. Fahlman, D.: Educational leadership for e-learning in the healthcare workplace. IRRODL 13
(2), 236–246 (2012)
12. Frehywot, S., Vovides, Y., Talib, Z., Mikhail, N., Ross, H., Wohltjen, H., et al.: E-learning
in medical education in resource constrained low- and middle-income countries. Hum. Res.
Health 11, 4 (2013)
13. Blake, H., Gartshore, E.: Workplace wellness using online learning tools in a healthcare
setting. Nurse Educ. Pract. 20, 70–75 (2016)
14. Chong, M.C., Francis, K., Cooper, S., Abdillah, K.S., Hmwe, T., Sohod, S.: Access to,
interest in and attitude toward e-learning for continuous education among Malaysian nurses.
Nurse Educ. Today 36, 370–374 (2016)
15. Corner, E.J., Handy, J.M., Brett, S.J.: eLearning to facilitate the education and implemen-
tation of the Chelsea critical care physical assessment: a novel measure of function in critical
illness. BMJ Open 6(4), e010614 (2016)
16. de Lazzari, C., Genuini, I., Pisanelli, D.M., D’Ambrosi, A., Fedele, F.: Interactive simulator
for e_learning environments: a teaching software for healthcare professionals. BioMed. Eng.
OnLine 13, 172 (2014)
17. Elmore, J.M.: CEVL e-learning teaches GUMS method to score hypospadias preoperatively
and predict postoperative outcomes. J. Pediatr. Urol. 11(5), 234–238 (2015)
18. Lahti, M., Kontio, R., Pitkanen, A., Valimaki, M.: Knowledge transfer from an e-learning
course to clinical practice. Nurse Educ. Today 34(5), 842–847 (2014)
zamfira@unitbv.ro
The Use of eLearning in Medical Education and Healthcare Practice 939
19. Munch-Harrach, D., Kothe, C., Hampe, W.: Audio podcasts in practical courses in
biochemistry-cost-efficient e-learning in a well-proven format from radio broadcasting.
GMS Z Med. Ausbild. 30(4), doc 44 (2013)
20. Murphy, J., Worswick, L., Pulman, A., Ford, G., Jeffery, J.: Translating research into
practice: evaluation of an e-learning resource for health care professionals to provide
nutrition advice and support for cancer survivors. Nurse Educ. Today 35(1), 271–276 (2015)
21. Polly, P., Marcus, N., Maguire, D., Belinson, Z., Velan, G.M.: Evaluation of an adaptive
virtual laboratory environment using Western Blotting for diagnosis of disease. BMC
MedEduc. 14, 222 (2014)
22. Reid, H.J., Thomson, C., McGlade, K.J.: Content and discontent: a qualitative exploration of
obstacles to eLearning engagement in medical students. BMC Med. Educ. 16, 188 (2016)
23. Rider, B.B., Lier, S.C., Johnson, T.K., Hu, D.J.: Interactive web-based learning: translating
health policy into improved diabetes care. Am. J. Prev. Med. 50(1), 122–128 (2016)
24. Sissine, M., Segan, R., Taylor, M., Jefferson, B., Borrelli, A., Koehler, M., et al.: Cost
comparison model: blended eLearning versus traditional training of community health
worker. Online J. Public Health Inform. 6(3), e196 (2014)
25. Thorne, C.J., Lockey, A.S., Bullock, I., Hampshire, S., Bugum-Ali, S., Perkins, G.D.:
E-learning in advanced life support – an evaluation by the Resuscitation Council (UK).
Resuscitation 90, 79–84 (2015)
26. van de Steeg, L., Ijkema, R., Langelaan, M., Wagner, C.: Can an e-learning course improve
nursing care for older people at risk of delirium: a stepped wedge cluster randomized trial.
BMC Geriatr. 14, 69 (2014)
27. George, P.P., Papachristou, N., Belisarion, J.M., Wang, W., Wark, P.A., Cotic, Z., et al.:
Online eLearning for undergraduates in health professions: a systematic review of the impact
on knowledge, skills, attitudes and satisfaction. J. Glob. Health 4(1), 010406 (2014)
28. Jayakumur, N., Brunckhorst, O., Dasgupta, P., Khan, M.S., Ahmed, K.: eLearning in
surgical education: a systematic review. J. Surg. Educ. 72(6), 1145–1157 (2015)
29. Kelly, C., Reid, E., Lohan, M., Alderdice, F., Spence, D.: Creating an eLearning resource to
improve knowledge and understanding of pregnancy in the context of HIV infection. Int.
J. Environ. Res. Public Health 11(10), 10504–10517 (2014)
30. Delf, P.: Designing effective eLearning for healthcare professionals. Radiography 19(4),
315–320 (2013)
31. Walsh, K.: The future of e-learning in healthcare professional education: some possible
directions. Ann. Ist. Super. Sanita. 50(4), 309–310 (2014)
zamfira@unitbv.ro
Efficiency and Prospects of Webinars
as a Method of Interactive Communication
in the Humanities
1 Introduction
The title of the webinar comes from the English words web and seminar. Despite the
name, the software designed for organizing webinars can be used for other forms of
educational process.
Aristotle pointed out that the impact of the speaker on the listener has a specific
purpose, the formation of a certain ideal, aimed at the prosperity of the state, the
development of personality. The learning process ultimately has the same goal.
Every teacher has a certain social order from the society - to educate students in
accordance with the highest standards of education, to lay them as much knowledge
and skills so they can benefit people and themselves.
zamfira@unitbv.ro
942 N.N. Petrova et al.
As a teacher can be viewed in two ways: as a wise mentor and as a person with a
wealth of experience and knowledge and transmitting information “by inheritance” to
his disciples, this scheme is divided into two components.
The first component of the scheme (Fig. 2):
zamfira@unitbv.ro
Efficiency and Prospects of Webinars 943
The first component of the scheme then takes the following form (Fig. 4):
zamfira@unitbv.ro
944 N.N. Petrova et al.
zamfira@unitbv.ro
Efficiency and Prospects of Webinars 945
will be a good platform for their future scientific work. In recent years, webinars are
becoming more common when creating a virtual scientific laboratories, inter-University
creative groups, conducting on-line conferences, student research forums and other
forms of scientific inter-University cooperation of students [5].
Both prior and subsequent work on the webinar requires serious preparation for
teachers and students. But sure it will give positive results. In the light of the ideas of
humanistic psychology and pedagogy, webinars open new possibilities of interaction
between teacher and students. They help to build good relations of cooperation, in
which open and friendly attitude to each other encourages openness, desire commu-
nication in the process of learning, independent thinking, reflection.
Direct “live” communication of the lecturer with the audience is hardly ever fading.
But webinars in our opinion, will occupy an important place in a number of promising
educational technologies in relation to their very simple, time saving and big oppor-
tunities of development of creative abilities of students [2, 3]. In the future, diverse
forms of organization of webinars will further extend and improve. Webinars have a
great development in network education as a form of functioning of network education
platform [4, 10]. They improve access to educational services, promote the develop-
ment of virtual mobility and flexible forms of employment [7].
Also, we conducted a survey in Kazan high schools such as KNRTU KAI, KFU
and KSUAE. The results are shown on the Fig. 6.
As can be seen from the presented data, the share of respondents always attending
webinars is 7.8%. More than a third of students try not to miss the webinars, but it is
not always possible at the time 36.5% and a quarter prefer to work with recordings of
webinars 24.2%, which may be a consequence of inconvenient class schedules. Each
sixth Respondent is not satisfied with the quality or content of the webinars (the total
figure in the total sample, 15.6%).
zamfira@unitbv.ro
946 N.N. Petrova et al.
zamfira@unitbv.ro
Efficiency and Prospects of Webinars 947
The most interesting solutions in the virtual mobility are Adobe Acrobat Connect
and Open meetings. Both technologies provide streaming video to the user’s web
browser and can be used for a wide range of problems related to distance learning
(e-learning). IT systems provide the following areas of eLearning:
• Virtual Classrooms;
• On-line training;
• The performance improvement system based on LMS (Learning Management
System).
Recognition of the possibility of virtual mobility is supported by various interna-
tional projects, which are actively involved in the Russian high schools [6, 8]. Along
with the classic e-learning tools there are some recent tools:
• Creation of special websites about the basic educational resources and mobility
programs, the introduction in educational institutions of the special chat-rooms for
the organizers of the mobility of students, teachers, etc., to discuss and exchange
information;
• E-voting system, designed to create interactive lectures, increase the participation of
students in the discussion of the proposed issue or point of view;
• Peer review system and technology of collective creation and editing (wiki);
• Asynchronous communication tools (forums and blogs);
• Social bookmarking and social networking;
• Podcasting and online lectures, video streaming etc.
All of these technologies are the components of an open virtual platform based on
Web 2.0, that allow you to organize an effective online collective and individual work
and integrate seamlessly into the virtual learning environment, providing the academic
mobility of both students and teachers.
5 Conclusion
The webinar is the most effective method of teaching to the maximum students a
particular topic, as it covers a lot of forms of influence on students. This is in contrast to
traditional forms; audio tools: communicate in real-time and placement of various
materials in the form of audio recordings; video tools: videos and other materials from
the Internet. This is the presentation of the teacher, which he demonstrates in the
Internet environment. It’s messaging not only between teacher and students but also
between students themselves and, of course, is the live chat as a means of exchanging
messages with one click.
Webinar in many ways helps to replace direct communication between teacher and
student, as absolutely to exclude the teacher from the educational process as a person is
impossible. Method of webinar’s work is indirectly-direct, as this online communi-
cation via the Internet may be seen as communication of close, free and personal. The
teacher for the student is actually “here” and “face to face”. Webinars offer the
opportunity to see and hear each other, even at a quite large distance. A conducting of a
webinar with a pre-filled audience with the second teacher there is also seems
zamfira@unitbv.ro
948 N.N. Petrova et al.
perspective. He helps to maintain order and also participates in the webinar that is more
costly for the University, but more useful for students. Two teachers is the ability for
greater control and deeper reasoning and active discussions.
Webinars in virtual mobility of students are also important. They provide access to
international educational resources at the minimum cost. The results of the study show
the growing impact of virtual mobility to improve the quality and accessibility of
educational services.
References
1. Absalyamova, S.G., Absalyamov, T.B.: Remote employment as a form of labor mobility of
today’s youth. Mediterr. J. Soc. Sci. 6(1S3), 227–231 (2015)
2. Apple, M.W., Kenway, J., Singh, M.: Globalizing Education: Policies, Pedagogies and
Politics. Peter Lang, New York (2005). 311 pages
3. Byram, M., Dervin, F.: Students, Staff and Academic Mobility in Higher Education.
Cambridge Scholars Publishing, Cambridge (2008). 320 pages
4. Dall Alba, G., Sidhu, R.: Australian undergraduate students on the move: experiencing
outbound mobility. Stud. High. Educ. 40(21), 721–744 (2015)
5. Petrova, N.N., Sidorenko, L.P.: Challenges and prospects of webinars as online-learning
method. Becтник Чyвaшcкoгo гocyдapcтвeннoгo пeдaгoгичecкoгo yнивepcитeтa им.
И.Я. Якoвлeвa. № 3 (91). C. 134–139 (2016)
6. Sakhapov, R.L., Absalyamova, S.G.: The use of telecommunication technologies in
education network. In: Proceedings of 2015 12th International Conference on Remote
Engineering and Virtual Instrumentation, REV 2015, Bangkok, pp. 14–17 (2015)
7. Sidorenko, L.P., Petrova, N.N.: Art of the lecture in real and virtual space. In:
Psycho-Pedagogical Research of the Quality of Education in the Conditions of Innovative
Activity of Educational Organization: VII All-Russia Scientific-Practical Conference,
1-Slavyansk-on-Kuban, pp. 163–167 (2014)
8. Loveland, E.: Student mobility in the European Union. Int. Educ. 17, 22–25 (2008)
9. Vázquez, L.K., Mesa, F.R., López, D.A.: To the ends of the earth: student mobility in
southern Chile. Int. J. Educ. Manag. 28, 82–95 (2014)
10. The Bologna Declaration of 19 June 1999, Berlin-Bologna-Web-Page. http://www.bologna-
berlin2003.de/pdf/bologna_declaration.pdf
zamfira@unitbv.ro
Port Logistics: Improvement
of Import Process Using RFID
Abstract. This paper describes a new system developed to improve the import
process of steel coils driven into a port terminal in the Port of Bilbao. A new
RFID based system minimizes mistakes in identification of the coils during the
inland movements of goods.
1 Introduction
Logistic has become a defining factor in industry 4.0. Around 90% of the worlds
merchandize and commodity trade is transported by ship. During last years, ports have
made an enormous effort on digitization, developing infrastructure to apply advanced
techniques for collection and analysis of information [1, 2]. However, the investment
cannot be profitable while port operators lack systems to automatically integrate
information about their processes [3].
This paper describes a new system developed to improve the import process of steel
coils driven into a port terminal in the Port of Bilbao. The large number of coils
transported complicates all different tasks associated with its intermediate storage until
they are sent to destination by road or rail. This translates into a significant error rate in
shipments while coil improper handling can cause defects in the goods. The project
developed uses radio frequency identification to validate the various stages in the
import process and provides the technological infrastructure for real-time reporting of
the execution of tasks [4].
2 Problem Description
Although historically the city of Bilbao has a great tradition in the manufacture of steel,
now the crisis of the sector demands the importation of a great number of coils to
supply the industry, mainly automotive, of the north and center of Spain. The Port of
Bilbao is one of the main entrees of the country for steel coils, that are transported by
road or rail to destination, accounting for 15% of total port traffic.
Although the average stay in the harbor of a coil is less than two weeks, some coils
can remain in the port warehouse for several months. During this period, intermediate
movements often take place to facilitate access to adjacent coils. These actions, usually
executed by the terminal staff without notification, lead to a progressive uncontrolled
stock that can cause errors when identifying a specific coil. Main features of coils
consist in material, thickness of turn, weight and width. The inland operations can
cause damages in the 2D codes that turn the identification in a very hard task. It is quite
common that during the final inspection, which is carried out when the coils are loaded
on the trucks or wagons, errors are detected that entail an important loss of time until
the real coils demanded are finally located and loaded. It is therefore necessary to
develop a system to maintain the location of the coils updated and ensure the correct
identification of each coil transported. In addition, one of the main requirements is to
achieve this objective without significantly altering the operation of the terminal staff.
3 Functional Description
The project covers all the activity of the port terminal, guaranteeing the traceability of
the coils since they are unloaded from the vessels until they leave the port. For this
purpose, the project is divided into three fundamental stages: ship unloading, inter-
mediate movements and terminal exit.
zamfira@unitbv.ro
Port Logistics: Improvement of Import Process Using RFID 951
• The operator must board the ship and proceed with a visual inspection of each coil.
The application allows to associate each coil to the list through the bar code being
able to add considerations and photographs on the status of the coil prior to dis-
charge when any defect is detected.
• Once the preliminary inspection is finished, starts the unload. This task is carried
out entirely by the port staff, hired by the terminal. Each time a coil lands on the
dock, the operator must perform another visual inspection to detect any defects
caused during the unload. Again, the application includes incident report with
multimedia. Once the coil is marked as unloaded, the operator takes out a RFID tag
from a bag and places it in a front of the coil, preferably over a freight. The mobile
computer includes an UHF module allowing to map the RFID EPC with the bar
code in the project data base.
Every time a coil is unloaded, the application reports in real time not only to the port
terminal but also to the port authority, providing important info that can improve the
planning about the vessel stay in the port.
zamfira@unitbv.ro
952 I. Angulo et al.
Finally, the operation differs between the coils being transported by road or rail. The
open trucks can enter the warehouse and the crane is responsible for loading the coils
directly. However, the coils which are transported to destination by rail are arranged in
the transit area of the warehouse to be loaded on the railroad wagons by forklifts.
4 Implementation Details
The development of the global system has required to implement two main subsystems:
the handheld application and the RFID embedded reader. In addition, the conditions of
stored merchandise, steel coils, posed a technological challenge when using a tech-
nology based on radiofrequency. It was necessary to perform an analysis of the
behavior of different RFID tags in the project scenario.
Table 1. RSSI signal received in the RFID reader at different distances by tested tags
Tag 1m 2m 3m 4m 5m
Confidex Ironside −26 −30 −39 - -
Confidex Ironside Micro −29 −36 - - -
Confidex Survivor −26 −29 −33 −36 −39
M 116431 Gao −28 −32 −37 −39 -
Confidex Carrier Tough −29 −32 −36 −41 -
zamfira@unitbv.ro
Port Logistics: Improvement of Import Process Using RFID 953
As can be seen in Table 1, the results vary markedly among the labels used. As a
result, the label chosen for the development of the pilot, due to its significantly cheaper
price, was the “Confidex Carrier”. A label for conventional use that is glued to the coil
using a methacrylate insulation with 1 cm thick.
One of the main problems was to locate the optimum position of the RFID tag on
the coil. Unfortunately, any possible position includes potential risks. The location on
the inside of the coil was ruled out by its poor electromagnetic behavior. In addition,
the frictions caused with cylinder of the forklift can damage the device. In this sense, as
it is usual to stack coils at various levels, and they are located supported by each other,
it was decided to avoid their placement on the outside of the coil. Therefore, the best
alternative is to place the label on one side of the coil. This involves placing two
antennas at both ends of the forklift cylinder and on both sides of the crane hook.
zamfira@unitbv.ro
954 I. Angulo et al.
Fig. 2. Operator of the port terminal validating the unloading of a set of coils.
The M6e embedded reader is a small size high performance RFID reader that
provides Support for two monostatic RF antennas and read and write levels, command
adjustable from −5 dBm to +30 dBm. Connection with the Raspberry Pi is done by an
UART port.
zamfira@unitbv.ro
Port Logistics: Improvement of Import Process Using RFID 955
5 Conclusion
To optimize the warehouse management, every single movement of any coil is con-
veniently reported to the adopted Port ERP, keeping the location always updated. This
automatic generation of a map of coils drastically reduces the time spent by operators in
locating the requested units and the number of intermediate movements in the ware-
house. Also, failures due to misidentification of coils are removed. The port community
system of the Port of Bilbao receives detailed information about the unload of ships
process allowing anticipate disagreements with planning that may affect the transit of
other ships. Finally, the port authority and customs authority receive updated input and
output of trucks in the terminal.
Furthermore, RFID eradicates mistaken coil transport reducing significantly the
error rate.
References
1. Wu, Y., Xiong, X., Gang, X., Nyberg, T.R.: Study on intelligent port under the construction
of smart city. In: 2013 IEEE International Conference on Service Operations and Logistics,
and Informatics (SOLI), Dongguan, pp. 175–179 (2013). doi:10.1109/SOLI.2013.6611405
2. Wang, Z., Subramanian, N., Abdulrahman, M.D., Cui, H., Wu, L., Liu, C.: Port sustainable
services innovation: Ningbo port users’ expectation. Sustain. Prod. Consumption, 23 August
2016. doi:10.1016/j.spc.2016.08.002. ISSN 2352-5509
3. Fontana, C.F., Papa, F., Marte, C.L., Yoshioka, L.R., Sakurai, C.A.: Intelligent transportation
system as a part of seaport terminal management system. Int. J. Syst. Appl. Eng. Dev. 8, 41–
46 (2014)
4. Kim, M., Kim, K.: Automated RFID-based identification system for steel coils. Prog.
Electromagnet. Res. 131, 1–17 (2012)
5. Jung, S.-C., Kim, M.-S., Yang, Y.: Baseband noise reduction method using captured TX
signal for UHF RFID reader applications. IEEE Trans. Ind. Electron. 59, 592–598 (2012).
ISSN 0278-0046
6. Kuo, S.K., Chen, S.L., Lin, C.T.: Design and development of RFID label for steel coil. IEEE
Trans. Ind. Electron. 57(6), 2180–2186 (2010). doi:10.1109/TIE.2009.2034174
zamfira@unitbv.ro
956 I. Angulo et al.
7. Ukkonen, L., Sydänheimo, L., Kivikoski, M.: Read range performance comparison of
compact reader antennas for a handheld UHF RFID reader. In: 2007 IEEE International
Conference on RFID, Grapevine, TX, pp. 63–70 (2007). doi:10.1109/RFID.2007.346151
8. Bong, G.H., Chang, Y.S., Oh, C.H.: A practical algorithm for reliability-based RFID event
management considering warehouse operational environment. Int. J. Adv. Logistics 3(3),
100–108 (2014)
9. Douglas, B., Kumar, D., Meeden, L., Yanco, H.: Pyro: a python-based versatile programming
environment for teaching robotics. J. Educ. Res. Comput. (JERIC) 4(3) (2004)
zamfira@unitbv.ro
Integration of an LMS,
an IR and a Remote Lab
Abstract. For over two decades an IR – Institutional Repository (at the time
referred to as Digital Library) and an LMS – Learning Management System
have been developed and integrated under the Maxwell System at Pontifícia
Universidade Católica of Rio de Janeiro (PUC-Rio). It supports traditional
face-to-face courses and offers distance and blended learning options. It is also a
publishing platform. This model has proved very practical for many reasons
mentioned in this work. To enhance the options for traditional, blended and
distance learning, a Remote Lab was added to the Maxwell System. Adding a
Remote Lab is an enhancement to the learning environment since it is a “real”
equipment and not only a software for numerical computation. This work
addresses this new integration and how it benefits from the original infrastruc-
ture of an IR and an LMS implemented as a single platform.
1 Introduction
ICT – Information and Communication Technology has provided a large number and a
wide variety of tools to support teaching and learning. Engineering Education has
benefitted from these tools. Many software products have allowed students to simulate
and/or solve problems. The same products help students get ready to go to the “real”
lab by simulating experiments in advance. Videos, interactive courseware, animations,
texts, etc. also support learning and are the contents that help blended learning
(b-learning) and distance learning (e-learning) to be accomplished.
Engineering Education requires experimentation. In Electrical Engineering, there
are lab classes for Electric Circuits, Analog Electronics, Digital Electronics, Control
Systems, Eletromechanical Energy Conversion, etc. They offer experiments with “real”
equipment and components, and prepare future engineers to deal with “real”
physical/technical problems. Remote Labs are a fairly new resource that is meant to be
added to the options to be used in Engineering Education. Remote Labs are “real” labs
that can remotely be used through computer networks, including the Internet. They can
be used in addition to traditional labs to provide one more step of preparation before the
traditional lab classes.
The use of ICT tools requires the management of multiple resources and different
platforms. This is of paramount importance for the students and instructors to easily use
them and commute among them in a seamless way. This integration – platforms,
resources and users – is a task for the technical staff. The technical staff must work very
closely with the users in order to provide solutions that suit their needs. This work
addresses such an integration. It presents the results of integrating a Remote Lab to a
systems that is at the same time an Institutional Repository and a Learning Manage-
ment System.
Section 2 introduces some technical definitions. The context at the university is
addressed in Sect. 3 that is very important due to its long experience in using ICT for
both digital resources management and as a learning support. Section 4 explains the
integration of the Remote Lab to the local platform. The use of the Remote Lab is
presented in Sect. 5 and final remarks are in Sect. 6.
2 Some Definitions
The title of this article contains two acronyms and an expression that must be defined
so that their uses are made clear.
• LMS – Learning Management System
Wright et al. [1] defined a Learning Management System as “An LMS is com-
prehensive, integrated software that supports the development, delivery, assess-
ment, and administration of courses in traditional face-to-face, blended, or online
learning environments.”
An LMS is a software environment to support different types of learning processes.
There are many products available. Some are commercial solutions and others are free
and open source products. There are also many “home grown” LMSs.
• IR – Institutional Repository
Lynch [2] created the expression Institutional Repository as “A university-based
institutional repository is a set of services that a university offers to the members
of its community for the management and dissemination of digital materials
created by the institution and its community members. It is most essentially an
organizational commitment to the stewardship of these digital materials, including
long-term preservation where appropriate, as well as organization and access or
distribution.”
Currently it is used worldwide and has taken the place of the expression digital
library that was very popular in the 1990s. It is very broad since it aims at digital
materials created by the institution – this means that articles, ETDs (Electronic Theses
and Dissertations), senior projects, monographs, etc. can be included. But not only
these – digital learning materials can be on the IR too.
An interesting aspect of this definition is that it addresses digital materials; this
means that an IR is not a catalog of non-digital items. It hosts both the descriptions and
zamfira@unitbv.ro
Integration of an LMS, an IR and a Remote Lab 959
the digital files of the documents that belong to the collection. Since digital learning
materials were mentioned, it is necessary to present some definitions related to them.
The first is LO – Learning Object, the second is SCO – Shareable Content Object and
the third is Asset. They follow:
• LO – Learning Object
The definition of a Learning Object comes from IEEE LTSC – The Institute of
Electrical and Electronics Engineers Learning Technology Standards Committee in
page 1 of its IEEE Standard for Learning Object Metadata [3]: “For this standard, a
learning object is defined as any entity – digital or non-digital – that may be used
for learning, education, or training.”
This definition is important in the context of Engineering Education because it
allows non-digital artifacts to be classified as LOs.
• SCO – Shareable Content Object
SCORM – Shareable Content Object Reference Model is defined as: “The
Shareable Content Object Reference Model (SCORM) is a model that references
and integrates a set of interrelated technical standards, specifications, and guide-
lines designed to meet high-level requirements for e-learning content and systems.”
in [4] page 3-3. The SCORM defines SCO as: “SCOs are the smallest logical units of
information you can deliver to your learners via an LMS.” [4] page 11-4.
SCOs are always digital since they are to be delivered via an LMS. This is a
difference they present when compared to LOs. At the same time, SCOs and LOs have
two common characteristics – they are units with educational purposes and are “seen”
by LMSs.
• Asset
Assets are defined by SCORM as: “Assets are electronic representations of
media, texts, images, sounds, HTML pages, assessment objects, and other pieces of
data. They do not communicate with the LMS.” [4] page 3-2.
In 2000, the terms reusable chuncks of instructional media, reusable instructional
components, reusable digital resources, reusable learning objects (LO) were introduced
by Wiley [5]. Later on, in 2009, the term Reusable Learning Object (RLO) was used by
Alsubaie [6].
Except for Asset, SCO, LO, RLO, instructional media, reusable instructional
components and reusable digital resources, have much in common and, therefore, fuzzy
boundaries. At the same time, reusable digital resources can be used for Assets.
• Remote Lab
The expression Remote Lab is associated with VRLs – Virtual & Remote Labs.
Virtual and Remote Labs are different from one another. Heradio et al. [7] presented
four possibilities of labs according to their physical natures and the ways they are
accessed:
• Local Access – Real Resource
• Local Access – Simulated Resource
zamfira@unitbv.ro
960 A.M. Beltran Pavani et al.
The context at the university is presented in three steps. The first is the system that is at
the same time an IR and an LMS; it is a single platform that hosts both functionalities.
The second is the first integration that was implemented with an external system. The
third is the Remote Lab that was integrated with the System.
zamfira@unitbv.ro
Integration of an LMS, an IR and a Remote Lab 961
zamfira@unitbv.ro
962 A.M. Beltran Pavani et al.
– self contained topics that can be used as references or to support other items;
(3) simulators and interactive exercises – items that allow students to practice; and
(4) texts of various natures (theses, dissertations, senior projects, monographs, articles)
that are at the same time products of the educational process and inputs to it.
zamfira@unitbv.ro
Integration of an LMS, an IR and a Remote Lab 963
The integration of VISIR and the Maxwell System had different aspects that are
complementary. All of them are necessary to achieve the model of use.
The model of use is based on three premises: (1) the Remote Lab, as the Maxwell
System, is an institutional resource that must be prepared to be used in different courses
with different instructors; this has consequences on the integration of both systems;
(2) the Remote Lab is part of the learning resources offered to students and faculty, and
for this reason is to be integrated with the platform where all other resources are made
available; and (3) digital materials are to be available for students to study and be
prepared for the use of VISIR, and to afterwards go to the traditional lab. The three
premises are discussed in the following subsections. Another subsection addresses the
installation of VISIR and the communication with the Maxwell System.
zamfira@unitbv.ro
964 A.M. Beltran Pavani et al.
– The sources that can be used in the corresponding experiment and their limits of
voltage and current;
– The components that can be used in the corresponding experiments. They are
subsets of the components listed in the Component List.
A Max List is associated with an experiment. Therefore, there is one Max List for
each experiment to be performed.
• Equipment Configuration File
The Equipment Configuration File is a text file identified as filename.cir. It is
generated by VISIR when used in the mode that allows experiments to be created. It
contains the following information:
– Equipment and components to be used in the experiment related to the Equipment
Configuration File;
– The components that can be used in the corresponding experiments. They are
subsets of the components listed in the Component List.
An Equipment Configuration File is associated with an experiment. Therefore,
there is one Equipment Configuration File for each experiment to be performed.
When an experiment is created in VISIR, the Equipment Configuration File is
automatically generated. The person creating it must save it on the local computer. In
the case addressed by this work, the creator must send it to the information processing
staff for the file to be described, added to the corresponding Collection (next para-
graphs) and stored on the system. The creator of the experiment can ask the staff to
upload the file to the Maxwell System.
The Equipment Configuration File on the Maxwell System is of paramount
importance in the integration process. This will be addressed in Subsect. 4.4.
It is clear the Component List is a document of VISIR. On the other hand, the Max
Lists and the Equipment Configuration Files are documents that refer to experiments.
In order to store, make available and preserve these documents, the IR character-
istic of the Maxwell System was used. The DCMES Standard classifies types of
resources, one of the types is a “collection”, which means a set of resources with
specific characteristics. The Maxwell System uses two other attributes to classify
resources – “subtype” and “nature”. Subtype adds a more specific characterization of
resources; for example, “text” is very wide and needs additional specification, such as
article, manual, ETD, etc. “Nature” is used to specify a focus to the resource; two
examples are “technical” and “educational”. Combining the three possibilities –
type/subtype/nature – two sets of collections were created:
• Documentação Técnica do VISIR – Virtual Instruments System in Reality
(Collection/Technical Documentation/Technical)
This collection holds the Component List, photographs of the protoboards with the
components, schematic representations of the components on the protoboards, the Data
Sheet of VISIR provided by the manufacturer (with the necessary authorization) and
the Manual Técnico de Utilização do VISIR [12] written by Barbosa.
zamfira@unitbv.ro
Integration of an LMS, an IR and a Remote Lab 965
Two remarks are important. The first is that the manual contains information about
VISIR and its installation at PUC-Rio. The second is that all the documents, except the
Data Sheet are updated along the time according to changes in the configuration.
Changes in the configuration may occur do to adding and/or deleting experiments.
• Name of the Experiment(*) (Collection/Remote Experiment/Educational)
There is one such collection for each experiment to be performed. Each one con-
tains the Experiments Descriptions & Assignments (text), a set of support digital
resources, the Max Lists for the experiments and the Equipment Configuration Files.
(*)
An example of title for a collection is Circuitos de Primera Ordem (First Order
Circuits).
Figures 1 and 2 show the catalog descriptions of two collections – the Technical
Documentation Collection and a Remote Experiment Collection. There are as many of
the second type collection as the number of experiments installed on VISIR.
A Remote Experiment Collection may hold all types of learning materials the
instructor chooses from the ones available on the Maxwell System. More on this
subject will be presented when the third premise is discussed.
At this point, there is a connection with the Assets that were presented in Sect. 2.
Some of the items of the collections do not have educational functions and have no
meaning outside the collections they belong; some examples are the Max Lists, the
Component List and the photos of the protoboards. For this reason, they were classified
as Assets. On the other hand, there are many resources that are not Assets; some
examples are the Experiments Descriptions & Assignments and the Manual Técnico.
zamfira@unitbv.ro
966 A.M. Beltran Pavani et al.
zamfira@unitbv.ro
Integration of an LMS, an IR and a Remote Lab 967
support the experiments. The reason for this is that before VISIR, the lab classes used
the Maxwell System only for support – hosting experiment descriptions, posting
agendas, etc. One example of such new resource is the Simulator Object, developed
using SciLab® Circuitos RLC de Segunda Ordem em Diferentes Topologias [18].
Figure 3 shows the opening screen of the object and Fig. 4 one of the configurations
that the object supports.
zamfira@unitbv.ro
968 A.M. Beltran Pavani et al.
Fig. 5. Remote Lab environment on the Maxwell System showing the link to “Entrar” (Enter),
in the ellipse, to perform Remote Experiment 3.
5 Using VISIR
VISIR is used by three different players – instructors, students and technical staff. Each
has a particular and specialized set of functions and the system must support all of
them. The Maxwell System has always been used by the same three players. Before the
profiles of the users are presented, it is important to remark that the system has always
identified persons by their roles and each role has a specific profile of functions and
authorization levels. Thus, only new functions had to be created.
As mentioned in Sect. 3, the decision to integrate the Remote Lab to the Maxwell
System yielded the saving of a lot of work – all users are already identified and have
profiles, information comes from the university administrative system, courseware with
defined levels of access are on the IR, tables of courses and classes are on the LMS, etc.
The players and the adjustments made to suit them follow.
• Technical Staff
The functions available to technical staff were not impacted since this set of users
has been managing persons and resources for many years. The integration of VISIR
and Maxwell was very comfortable for this group.
• Instructors
Some work has been devoted to add functions for instructors to use and manage
VISIR from the System. Two were adjustments of functions already available for the
use and management of the “Classroom” and/or the “Virtualroom”: (1) configuration of
the environment to suit the needs and/or preferences of the instructors; and
(2) assessment of accesses by students. Two were new functions related to information
zamfira@unitbv.ro
Integration of an LMS, an IR and a Remote Lab 969
on the IR: (1) browsing Remote Experiments Collections; and (2) browsing Technical
Documentation collections. One was a completely new function that is performed from
the system on the VISIR equipment – the Creation of a New Experiment. To perform
this function VISIR has a feature that is not available to students – a button with a “+”
sign that allows the inclusion of components. Figure 6 shows the VISIR interface when
a remote experiment can be created – the “+” sign is in the ellipse.
The other functions – posting news on the bulletin board, mailing list, posting
grades, assigning activities, posting bibliography, etc. – were not affected. They are
used as they have always been.
Figures 7 and 8 show, respectively, the instructors menu and an online accesses
report (students’ names were erased).
• Students
Students maintained the same functions they have always had and got a new one –
access to VISIR. Figure 5 shows the list of remote experiments where only the last is
available due to the scheduling defined by the instructors. Figure 9 shows the envi-
ronment offered to students to get ready to experiments – the courseware is organized
according to the configuration created using the function in the Instructors’ menu. The
upper part contains the Reference Resources and the lower part (partially shown)
contains the Experiments Assignments.
zamfira@unitbv.ro
970 A.M. Beltran Pavani et al.
zamfira@unitbv.ro
Integration of an LMS, an IR and a Remote Lab 971
6 Final Remarks
This is a project under way. The implementation is compliant with the premises that
were stated.
The first run of VISIR at the university was in the second semester of 2016 which
was the first semester of the integration. It may happen that new functions will be
necessary and current functions will need enhancements.
At the end of the term, questionnaires were handed to students and are currently
under analysis. Adjustments, new functions, etc. can be implemented as consequence
of the surveys.
Beginning next March the course of General Electricity will start using VISIR, thus
new instructors and their students will be able to contribute with suggestions.
References
1. Wright, C.R., Lopes, V., Montgomerie, T.C., Reju, S.A.: Selecting a learning management
system: advice from an academic perspective. EDUCAUSE Rev., 21 April 2014. http://
www.educause.edu/ero/article/selecting-learning-management-system-advice-academic-
perspective. Accessed 05 Feb 2015
2. Lynch, C.: Institutional Repositories: essential infrastructure for scholarship in the digital
age, ARL Bimonthly report, 226, United States, February 2003. http://www.arl.org/
resources/pubs/br/br226/. Accessed 05 Feb 2015
3. IEEE Standard for Learning Object Metadata, 1484.12.1TM (2002). http://dx.doi.org/10.
1109/IEEESTD.2002.94128. Accessed 05 Feb 2015
zamfira@unitbv.ro
972 A.M. Beltran Pavani et al.
4. Advanced Distributed Learning: ADL Guidelines for Creating Reusable Content with
SCORM 2004, July 2008. http://www.adlnet.gov/wp-content/uploads/2011/07/ADL_
Guidelines_Creating_Reusable_Content.pdf. Accessed 05 Feb 2015
5. Wiley II, D.A.: Learning object design and sequencing theory. Ph.D. dissertation, Brigham
Young University, United States, June 2000. http://opencontent.org/docs/dissertation.pdf.
Accessed 07 Feb 2015
6. Alsubaie, M.: Reusable objects: learning object creation lifecycle. In: Proceedings of the
Second International Conference on Development of eSystems Engineering, DeSE 2009,
Abu Dhabi, UAE, pp. 321–325 (2009). http://dx.doi.org/10.1109/DeSE.2009.63. Accessed
07 Feb 2015
7. Heradio, R., de la Torre, L., Galan, D., Cabrerizo, F.J., Herrera-Viedma, E., Dormido, S.:
Virtual and remote labs in education: a bibliometric analysis. Comput. Educ. 98, 14–38
(2016). http://dx.doi.org/10.1016/j.compedu.2016.03.010
8. Pavani, A.M.B.: Creating a collection of assets in electrical engineering – a project under
way. In: Proceedings of ICEE 2015 – International Conference on Engineering Education on
Flash Memory, Croatia, pp. 315–322 (2015). ISBN 978–953-246-232-6. http://icee2015.
zsem.hr/images/ICEE2015_Proceedings.pdf
9. Marques, M.A., Viegas, M.C., Costa-Lobo, M.C., Fidalgo, A.V., Alves, G.R., Rocha, J.S.,
Gustavsson, I.: How remote labs impact on course outcomes: various practices using VISIR.
IEEE Trans. Educ. 57(3), 151–159 (2014). doi:10.1109/TE.2013.2284156
10. Tawfik, M., Sancristobal, E., Martín, S., Gil, C., Pesquera, A., Lozada, P., Díaz, G., Peire, J.,
Castro, M., García-Zubia, G., Hernández, U., Orduña, P., Angulo, I., Costa-Lobo, M.C.,
Marques, M.A., Viegas, M.C., Alves, G.R.: VISIR deployment in undergraduate engineer-
ing practices. In: Proceedings of the 2011 First Global Online Laboratory Consortium
Remote Laboratories Workshop, USA, pp. 1–7, October 2011. http://dx.doi.org/10.1109/
GOLC.2011.6086786
11. Alves, G.R., Viegas, M.C., Marques, M.A., Costa-Lobo, M.C., Silva, A.A., Formanski, F.,
Silva, J.B.: Student performance analysis under different Moodle course designs. In:
Proceedings of the 2012 15th International Conference on Interactive Collaborative
Learning, Austria, pp. 1–5, September 2012. http://dx.doi.org/10.1109/ICL.2012.6402181
12. Barbosa, W.S.: Manual Técnico de Utilização do VISIR, October 2016. http://www.
maxwell.vrac.puc-rio.br/27695/27695.PDF
13. http://www.maxwell.vrac.puc-rio.br/27810/27810.PDF
14. http://www.maxwell.vrac.puc-rio.br/series.php?tipBusca=dados&nrseqser=8
15. http://www.maxwell.vrac.puc-rio.br/series.php?tipBusca=dados&nrseqser=5
16. http://www.maxwell.vrac.puc-rio.br/series.php?tipBusca=dados&nrseqser=12
17. http://www.maxwell.vrac.puc-rio.br/series.php?tipBusca=dados&nrseqser=14
18. http://www.maxwell.vrac.puc-rio.br/Busca_etds.php?strSecao=resultado&nrSeq=27280@1
zamfira@unitbv.ro
Artificial Intelligence and Collaborative Robot
to Improve Airport Operations
1 Introduction
As air traffic continues to grow, security remains strict and capacity is stretched to the
limit, the need for efficient and safe airport operations is of first importance. The
problems created by poorly run systems are felt by everyone involved in air travel.
Time and safety are critical issues. Therefore reducing and improving airport operations
are ones of the biggest challenges which airlines are facing in these days when the
market is oversaturated with competition.
In order to optimize occupation of stopover zones and avoid airport congestion, the
European CO-FRIEND project, coordinated by AKKA Technologies, aims to use the
latest technologies in video-surveillance, video-tracking and artificial intelligence to
monitor airport operations. If the advancement of the operation schedules is updated in
real time, the actions to stay on schedule can be envisioned.
Airplanes are inspected periodically during maintenance operations on an airport
between flights. The reduction in inspection time is a major objective for airlines. If
maintenance operations are faster, this will optimize the availability of aircraft and
reduce operating costs. Nowadays, the inspection is performed by human operators
mainly visually, sometimes with some tools to evaluate defects. The French
multi-partner AIR-COBOT project of the Aerospace Valley, led by AKKA Tech-
nologies, aims to improve maintenance time and also traceability.
2 CO-FRIEND
The CO-FRIEND - COgnitive & Flexible learning system operating Robust Interpre-
tation of Extended real sceNes by multi-sensors Datafusion - project uses a series of
fixed and dome cameras, to automatically detects and monitors all stopover operations
around the aircraft and therefore contributes to improving the safety of people and
equipment. This project is a European project. The partners are AKKA Technologies,
Inria of Sophia Antipolis, University of Hamburg (Cognitive Systems Laboratory),
University of Leeds, University of Reading (Computational Vision Group),
Toulouse-Blagnac Airport. It begins in February 2008, lasts three years and follows up
the AVITRACK project [1–3].
The main objectives of this project and its predecessor are to develop techniques to
recognize and learn automatically all servicing operations around aircraft parked on
aprons. Both projects support a more efficient, safer and prompter management of
apron areas, in which ground operations such as fueling, baggage loading, etc., can
have a negative impact on the functioning of the airport, including traffic time delays.
2.1 Equipment
A number of main events, both static and dynamic, can be recognized and interpreted
by the system: aircraft, vehicles and people. In order to support this surveillance, the
system uses a set of fixed and Pan-Tilt-Zoom cameras with overlapping fields of view.
Figure 1 presents six fields of view from the airport cameras. All the cameras are
placed at strategic points in the apron areas. The camera streams are temporally syn-
chronized by a central video server.
zamfira@unitbv.ro
Artificial Intelligence and Collaborative Robot 975
zamfira@unitbv.ro
976 F. Donadio et al.
Spatial and temporal properties from detected mobile objects are modeled
employing soft computing relations, that is, spatial-temporal relations graded with
different strengths. This system is composed of three modules: the trajectory speed
analysis module, the trajectory clustering module, and the activity analysis module.
The first is aimed at segmenting the trajectory into segments of fairly similar speed.
The second is aimed at obtaining behavioral displacement patterns indicating the origin
and destination of mobile objects observed in the scene. It is achieved by clustering the
mobile trajectories and also by discovering the topology of the scene. The latter module
is aimed at extracting more complex patterns of activity, which include spatial infor-
mation coming from the trajectory analysis and temporal information related to the
interactions of mobiles observed in the scene, either between themselves or with
contextual elements of the scene. Spatial and temporal properties from detected mobile
zamfira@unitbv.ro
Artificial Intelligence and Collaborative Robot 977
objects are modeled employing soft computing relations. These can then be aggregated
employing typical soft-computing algebra. A clustering algorithm based on the tran-
sitive closure calculation of the final relation allows finding spatial-temporal patterns of
activity. This approach has been applied to dock-station monitoring at the
Toulouse-Blagnac airport [4–7].
Fig. 5. At left, full image. At right, zoomed image on the front landing gear.
3 AIR-COBOT
Previous robotic solutions for aircraft inspection focus on aircraft surface skin
inspection with robot crawling on the airplane surface. The AIR-COBOT - Aircraft
enhanced Inspection by smaRt & Collaborative rOBOT - projet chooses a different path
which leads to a collaborative mobile robot with cameras and a three-dimensional
scanner. Thanks to its acquisitions, a database dedicated to each airplane containing
images and scans, will be updated after each maintenance check. Researches have been
made on three main problematics which are: autonomous navigation, Non-Destructive
Testing (NDT), Human-Robot Interaction (HRI).
zamfira@unitbv.ro
978 F. Donadio et al.
The partners of this French project are AKKA Technologies, Airbus Group,
LAAS-CNRS, Institut Clément Ader of Albi, Stéréla, M3 Systems and 2MoRo
Solutions. It begins in January 2013, lasts three years. At the end, the partners decided
to keep the demonstrator in activity to continue research on these domains and improve
the cobot.
To navigate in the airport, the robot can go to an airplane parking thanks to
geolocalization data, or by following its human operator. To autonomously navigate
around the airplane, the robot is able to use laser and vision methods to localize itself
compared to the aircraft. Obstacle recognition and avoidance are also use in navigation
mode. The robot can inspect visually some items of the aircraft such as probes, static
ports, trapdoors, latches and scan some fuselage parts. It has a tasks checklist to follow.
The human operator controls the inspection diagnoses on its tablet. He also checks
visually the aircraft and can request additional NDT checks.
Fig. 6. At left, 4MOB platform. At right, AIR-COBOT is equipped with many sensors.
zamfira@unitbv.ro
Artificial Intelligence and Collaborative Robot 979
The open source framework Robot Operating System (ROS) has been used for
integrating device drivers and navigation algorithms. The robot has two industrial
computers, one running on Linux for the autonomous navigation module and the other
on Windows for the NDT module. The whole cobot weighs 230 kg.
The tablet interface provides several control panels to perform different actions:
changing the mission tasks or the navigation mode; checking the pose estimations or
the NDT results; reading robot warnings or interaction requests. Figure 7 presents a
view of the control panel for the NDT sensors.
Fig. 7. At left, control panel for the NDT sensors. At right, 3D scan visualization on the tablet.
At the end, the robot provides its diagnoses and asks the human to validate or refute
them. The operator can easily manipulate the pictures or the 3D scans for zooming or
rotating, see Fig. 7. Color representations of the results are put on the pictures or the 3D
scans to help the user comprehension.
zamfira@unitbv.ro
980 F. Donadio et al.
Near the aircraft, a point cloud in three dimensions is acquired thanks to the laser
scanning sensors fixed on pan-tilt units. Matching between the model of the aircraft and
the scene point cloud is performed to estimate the static pose of the robot. Figure 8
provides an example in an in-door context [9].
Fig. 8. Robot is located in back left of the aircraft in an inside environment. At left, 3D data is
acquired with a Hokuyo laser range finder moved thanks to a pan-tilt unit. At right, the matching
result is made of data (blue) with model (red).
The robot moves and holds this pose by considering the IMU, the wheel odometry
and the visual odometry. Laser data are also used horizontally in two dimensions. Pose
estimation of the robot is computed when enough elements from the landing gears and
engines are visible [9].
For visual localization, the robot estimates its pose relative to the aircraft using
visual elements (doors, windows, tires, static ports etc.) of the aircraft. Pattern recog-
nitions or extractions of features are used to detect those visual landmarks [10, 11]. By
detecting and tracking them, see Fig. 9, in addition to estimating its pose relative to the
aircraft, the robot can perform a visual servoing [12].
zamfira@unitbv.ro
Artificial Intelligence and Collaborative Robot 981
A first confidence index is computed based on the number of items visible in laser
data. A second confidence index is computed based on the number of visual features. If
good data confidence is achieved, the pose is updated. Artificial intelligence arbitrating
between those pose estimation results is in development [9, 11].
The laser data coming from laser range finders and visual data coming from the
cameras are used for detection, classification (moving, motionless) and recognition
(human, vehicle, other) of the obstacles [12]. The detection and the classification are
easier in the two-dimensional laser data, while identification is better in the images. The
two methods are complementary. Three kinds of avoidances are considered: stop and
wait for a free path, spiral obstacle avoidance and path planning. The chosen avoidance
approach depends on the robot’s surroundings (navigation corridor, tarmac area
without many obstacles, cluttered indoor environment etc.) at the time of the encounter
with an obstacle.
Fig. 10. Examples of items to inspect. From left to right, static port with its protection, open air
inlet valve, AOA probe, trap door with unlocked handle, reactor with foreign element. One can
note that light conditions are very different.
The detection uses shape recognition with regular shapes (rectangles, circles,
ellipses) or more complex shapes obtained with the projection in the image plane of the
3D model of the element to be inspected. The evaluation is based on indices such as the
uniformity of segmented regions, convexity of their forms, or periodicity of the image
pixels’ intensity [13, 14].
zamfira@unitbv.ro
982 F. Donadio et al.
Feature extraction using Speeded Up Robust Features (SURF) can also be applied
to perform the inspection of certain elements having two possible states, such as pitot
probes or static ports being covered or not covered, see Fig. 11. For such items, in
order to decrease the mission time, visual inspection with the navigation cameras
during displacements around the aircraft is under consideration [11].
Fig. 11. Two king of non-destructive testing. At left, visual inspection of the static port with
SURF method. At right, 3D inspection, tridimensional scan of the aircraft surface, a bump is
visible in the middle. The writing, visible in the top, helps to locate precisely the default.
At given positions, the pantograph elevates the 3D scanner at the fuselage level.
A pan-tilt unit moves the Eva scanner to acquire the hull. Figure 11 shows a 3D scan.
By comparing the data acquired to the aircraft model, algorithms are able to diagnose
any faults in the fuselage structure and provide information on their shape, size and
depth.
As explained in [9] by moving the pan-tilt units of the laser range finders, it is also
possible to obtain a point cloud in three dimensions. It is planned to make targeted
acquisitions, simpler in terms of movement, to verify, for example, the absence of
chocks in front of the landing gear wheels, or the proper closing of latches.
3.4 Collaboration
The robot has three possible navigation modes: autonomous mode; follower mode and
remote control mode. The level of robot autonomy is decreasing between these three
modes and the human-robot interactions are adapted in consequence. The collaboration
between the robot and its human operator are described in [8].
In the autonomous mode, like explained in the previous sections, the robot per-
forms a list of tasks autonomously such as moving to a pose in the airport frame or in
the aircraft frame, inspects an item or avoids obstacles. The human operator has to stay
at proximity of the robot and check sometimes the robot behavior. Even so, he can
perform his own inspection tasks in the meantime.
In the follower mode, the robot follows the human operator until a change of mode.
The robot has to avoid obstacles and recognize its operator between other humans. In
the remote control mode, the human operator can displace the robot to a specific
location thanks to the remote control or specify an NDT task thanks to the tablet
interface. The human operator is in charge of the mobile platform and the NDT sensors.
zamfira@unitbv.ro
Artificial Intelligence and Collaborative Robot 983
Classical warnings of the robot can emerge if there is a crashing code problem, or a
material dysfunction. If possible, it continues the mission with its reduced capacities
and skips tasks linked to these problems until someone intervenes. One example during
navigation tasks, if the GPS signal is too weak, then the robot sends a soft warning to
the human operator and move a bit updating its pose with odometry measurements. But
at some point, it has to receive the GPS signal otherwise its confidence level of its pose
estimation would be too low and the robot would stop and send a strong warning to the
user. Second example during inspection tasks, if the elevator of the 3D scanner has a
mechanic malfunction and it is not elevating correctly then the human operator has to
check it.
During its navigation tasks, the robot has to follow navigation corridors and safety
trajectories around the aircraft. It warns its operator if it is stuck and it cannot avoid
safely an obstacle without leaving the navigation corridor or being too far safety
trajectory. In that case, the operator can choose the follower or remote control modes to
lead the robot away from the problem or move the obstacle that blocks its path. Alerts
are also sent if the robot enters a prohibited area or exceeds a given speed.
During its inspection tasks, the robot can warn the human operator that something
is wrong. For example, it did not find the element to inspect in the image or the 3D scan
seems incorrect. The cobot can ask for a fast human intervention for examples if there
is still the chock in front of the landing gear or the protection on a pitot probe.
The operator is also visually checking the aircraft. He could ask for a NDT check if
he thinks there might be a problem on the aircraft which is not taken into consideration
in the robot tasks. After moving the robot, he can control the sensors and asks the robot
to perform some tests. In Fig. 12, the human operator asks for a 3D scan.
Fig. 12. From left to right, elevation of the scanner in order to perform a scan of A320 aircraft in
a hangar of Air France Industries, tridimensional scan of the aircraft where a crack and a bump
are visible, inspection result. The inspection algorithms provide shape, size and depth of those
imperfections with visual color representation to help the human operator.
If the robot confirms a default for example a bump, the operator can add this check
for this particular aircraft. The robot remembers its pose compared to the aircraft and
the performed NDT check so it can do it the next time that it encounters this aircraft.
Figure 12 presents a scan of the aircraft and one associated diagnostic.
At regular intervals, the robot sends its pose estimation in airport frame or aircraft
to the tablet. The human operator can check it on a facility map for the first one, and on
an aircraft representation for the second one. The human operator has an access to the
mission plan status and the diagnostics in real time. He can, for example, check camera
preview before image analysis to check the camera pointing.
zamfira@unitbv.ro
984 F. Donadio et al.
Since the human operator has a better understanding of the environment, he can
take the control of the robot to avoid problems before they arrived or just change the
order of the list of tasks. The human is better adapted to understand if another worker
interferes with the robot mission and at the opposite, to take into account whether the
robot interferes with another worker. In conclusion, he is responsible for choosing
which one has the priority.
zamfira@unitbv.ro
Artificial Intelligence and Collaborative Robot 985
requested checking tasks into automatic tasks for a specific aircraft; learning new
obstacle to be able to perform recognition; developing its artificial intelligence.
5.1 Conclusions
The two projects presented in this paper can increase the efficiency of airport opera-
tions. Their monitoring is beneficial to detect problems and adapt them in consequence.
Other vehicles or operators could intervene in case of difficulty to keep the time
schedule. The human-robot collaboration will increase the efficiency and the reliability
of inspection, reduces the risk and uncertainties, self-adapt to different types of air-
crafts, service types, investigation contexts, and operational circumstances.
At the same time, the application of collaborative robots may add complexity to
airport management. The experiments in real airport environment are necessary and the
feedbacks from airlines, airport management companies and ground-support service
providers are important. In the future, it will be necessary to consider in details the
tradeoff between management complexity and robotic intelligence.
5.2 Prospects
The mobile platform is made for inspecting the lower parts of the aircraft. It is envi-
sioned to use it with a drone for the upper parts. The partnership between them is
beneficial due to complementary inspections from different point of view (ground, air)
and better adaptability.
AKKA Technologies envisions also Cyber Physical System (CPS) to monitor not
only the aircraft operations around the aircraft but all the activities in the airport to
provide better predictions and adapt the actions of human operators, future robotic
assistants, and autonomous vehicles.
Thanks to the traceability of maintenance and other actions, aircraft operations will
not be developed according to an aircraft model but to each particular aircraft. State of
the aircrafts will be more closely studied and constraining maintenance operations
could also be predicted in advance.
zamfira@unitbv.ro
986 F. Donadio et al.
References
1. Borg, M., Thirde, D., Ferryman, J., Fusier, F., Brémond, F., Thonnat, M.: An integrated
vision system for aircraft activity monitoring. In: Proceedings of the 6th IEEE International
Workshop on Performance Evaluation of Tracking and Surveillance (PETS) (2005)
2. Ferryman, J., Borg, M., Thirde, D., Fusier, F., Valentin, V., Bremond, F., Thonnat, M.,
Aguilera, J., Kampel, M.: Automated scene understanding for airport aprons. In: Advances
in Artificial Intelligence, AI 2005, vol. 3809, pp. 593–603 (2005)
3. Borg, M., Thirde, D., Ferryman, J., Fusier, F., Valentin, V., Brémond, F., Thonnat, M.: A
real-time scene understanding system for airport apron monitoring. In: Proceedings of IEEE
International Conference on Computer Vision Systems (ICVS) (2006)
4. Chau, D.P., Bremond, F., Thonnat, M., Corvee, E.: Robust mobile object tracking based on
multiple feature similarity and trajectory filtering. In: Proceedings of the International
Conference on Computer Vision Theory and Applications (VISAPP) (2011)
5. Patino, L., Brémond, F., Thonnat, M.: Activity discovery from video employing soft
computing relations. In: Proceedings of the IEEE International Joint Conference on Neural
Networks (IJCNN) (2010)
6. Patino, L., Brémond, F., Evans, M., Shahrokni, A., Ferryman, J.: Video activity extraction
and reporting with incremental unsupervised learning. In: Proceedings of the 7th IEEE
International Conference on Advanced Video and Signal-Based Surveillance (AVSS) (2010)
7. Greenall, J.P.: High-level activity learning and recognition in structured environments. Ph.D.
thesis, University of Leeds (2012)
8. Donadio, F., Frejaville, J., Larnier, S., Vetault, S.: Human-robot collaboration to perform
aircraft inspection in working environment. In: Proceedings of the 5th International
Conference on Machine Control and Guidance (MCG) (2016)
9. Frejaville, J., Larnier, S., Vetault, S.: Localisation à partir de données laser d’un robot
naviguant autour d’un avion. In: Proceedings of the Reconnaissance des Formes et
l’Intelligence Artificielle (RFIA) Congress (2016)
10. Jovančević, I., Viana, I., Orteu, J., Sentenac, T., Larnier, S.: Matching CAD model and
image features for robot navigation and inspection of an aircraft. In: Proceedings of the 5th
International Conference on Pattern Recognition Applications and Methods (ICPRAM),
pp. 359–366 (2016)
11. Villemot, T., Larnier, S., Vetault, S.: Détection d’amers visuels pour la navigation autour
d’un avion et son inspection. In: Proceedings of the Reconnaissance des Formes et
l’Intelligence Artificielle (RFIA) Congress (2016)
12. Futterlieb, M., Cadenat, V., Sentenac, T.: A navigational framework combining visual
servoing and spiral obstacle avoidance techniques. In: Proceedings of the 11th International
Conference on Informatics in Control, Automation and Robotics (ICINCO), pp. 57–64
(2014)
13. Jovančević, I., Arafat, A., Orteu, J., Sentenac, T.: Airplane tire inspection by image
processing techniques. In: Proceedings of the 5th Mediterranean Conference on Embedded
Computing (MECO) (2016)
14. Jovančević, I., Larnier, S., Orteu, J., Sentenac T.: Automated exterior inspection of an
aircraft with a pan-tilt-zoom camera mounted on a mobile robot. J. Electron. Imag. 24(6)
(2015)
zamfira@unitbv.ro
Methodological Proposal for Use of Virtual
Reality VR and Augmented Reality
AR in the Formation of Professional Skills
in Industrial Maintenance
and Industrial Safety
1 Introduction
Different authors have developed applications that favor the understanding of good
practice in engineering based of augmented reality and virtual reality [1]. The main
areas in the engineering service in with VR and AR are oriented to the area of industrial
plants and aerospace and the main engineering services intervened are in their order;
maintenance, training and machine inspection [1].
These two methods; Virtual Reality VR and Augmented Reality AR share modes of
operation and similar principles of development, however the basic principle of Virtual
Reality takes the essential characteristics of real elements and through computer
applications are virtualized in environments of low interaction [2]. The virtual elements
replace the physical reality [3] and the environments are pre-designed, based on it the
mechanisms of evaluation these applications are concentrated in the degree of approach
with reality, its coherence with the physical environment (laws and behaviors) and
interaction with the user [4].
While the AR proposal seeks to take advantage of the real environment and on this
tool, generates points of interaction with the user [5], increasing the connection of the
senses with the recreated phenomenon, especially the perception of location and
movement.
The degree of attachment of person is greater and their perception of reality is
higher than VR. The main characteristics in which these technologies are evaluated are
focused on: Reliability, Sensitivity and Agility [6]. However, a variety of techniques
are known that make a continuous development between virtual reality and real reality.
Although many AR and VR combinations are shown there are several clearly
defined combinations. In Fig. 1, it is shown how the development of these tools is a
continuous development from virtual reality to the total real.
Fig. 1. Classification of reality concepts according to correlation between perception and action
and level of interaction [7].
In order to VR and AR tools to fulfill the purposes for which they have been designed,
it is necessary to follow a methodology of development and implementation within the
context of the use of engineering laboratories, which starts from the training objectives
until the evaluation in their different aspects.
To this end, an integrated methodology has been proposed for the elaboration of
laboratories in different modalities; Face-to-face, remote and virtual (local and virtual in
the cloud).
zamfira@unitbv.ro
Methodological Proposal for Use of Virtual Reality VR 989
This methodology called SMART (by its acronym in Spanish Modular Temporal
Rapid Access System), is composed of phases that incorporate documents and infor-
mation in order to make a robust and coherent proposal. The final objective teaching
besides generating the object of learning of VR or AR in laboratories is to incorporate
the proposal generated in a broader and scalable environment, until reaching to a
confederation of laboratories, teaching (Fig. 2).
2.1 Requirements
In this phase the requirements of curricular design with the training strategies are
incorporated in order to develop the professional competences that evidence the degree
of development of superior order in the formation of students in engineering.
Some of the best known standards are ABET [8], Bologna, ISCED [9], World Bank
and OECD [10]. One of the most important aspects is to take advantage of the structure
of the standard.
2.2 Architecture
The next phase corresponds to the design of the learning object, that is getting char-
acteristics of the student, equipment and systems available and normative standards.
For example, for Dini [1], the three main methods of aiding the generation of aug-
mented reality experiences are optical combination, video mixing and image
projection.
zamfira@unitbv.ro
990 J.D. Velosa et al.
2.3 Construction
This phase includes two components: the preparation to put into use the first prototype.
Here we determine the definition of the general laboratory structures implemented by
the institution, the elements for its implementation with students and the definition of
mechanisms of control and evaluation mechanisms of students.
2.4 Evolution
As a final phase, the development of the VR and AR learning experience should be
evaluated. This verification is done in three aspects: Evaluation of the implementation,
Evaluation of the competences reached by students and the Evaluation of the Learning
students have gotten. The evaluation of the implementation is done with methods that
evaluate the student’s perception: Technology Acceptance Model (TAM) and ARI [11].
The evaluation of competencies is carried out using learning process evaluation
tools oriented by a certification organization in the ABET study and learning with
Conceiving — Designing — Implementing — Operating (CDIO) MR [12] and
OTSM-TRIZ [13] and Learning Analytics [14, 15].
The methodology SMART (Sistema Modular de Acceso Rápido Temporal for
abbreviators in Spanish), used to integrate all the phases before mentioned, is presented
in the next Fig. 3.
The methodology is valid for any type of engineering competence generation
experience. The way to integrate the features in each step is done through the QFD
(Quality Function Deployment) tools.
Traditionally, the maintenance and industrial safety course of the EAN University
(Colombia) for manufacturing engineers has been worked through scenarios and
hypothetical situations, trying to represent reality through assumptions.
Particularly in the case of maintenance subject it, has been made practices of
element description, but in its application it is not always possible to relate and inte-
grate in a single experience; Machine characteristics, maintenance characteristics,
equipment status and performance against the equipment.
For this experience, the student develops activities with the help of the professor
and laboratory professional in short and always assisted moments, almost always
generating superficial results with fragments of information, highly directed work and
lack of additional information.
Similarly, in the case of occupational hazards, the experience cannot be brought
into the context of a real decision-making situation with multiple sources of infor-
mation to a student without a serious risk of accident or Low understanding of the
phenomenon by being concentrated in the risky situation.
Based on this, the questions that this article seeks to answer are: how can the use of
augmented and virtual reality technological and computer tools facilitate training in the
zamfira@unitbv.ro
Methodological Proposal for Use of Virtual Reality VR 991
zamfira@unitbv.ro
992 J.D. Velosa et al.
The competences that were to be evaluated from the ABET standard were:
• An ability to perform standard tests and measurements, and to conduct, analyze and
interpret experiments.
• Ability to apply written, oral and graphic communication in technical and
non-technical environments; And the ability to identify and use appropriate tech-
nical literature.
• Ability to identify, analyze and solve engineering technology problems.
The physical structure is comprised of a Smartphone (with spherical image viewing
applications, file editing and selection and WiFi connection), virtual reality vision
glasses, control computer with Internet access and a closed space (Fig. 5).
zamfira@unitbv.ro
Methodological Proposal for Use of Virtual Reality VR 993
For the construction phase of the experiment, four photos were taken of the three
most important risk factors that were evidenced in different spaces of the laboratory. An
example of this is shown in the following photo (Fig. 6).
After that, the photos are loaded into the memory of the Smartphone that is edited
with the Virtual Reality App.
The constriction or development of the experience is done in the laboratory space.
Students visualize the spherical photo pre-designed with the risks and based on their
knowledge, first identifies the risks presented, then analyzes the degree of severity of
the same and finally proposes the best way to mitigate or avoid it. The identification
can be assisted by the professor or by other students.
Finally, the professor, with the help of the evidence gathered in the experience,
evaluates the three aspects of the student and his experience: Evaluation of the
implementation, Evaluation of the competences reached and the Evaluation of the
Learning. The evaluation of the implementation is carried out by means of a survey
made to the students after the reality virtual and augmented of the experience; this one
is based on the perception of the technology, through the proposal Technology
Acceptance Model (TAM) on perception of the technological acquisition.
The evaluation of the competences and abilities is based on the proposal of certi-
fication of ABET and the elements delivered by the student, among them.
• Risk overview
• Risk Matrix
• Audit and risk control report
zamfira@unitbv.ro
994 J.D. Velosa et al.
The methodology takes as a principle the proposal of [18] and the SMART proposal,
on the application of augmented reality in operating environments and the training
process in manufacturing engineers in the context of engineering design [19, 20].
Based on this integration proposal, a virtual reality activities was designed; First
real spherical photos of risk situations were taken in the laboratory and presented to
students through virtual reality helmets. See Fig. 1(b). Students with the support of
prior information and teacher instructions, and real-time laboratory professional could
interpret the situations presented to them. To See Fig. 7(a).
Fig. 7. Photos of the experience (a) spherical photo identification of equipment (b) Students
developing the industrial safety activity with AR glasses
4.1 Results
The data focus on the perception of the implementation and use of the technology of
the professional versus the use of the activities implemented. To this end, the survey
tool was applied which probes the characteristics of Perceived Utility and ease of use.
The instrument takes into account the TAM (Technology Acceptance Model) method
described by [21, 22].
zamfira@unitbv.ro
Methodological Proposal for Use of Virtual Reality VR 995
Through this tool the perception of the usefulness of the technological imple-
mentation designed and the ease of use is investigated. The survey was applied after the
experience. Facing the perception that the students have evaluated five questions, each
on a Likert scale of 1 to 10 where 10 is the most accepted degree. The results of the
students are shown in graphs 8 and 9.
Based on this it is observed that the highest values are obtained in the question: Q3
The tools used (equipment and 3D material and software) are clear. This corroborates
that the strategy of using photos 360 of the laboratory or the place where the practice is
performed the student achieves a good perception of reality.
Meanwhile, the lowest score in this respect is found in the question; Q0 - Your
level of knowledge of practice laboratories is (at the end), this got a 8.4/10. One of the
main reasons for this value is the limited time between the delivery of the support
material and the practice. The other aspects are enclosed in high ratings.
Perceived utility has generally high values. The most valued answer was: Q4. The
overall evaluation of laboratory integration is positive, confirming the acceptability of
these novel proposals (Fig. 8).
And the one of lower qualification is the one that involves the participation and
intervention of the teacher. This is especially evident in the component that evaluates
the perception of use. The main reason for this appreciation is the orientation given to
the activity by the teacher on the tools that are used (Fig. 9).
The questions were formulated against the characteristics of the MAT model; these
are:
• Knowledge:
Knowledge Achieved when implementing the tool and when used by the student, in
conjunction with the application that the student can give without needing assistance
from the teacher. The related questions are Q0 and Q6.
zamfira@unitbv.ro
996 J.D. Velosa et al.
• Use
Characteristic observed in the perception that this tool can have the training by the
student. And in the ease of using the tool with the basic knowledge gained. The related
questions are Q1, Q9 and Q10.
• Coherence
Characteristic that looks for the strong relation between the reality and the situations
that normally could be faced an Engineer. The related questions are Q2.1 and Q5.
• Clarity
Both in the instructions and in the use that is given to the implementation of the
experience. The related questions are Q3 and Q8.
• Degree of acceptance
It measures the degree of conformity to what the student expects and what ultimately
results from his/her experience. The related questions are Q4 and Q7.
In the following graph it is observed how the characteristics that keep an equal or
similar valuation is the Usage and the coherence and the characteristic that has a greater
difference between the Perceived Utility and the ease of use is the degree and
knowledge that can be reached (Fig. 10).
zamfira@unitbv.ro
Methodological Proposal for Use of Virtual Reality VR 997
Fig. 10. Value of the characteristics of each of the factors of the TAM methodology
With the help of this matrix and the level of the indicator of the degree of devel-
opment of the competences evaluated, a tool of evaluation of the competition was
proposed based on the application of the experience of virtual reality.
The main elements of the rubric and the average grade of the group evaluated are
presented in the following Table 2.
zamfira@unitbv.ro
998 J.D. Velosa et al.
With the help of exercise and VR and augmented virtual activities AR will establish
practices consistent with the skills and competencies required by the course, including
identification of tacit and active risks, risk rating and action plan. More tests are
currently being made with environments developed with different groups. The expected
results are:
• To corroborate the effectiveness of the use of these augmented reality AR and
virtual VR tools in industrial safety practices. To reduce the risk to the student of
dangers increasing the degree of depth in the situations. Involve the outcome of this
project in a larger project or the one of hybrid laboratories for laboratories federations.
zamfira@unitbv.ro
Methodological Proposal for Use of Virtual Reality VR 999
References
1. Dini, G., Mura, M.D.: Application of augmented reality techniques in through-life
engineering services. Procedia CIRP 38, 14–23 (2015)
2. Weyrich, M., Drews, P.: An interactive environment for virtual manufacturing: the virtual
workbench. Comput. Ind. 38(1), 5–15 (1999)
3. Vora, J., Nair, S., Gramopadhye, A.K., Duchowski, A.T., Melloy, B.J., Kanki, B.: Using
virtual reality technology for aircraft visual inspection training: presence and comparison
studies. Appl. Ergon. 33(6), 559–570 (2002)
4. Murat, A., Akcayir, G., Pektas, H.M., Ocak, M.A.: Augmented reality in science
laboratories: the effects of augmented reality on university students’ laboratory skills and
attitudes toward science laboratories. Comput. Hum. Behav. 57, 334–342 (2016)
5. Chang, Y., Liu, H., Kang, Y.: Using augmented reality smart glasses to design games for
cognitive training, pp. 246–247 (2016)
6. Elia, V., Gnoni, M.G., Lanzilotto, A.: Evaluating the application of augmented reality
devices in manufacturing from a process point of view: an AHP based model. Expert Syst.
Appl. 63, 187–197 (2016)
7. Schnabel, M.A., Wang, X., Seichter, H., Kvan, T.: From virtuality to reality and back. In:
Proceedings of the International Association of Societies of Design Research, pp. 1–15
(2007)
8. A. Commission Technology Accreditation. Criteria for Accrediting Engineering Technology
(2014)
9. UNESCO: The International Standard Classification of Education 2011 (2013)
10. Tremblay, K., Lalancette, D., Roseveare, D.: Assessment of higher education learning
outcomes (AHELO) feasibility study. Feasibility Stud. Rep. 1, 113–126 (2013)
11. Georgiou, Y., Kyza, E.A.: crossmark, vol. 98, July 2016, pp. 24–37 (2017)
12. Cdio: The CDIO Initiative. vol. 0, no. 26 January 2011, pp. 1–14 (2010)
13. Becattini, N., Cascini, G., Rotini, F.: OTSM-TRIZ network of problems for evaluating the
design skills of engineering students. Procedia Eng. 131, 689–700 (2015)
14. Orduña, P., Almeida, A., López-De-Ipiña, D., Garcia-Zubia, J.: Learning Analytics on
federated remote laboratories: tips and techniques. In: IEEE Global Engineering Education
Conference EDUCON, no April, pp. 299–305 (2014)
15. Serrano-Laguna, Á.: Computer Standards & Interfaces, vol. 50, September 2016,
pp. 116–123 (2017)
16. de Trabajo, M.: Decreto 1072 de 2015 (26 May 2015). Por medio del cual se expide el
Decreto Único Reglamentario del Sector Trabajo EL, vol. Version ac, pp. 1–326 (2015)
17. Gutiérrez, A.: Guía técnica para el análisis de exposición a factores de riesgo ocupacional
(2011)
zamfira@unitbv.ro
1000 J.D. Velosa et al.
18. Porcelli, I., Rapaccini, M., Espíndola, D.B., Pereira, C.E.: Technical and organizational
issues about the introduction of augmented reality in maintenance and technical assistance
services. In: IFAC Proceedings, pp. 257–262 (2013)
19. Abdullah, F., Ward, R., Al, H., Abdallah, F., Barbar, K., Aorks, T.: Comparing copresent
robots, telepresent robots and virtual agents. Comput. Hum. Behav. 55(2), 1–10 (2015)
20. Di Donato, M., Fiorentino, M., Uva, A.E., Gattullo, M., Monno, G.: Text legibility for
projected Augmented Reality on industrial workbenches. Comput. Ind. 70, 70–78 (2015)
21. De Aceptación, M., Tam, T., Cataldo, A.: Una revisión de la literatura, no. 45, pp. 1–9
(1986)
22. Phatthana, W., Mat, N.K.N.: The application of technology acceptance model (TAM) on
health tourism e-purchase intention predictors in Thailand. In: International Conference on
Business and Economics Research, vol. 1, pp. 196–199 (2010)
zamfira@unitbv.ro
Sketching 3D Immersed Experiences Rapidly
by Hand Through 2D Cross Sections
1 Introduction
Sketching is used by designers to rapidly capture and communicate ideas [1]. A sketch
can take many forms. It may be a hand drawing, constructed using computer tools, or
composed using other objects. It has been argued that computer tools generally are slow
to use, and that the delays incurred by operating the software compromises the creative
process [2]. Ideas arrive as a wave of thought and thus emerge in an instant and may also
disappear in an instant. It is therefore important to capture and express the idea in a
timely manner. Many argue for using simple sketching tools, such as drawing by hand.
Hand drawings also have the added advantage that they appear unfinished, yet
inspirational through their organic appearance [3]. Hand drawing is relatively easy to
achieve when expressing ideas for two-dimensional user interfaces [4, 5], and
sketching three-dimensional scenes and objects in perspective. However, such
three-dimensional sketches are static, representing the view in one direction from one
position. To sketch immersive interactive experiences often require that the designer
2 Background
zamfira@unitbv.ro
Sketching 3D Immersed Experiences Rapidly by Hand 1003
human interprets the semantics of shapes, which is relatively easy for humans and hard
for machines and the computer perform the other tasks that are hard for the human and
easy for the machine [28]. Another approach is to rely on fuzzy logic where uncertainty
in the 2D input is propagated to the 3D output [24, 29].
To capture complex shapes with continuous contours detailed line grids are needed.
Another approach for sketching complex 3D shapes is the use of shading [30] where
shades are used to express the shape of an object in 3D they way light reflects from the
surface of the object. A very different approach is to use the input sketch as a query into
a database of 3D models and then fit the model into a given scene [31]. This approach
obviously needs database content that matches the needs of the designer. Moreover,
using a sketch as a database query is not trivial.
Silhouette modelling has also been proposed, where a three-dimensional shape is
based on the silhouette of a two-dimensional sketch. For instance, in the Teddy system
[32] the users draws the outline of a cuddly toy in two-dimensions and the
three-dimensional shape is derived from the silhouette with a rounded shape, where
small parts are thinner and large parts become thicker. Two-dimensional silhouettes can
also be used to define three-dimensional cross sections [33], which is similar to the
approach proposed herein.
An interesting method proposed by Tolba et al. [34] mapped the perspective grid
with its vanishing lines to a two-dimensional sketch of a 3D sketch drawn in per-
spective and thereby capture the model. They also had a hybrid approach which was
not solely focused on capturing the three-dimensional model but on the viewing
experience as they also experimented with panoramic sketches where panoramic
images were composed using four 3D sketches of the four viewing directions mapped
onto a unit sphere, this is similar to the approach proposed herein.
It has been proposed to sketch directly in the equirectangular panoramic domain
[35]. This approach allows designers to use their existing three-dimensional sketching
experience. Sketches are still drawn from one position, but the omnidirectional nature
of the sketch allows the sketches to be viewed using panoramic viewing software
giving the viewer a stronger sense of presence. The equirectangular panoramic map-
ping is a projection of the world onto a sphere represented using a geographical
coordinate system with latitude and longitude [36, 37]. A complete panorama repre-
senting all directions from –180 to 180° horizontally and –90 to 90° vertically becomes
a panoramic image with an aspect ratio of 2:1.
One property of the equirectangular projection is that vertical lines remain vertical,
while horizontal lines become curved. It thus require some experimentation and skill to
draw realistic models directly in the equirectangular domain. It was therefore proposed to
use equirectangular grid lines [38] where planes represented as grids are projected onto
the equirectangular panorama. By tracing the various lines along the x, y and z directions
it is possible to make perfect panoramic images. Attempts at translating the viewer inside
a panoramic image has also been attempted [39].
Although the grid lines reportedly helped, it is still non-trivial to sketch directly in
the panoramic domain as humans are not used to process entire panoramas on one go.
The work proposed herein can thus achieve the same results without having to operate
directly in the panoramic domain. It is based on flat two-dimensional cross sections and
humans are good at working with two-dimensional representations.
zamfira@unitbv.ro
1004 F.E. Sandnes
3 Method
zamfira@unitbv.ro
Sketching 3D Immersed Experiences Rapidly by Hand 1005
used instead of white and cyan to represent the background and grid lines, respectively.
Line drawing sketches will therefore become transparent wire-frames, while sketches
where regions are colored are modelled as non-transparent faces.
The 3D model is built as follows. Given an image I at an angle a at distance d with
width w. The pixel at image I at location [x, y] is given by I(x, y). It is assumed that the
viewing normal vector intersects the image in the middle. An image pixel I(x, y) is
color therefore mapped into the following coordinate of the 3D-model.
width width
xi;j ¼ i ; i 2 ½0::pixelsx ð1Þ
pixelsx 2
height height
yi;j ¼ j ; j 2 0::pixelsy ð2Þ
pixelsy 2
zi;j ¼ 1 ð3Þ
where pixelsx and pixelsy are the number of pixels in the sketch plane image in the
horizontal directions, respectively. The image size is given by
where R is the radius of the viewing sphere, set to 1 herein, and a and b are the
horizontal and vertical half-angle offsets, or angular sizes, of the sketch plane image.
Having obtained the 3D model, it is possible to render the model in various ways, such
as using a virtual reality viewer. In this work the 3D models were used to render
equirectangular panoramic images.
The panorama was build up as follows. First, any background panorama was
painted direction onto the target panorama first, and then all the points in the model
were rendered in decreasing distance to the observer. This is the well-known painter’s
algorithm where the foremost objects are added last such that one achieves hidden line
removal. The conversion from the Cartesian coordinate system to the geographical
coordinate system used in the equirectangular projection were as follows.
First, the point of intersection [x′, y′, z′] between the viewing sphere S and the line
going from the center of the sphere to an image pixel point [x, y, z] is computed as:
h 0 0 0i R
x ; y ; z ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ½x; y; z ð6Þ
x þ y2 þ z2
2
The point of intersection is thus defined by the vector with length R along the line
going from the sphere origin to the grid point. Finally, the intersection point [x′, y′, z′] is
transformed into geographical spherical coordinates [u, /] analogous to latitude and
longitude, using the following expressions
zamfira@unitbv.ro
1006 F.E. Sandnes
0 0
/ ¼ tan1 ðx ; y Þ ð7Þ
0
z
1
u ¼ sin ð8Þ
R
The arctan2 function is used as it reports angles in the full range from –180 to 180°
compared to the arctan function that reports angles between –90 to 90°.
The resulting panoramic images are imported into a standard panoramic viewer or
cloud point viewer. The panoramas rendered in this paper were generated with the
FSPViewer [11] and models were rendered with Cloud Compare [47].
zamfira@unitbv.ro
Sketching 3D Immersed Experiences Rapidly by Hand 1007
The proposed framework assumes that the viewer is located in the center, that is,
the viewer is in the center of the coordinate system. This is practical form the viewpoint
of constructing the 3D scene and for viewing using panoramic viewers. However, the
method is not limited to the panoramic views from one position. Once the
two-dimensional sketches are imported into the tool, the three-dimensional model can
be used to generate views from any location.
zamfira@unitbv.ro
1008 F.E. Sandnes
Fig. 3. Cage walls and roof (left) and cage floor (right)
4 Experiments
zamfira@unitbv.ro
Sketching 3D Immersed Experiences Rapidly by Hand 1009
Fig. 4. Panoramic images with a background panoramic image by James Kennedy Monash and
the cage sketch.
Figure 4 shows two panoramic images with the cage sketch superimposed. The top
panoramic image shows the cage centered around the origin, while the cage is lifted
relatively in the bottom panorama giving a strengthened sensation of the observer
sitting at the bottom of the cage.
Figure 5 shows renderings of the example panoramas presented herein using the
FSPviewer and Fig. 6 shows views of the resulting 3D model of the cage using the
CloudCompare point cloud viewer. The figure shows how the panoramic image is
modelled using a sphere (top left) and that this sphere provides a suitable background
when the model is viewed close to the center of the sphere. Figure 6 also demonstrates
the viewing distortions when the viewer is moved towards the border of the panorama
sphere wall where one can see both the closest side and the far side of the sphere (top
right).
zamfira@unitbv.ro
1010 F.E. Sandnes
Fig. 6. Real 3D views rendered using ClouldCompare based on the resulting point cloud model.
5 Conclusions
zamfira@unitbv.ro
Sketching 3D Immersed Experiences Rapidly by Hand 1011
rendered as panoramic images that were viewed with standard panoramic viewing
software. An advantage of the proposed approach is that the quality of the captured
three-dimension model is insufficient for professional production purposes. The models
therefore only exist with the purpose of being sketches during ideation.
References
1. Buxton, B.: Sketching User Experiences: Getting the Design Right and the Right Design:
Getting the Design Right and the Right Design. Morgan Kaufmann, San Francisco (2010)
2. Black, A.: Visible planning on paper and on screen: the impact of working medium on
decision-making by novice graphic designers. Behav. Inf. Technol. 9, 283–296 (1990)
3. Sandnes, F.E., Jian, H.L.: Sketching with Chinese calligraphy. Interactions 19, 62–66 (2012)
4. Landay, J.A., Myers, B.A.: Interactive sketching for the early stages of user interface design.
In: Proceedings of the SIGCHI Conference on Human Factors in Computing Systems,
pp. 43–50. ACM Press/Addison-Wesley Publishing Co. (1995)
5. Landay, J., Myers, B.: Sketching interfaces: toward more human interface design. Computer
34, 56–64 (2001)
6. Olsen, L., Samavati, F.F., Sousa, M.C., Jorge, J.A.: Sketch-based modeling: a survey.
Comput. Graph. 33, 85–103 (2009)
7. Kondo, K.: Interactive geometric modeling using freehand sketches. J. Geom. Graph. 13,
195–207 (2009)
8. Huang, Y.P., Wang, S.S., Sandnes, F.E.: RFID-based guide gives museum visitors more
freedom. IT Prof. Mag. 13, 25 (2011)
9. Xiong, Y., Turkowski, K.: Registration, calibration and blending in creating high quality
panoramas. In: Fourth IEEE Workshop on of Applications of Computer Vision (WACV
1998), pp. 69–74. IEEE Press (1998)
10. Kweon, G., Choi, Y.: Image-processing based panoramic camera employing single fisheye
lens. J. Opt. Soc. Korea 14, 245–259 (2010)
11. Senore, F.: FSPViewer. http://www.fsoft.it/FSPViewer/. Accessed 20 Nov 2015
12. Keane, T.P., Cahill, N.D., Rhody, H., Hu, B., Tarduno, J., Jacobs, R., Pelz. J.: Sphere 2:
Jerry’s rig, an OpenGL application for non-linear panorama viewing and interaction. In:
Image Processing Workshop (WNYIPW), pp. 13–16. IEEE Press (2012)
13. Anguelov, D., Dulong, C., Filip, D., Frueh, C., Lafon, S., Lyon, R., Ogale, A., Vincent, L.,
Weaver, J.: Google street view: capturing the world at street level. Computer 6, 32–38
(2010)
14. Kwiatek, K., Woolner, M.: Transporting the viewer into a 360 heritage story: panoramic
interactive narrative presented on a wrap-around screen. In: 16th International Conference on
Virtual Systems and Multimedia (VSMM), pp. 234–241. IEEE Press (2010)
15. Ku, D.C., Qin, S.F., Wright, D.K.: What is on the backside of the paper? From 2D sketch to
3D model. In: The 20th BCS HCI Group Conference, British Computer Society (2006)
16. Naya, F., Jorge, J., Conesa, J., Contero, M., Gomis, J.M.: Direct modeling: from sketches to
3D models. In: Proceedings of the 1st Ibero-American Symposium in Computer
Graphics SIACG, pp. 109–117 (2002)
17. Contero, M., Naya, F., Gomis, J.M., Conesa, J.: Calligraphic interfaces and geometric
reconstruction. In: 12th ADM International Conference on Design Tools (2001)
18. Cherlin, J.J., Samavati, F., Sousa, M.C., Jorge, J.A.: Sketch-based modeling with few
strokes. In: Proceedings of the 21st Spring Conference on Computer Graphics, pp. 137–145.
ACM (2005)
zamfira@unitbv.ro
1012 F.E. Sandnes
19. Matondang, M.Z., Mardzuki, S., Haron, H.: Transformation of engineering sketch to valid
solid object. In: Proceedings of International Conference of the 9th Asia Pacific Industrial
Engineering & Management Systems (APIEMS 2008) Conference and The 11th Asia Pacific
Regional Meeting of International Foundation for Production Research, pp. 2707–2715
(2008)
20. Eggli, L., Hsu, C., Brüderlin, B.D., Elber, G.: Inferring 3D models from freehand sketches
and constraints. Comput. Aided Des. 29, 101–112 (1997)
21. Contero, M., Naya, F., Jorge, J., Conesa, J.: CIGRO: a minimal instruction set calligraphic
interface for sketch-based modeling. In: Computational Science and Its Applications—
ICCSA 2003, pp. 549–558. Springer, Berlin Heidelberg (2003)
22. Huang, Y.P., Chang, T.W., Chen, Y.R., Sandnes, F.E.: A back propagation based real-time
license plate recognition system. Int. J. Pattern Recognit. Artif. Intell. 22, 233–251 (2008)
23. Huang, Y.P., Hsu, L.W., Sandnes, F.E.: An intelligent subtitle detection model for locating
television commercials. IEEE Trans. Syst. Man Cybern. Part B 37(2), 485–492 (2007)
24. Qin, S.F., Wright, D.K., Jordanov, I.N.: From on-line sketching to 2D and 3D geometry: a
system based on fuzzy knowledge. Comput. Aided Des. 32, 851–866 (2000)
25. Yang, C., Sharon, D., van de Panne, M.: Sketch-based modeling of parameterized objects.
In: EG Workshop on Sketch-Based Interfaces and Modeling, pp. 63–72 (2005)
26. Fiorentino, M., Monno, G., Renzulli, P.A., Uva, A.E.: 3D sketch stroke segmentation and
fitting in virtual reality. In: Proceedings of GRAPHICON, pp. 188–191 (2003)
27. Kim, D.H., Kim, M.J.: A new modeling interface for the pen-input displays. Comput. Aided
Des. 38, 210–223 (2006)
28. Shtof, A., Agathos, A., Gingold, Y., Shamir, A., Cohen-Or, S.: Geosemantic snapping for
sketch-based modeling. Comput. Graph. Forum 32, 245–253 (2013)
29. Roth-Koch, S.: Digitalization of paper sketches integration of the non digital draft. In:
Proceedings of the 2011 Conference on Designing Pleasurable Products and Interfaces,
p. 39. ACM (2011)
30. Kerautret, B., Granier, X., Braquelaire. A.: Intuitive shape modeling by shading design. In:
Smart Graphics, pp. 163–174. Springer, Heidelberg (2005)
31. Shin, H., Igarashi, T.: Magic canvas: interactive design of a 3-D scene prototype from
freehand sketches. In: Proceedings of Graphics Interface 2007, pp. 63–70. ACM (2007)
32. Igarashi, T., Matsuoka, S., Tanaka, H.: Teddy: a sketching interface for 3D freeform design.
In: ACM SIGGRAPH 2007 Courses, p. 21. ACM (2007)
33. Tai, C.L., Zhang, H., Fong, J.C.K.: Prototype modeling from sketched silhouettes based on
convolution surfaces. Comput. Graph. Forum 23, 71–83 (2004)
34. Tolba, O., Dorsey, J., McMillan, L.: Sketching with projective 2D strokes. In: Proceedings
of the 12th Annual ACM Symposium on User Interface Software and Technology, pp. 149–
157. ACM (1999)
35. Sandnes, F.E.: Communicating panoramic 360 degree immersed experiences: a simple
technique for sketching in 3D. In: Antona, M., Stephanidis, C. (eds.) Proceedings of HCI
International 2016, Universal Access in Human-Computer Interaction. Interaction Tech-
niques and Environments. LNCS, vol. 9738, pp. 338–346. Springer, Heidelberg (2016)
36. Sandnes, F.E.: Where was that photo taken? Deriving geographical information from image
collections based on temporal exposure attributes. Multimed. Syst. 16, 309–318 (2010)
37. Sandnes, F.E.: Determining the geographical location of image scenes based on object
shadow lengths. J. Signal Process. Syst. 65, 35–47 (2011)
38. Sandnes, F.E.: PanoramaGrid – a graph paper tracing framework for sketching 360-degree
immersed experiences. In: Proceedings of the International Working Conference on
Advanced Visual Interfaces (AVI 2016), pp. 342–343. ACM (2016)
zamfira@unitbv.ro
Sketching 3D Immersed Experiences Rapidly by Hand 1013
39. Sandnes, F.E., Huang, Y.P.: Translating the viewing position in single equirectangular
panoramic images. In: Proceedings of SMC 2016. IEEE Press (2016)
40. Shi, F., Laganiere, R., Dubois, E., Labrosse, F.: On the use of ray-tracing for viewpoint
interpolation in panoramic imagery. In: Canadian Conference on Computer and Robot
Vision (CRV 2009), pp. 200–207. IEEE Press (2009)
41. Kolhatkar, S., Laganiere, R.: Real-time virtual viewpoint generation on the GPU for scene
navigation. In: Canadian Conference on Computer and Robot Vision (CRV), pp. 55–62.
IEEE Press (2010)
42. Zhang, C., Zhao, Y., Wu, F.: Triangulation of cubic panorama for view synthesis. Appl. Opt.
50, 4286–4294 (2011)
43. Cohen, J.M., Hughes, J.F., Zeleznik. R.C.: Harold: a world made of drawings. In:
Proceedings of the 1st International Symposium on Non-photorealistic Animation and
Rendering. ACM (2000)
44. Fei, G.: 3D animation creation using space canvases for free-hand drawing. In: Proceedings
of the 7th ACM SIGGRAPH International Conference on Virtual-Reality Continuum and Its
Applications in Industry. ACM (2008)
45. Dorsey, J.: The mental canvas: a tool for conceptual architectural design and analysis. In:
15th Pacific Conference on Computer Graphics and Applications (PG 2007). IEEE (2007)
46. Greenberg, S., Carpendale, S., Marquardt, N., Buxton, N.: Sketching User Experiences: The
Workbook. Elsevier, Waltham (2011)
47. Girardeau-Montaut, D.: Cloudcompare-open source project. OpenSource Project (2011)
zamfira@unitbv.ro
Analyzing Modular Robotic Systems
Reem Alattas(&)
Abstract. This paper surveys modular robot systems, which consist of multiple
modules and aim to create versatile, robust, and low cost systems. The modu-
larity allows these robots to self-assemble, self-reconfigure, self-repair, and
self-replicate. Therefore, the surveyed research covered the previous charac-
teristics along with evolutionary robotics and 3D printed robots. These fields are
interdisciplinary, so we organize the implemented systems according to the main
feature in each one. The primary motivation for this is to categorize modular
robots according to their main function and to discover the similarities and
differences of implementing each system.
1 Introduction
Modular robots are composed of various units or modules, hence the name. Each
module involves actuators, sensors, computational, and communicational capabilities.
Usually, these modules are homogeneous; however they could be heterogeneous to
maximize versatility [1].
Modularity allows robots for self-assembly, self-reconfiguration, and self-repair [2]
those are discussed in Sects. 2, 3, 4 and 5 respectively. Section 6 presents self-reproduction
robots. Since developing methods for evolving controllers has been of great interest, we
cover evolutionary robots in Sect. 7, followed by printable robots in Sect. 8 and automatic
manufacturing in Sect. 9. To conclude, we summarize the article with Sect. 10.
2 Modularity
The concept of modularity has emerged in the past few decades, which led to suc-
cessfully implementing a number of prototypes. CEBOT is one of the first modular
robots. It was developed by Fukuda and Kawauchi in 1990, as a distributed robotic
system consisting of cells that could attach together to perform a function. CEBOT is
capable of dynamically self-reconfiguring and self-repairing. The cells are operated by
the communication network COMBUS [3]. Figure 1 shows the geometry of a mobile
and an object cell.
In 1993, Yim created a set of modular robots that can employ several locomotion
strategies [4]. Such systems are called self-reconfigurable robots; one example is
PolyBot that was implemented in 2000 to explore how realistic is to implement robots
using several homogeneous hardware modules. These modular self-reconfigurable
systems have three characteristics; versatility, robustness, and low cost. The first two
generations of PolyBot prove versatility by executing locomotion over a variety of
terrain. However, as the number of modules increases, cost increases, and robustness
decreases due to software scalability and hardware dependency issues. Currently the
maximum number of modules utilized in one connected PolyBot system is 32 with
each module having 1 DOF [5]. The third generation deals with 200 modules to show a
variety of capabilities, including moving like a snake, lizard or centipede as well as
humanoid walking and rolling in a loop [6].
Chiang and Chirikjian introduced a metaphoric system in 2001 to form structures
by rolling over each other in a plane. Also, a cost function was introduced to measure
reconfiguration fitness and to bisect shapes. This can be viewed as a geometric figures
pattern-matching problem under rigid body motions [7].
In the same year, Rus and Vona developed Crystalline atoms that have 3 DOF,
which allows expansion and contraction by a factor of two. Robots are formed by
expanding and contracting each atom frame in order to move relatively to the other
atoms. These movements simulate muscles actuation mechanism. Moreover, Crys-
talline robots are capable of self-reconfiguration very fast in O(n2) time, where n is the
number of atoms [8]. Earlier in 1998, Rus et al. have developed modules for building
self-reconfigurable robots called Molecules that support various locomotion modalities
by organizing autonomously as geometric structures to best fulfill the task in hand.
These Molecules have 2 DOF and can be aggregated to form 3D structures. Finally,
motion planning is done in O(n) time, where n is the number of molecules [9, 10].
Suh et al. in 2002 introduced the Telecubes that are cubic modules with 6 prismatic
DOF and sides capable of expanding more than twice its original length. Those cubes
can form a modular self-reconfigurable robot by attaching and detaching magnetically
to other cubes [11].
zamfira@unitbv.ro
1016 R. Alattas
As mentioned earlier, robotic modules are equipped with sensors in order to collect
data and provide necessary feedback that can be used locally on the module to guide
self-reconfiguration. Støy et al. proposed a methodology where raw sensor values can
be used globally, and combined it with role based control method for the
self-reconfigurable robot CONRO [12, 13].
Molecubes is an open hardware and software platform for modular robotics that
was developed to remove entry barriers to the field and accelerate progress. Different
types of active modules; such as gripper, actuated joint, controller, camera, and wheel
along with a number of passive modules were presented. Evolutionary search was to
design different types of robots rapidly [14, 15]. The following Tables 1, 2 and 3
compare the aforementioned systems on a number of parameters including geometrical,
electrical, and physical properties.
Chirikjian 2D 3 Hexagonal
Crystalline 2D 1 Square
Telecubes 3D 1 Cubic
CONRO 3D 2 None
Molecubes 3D 4 Cubic
zamfira@unitbv.ro
Analyzing Modular Robotic Systems 1017
3 Self-assembly
One of the main benefits of modularity is the capability of self-assembly, which is the
natural construction of complex multi-unit system using simple units governed by a set
of rules. Self-assembly process is ubiquitous in nature as it generates much of the living
cell functionality [16]. However, it is uncommon in technical field, because it is
considered as a new concept relatively in that arena although it could help in lowering
costs and improving versatility and robustness; which are the three promises of
modular robotics. The ability to form a larger stronger robot using smaller modules
allows self-assembly robotics to perform tasks in remote and hazardous environment.
Jones and Mataric in 2003 introduced intelligent self-assembly system using
assembly agents and a transition rule set compiler, which takes a goal shape as an input
and gives a set of rules as an output that can be utilized by the assembly agents to
assemble the target shape [17]. In additional work, Kelly and Zhang described a model
for optimizing the size of the rule sets used to build a structure [18]. Furthermore,
Werfel studied assembling complex structures using transition rule sets and artificial
swarms to automate construction [19].
Stochastically driven self-assembly systems were studied by White et al. in 2004 as
they developed algorithms and hardware for few systems. One system uses square units
with electromagnets that self-assembled into an L-shape and then into a line. The other
system uses triangular units with swiveling permanent magnets that self-assembled into
a line and then changed their sequence within the line. Both systems lack batteries, and
the modules only receive power after they connect to the structure being self-assembled
[20]. Tolley et al. extended that 2D system to 3D. Their evolutionary approach takes a
target function as input and yields a shape to achieve the input function, and directs the
shape’s assembly. However, the units are unable to move on their own as they need to
circulate in turbulent fluid to accrete onto the structure. This fluidic system could be
scaled down to produce micro-scale modules [21].
In 2005, Bishop et al. built triangular programmable parts that can be assorted on an
air table by overhead oscillating fans to self-assemble various shapes according to the
mathematics of graph grammars. The modules can communicate and selectively bond
using mechanically driven magnets, without global knowledge of the full shape.
Despite planning to build approximately 100 parts, only six parts were built for design
simplicity reasons. Those six parts were used in an experiment that showed these parts
react similarly to chemical systems [22]. Then, Napp et al. added measurements of
kinetic rate data to the previous work of graph grammar in order to yield a Markov
zamfira@unitbv.ro
1018 R. Alattas
zamfira@unitbv.ro
Analyzing Modular Robotic Systems 1019
4 Self-reconfiguration
Recently, modular robotics has gotten attention from researchers in the robotics field due
to their ability to self-reconfigure [2]. Modular self-reconfigurable robots involve var-
ious modules that can combine themselves autonomously into a meta-module or a
structure that is capable of performing a specific task under certain circumstances [1].
Self-reconfigurability allows these robots of metamorphosis, which in turn makes them
capable of performing different sorts of kinematics. For instance, a robot may recon-
figure into a manipulator, a crawler, or a legged one [2]. This sort of adaptability enables
self-reconfigurable robots to accomplish tasks in unstructured environments; such as
space exploration, deep-sea applications, rescue missions, or reconnaissance [3].
Yim et al. in 2002 classified reconfigurable robots into three classes of architecture:
lattice, chain, and mobile based on how they reconfigure [25]. Then, they added
deterministic and stochastic reconfigurations in 2007 [26].
Lattice architectures have modules that are connected in a 3D pattern that can be
used as a guide for modules to determine their positions and form the new shape
accordingly. Chain/Tree architectures have modules that are connected together in a
string or tree topology. The underlying architecture is serial such that each chain is
always attached to the rest of the modules at one or more points, and they reconfigure
by attaching and detaching to and from themselves. Mobile architectures change shape
by having modules detach themselves from the main body and move independently
[25]. Deterministic Architecture relies on units moving or being directly manipulated
into their target location during reconfiguration. Stochastic Architecture relies on units
moving around using statistical processes; e.g. Brownian motion, that are used to
guarantee reconfiguration times as well [26].
Since reconfigurable robotics field has a great interest in robotics community, we
have seen many prototypes implementations. Among them is M-TRAN (Modular
transformer) a distributed lattice-based self-reconfigurable system composed of
homogeneous robotic modules. The special design of M-TRAN module realizes both
reliable and quick self-reconfiguration and versatile robotic motion. M-TRAN is able to
metamorphose into robotic configurations such as a legged machine and hereby gen-
erate coordinated walking motion without any human intervention. The actual system
that was built using ten modules was examined through experiments to demonstrate the
basic operations of self-reconfiguration and motion generation. In order to drive
M-TRAN hardware, a series of software programs has been developed including a
kinematics simulator, a UI to design appropriate configurations and motion sequences
for given tasks, and an automatic motion planner for a regular cluster of M-TRAN
modules. These software programs are integrated into the M-TRAN system supervised
by a host computer [27].
In the second prototype, M-TRAN II, various improvements were integrated in order
to allow complicated reconfigurations and versatile whole body motions. Those
improvements contain reliable attachment/detachment mechanism, on-board multi-
computers, high speed inter-module communication system, low power consumption,
and precise motor control. Developed software are also integrated to design
zamfira@unitbv.ro
1020 R. Alattas
M-TRAN 3D 2 Cubic
zamfira@unitbv.ro
Analyzing Modular Robotic Systems 1021
5 Self-repair
Fitch et al. built on the previous work of Yoshida et al. to accomplish self-repair
using the self-reconfiguring Crystalline robots with a focus on geometric motion
planning. The aforementioned Crystalline robots consist of modules are actuate by
expanding and contracting, as shown in Fig. 5. This actuation mechanism is used for
self-repair as the process consists of three phases: detect failure, eject the failed module,
and replace the failed module [35].
zamfira@unitbv.ro
1022 R. Alattas
6 Self-reproduction/Self-replication
7 Evolutionary Robotics
zamfira@unitbv.ro
Analyzing Modular Robotic Systems 1023
not be captured by a model based approach. On the other hand, more complex tasks
expose limits of reactive architectures. However, very complex tasks such as garbage
collection and battery recharging show that emergent modular structures allowed the
decomposition of the global behavior into basic behaviors to emerge spontaneously.
Furthermore, the achieved decomposition did not correspond to a distal decomposition
an external designer would naturally expect, and outperformed other manually
designed decompositions [41].
According to Lipson, each robot comprises two major parts: controller (brain) and
morphology (body). Robot controllers can be represented in any one of a number of
ways: as logic functions, programs, differential equations, or neural networks. Various
experiments represent the controller as a neural network that maps sensory input to
actuator outputs. These neural networks can have different architectures, such as
feed-forward or recurrent. Sometimes the choice of architecture is left to the synthesis
algorithm [42].
Nolfi and Floreano described an experiment of using evolutionary methods to
evolve a controller that would make a legged robot, which is equipped with actuators
and sensors, locomote towards a high chemical concentration area [43]. Bongard
explored the same concept on a legged robot in a physically realistic simulator. The
robot has four legs and eight rotary actuators. A neural controller that maps sensors to
actuators determines the behavior of the machine. Trying out a candidate controller in
four different concentration fields, and summing up the distance between the final
position of the robot and the highest concentration point evaluated the fitness. The
shorter the distance is considered better. In this experiment, 200 candidate controllers
were evolved for 50 generations and the robot learned to move and to change direction
towards the high concentration [44].
Zykov et al. used evolving controllers for a real dynamical-legged robot in 2004.
The nine-legged machine, demonstrated in Fig. 6, is composed of two Stewart plat-
forms back to back. The author used force-actuators which exact extension can be set.
zamfira@unitbv.ro
1024 R. Alattas
The controller architecture for this machine was an open-loop pattern generator that
determines when to open and close pneumatic valves. The on-off pattern was evolved
and candidate controllers were evaluated by trying them out on a robot in a cage.
Fitness was measured using a camera that tracks the red ball on the foot of one of the
legs of the machine [45].
Paul and Bongard designed dynamic bipedal robot controllers in simulation using
evolutionary process. The robot consists of the bottom half of a walker with six motors,
a touch sensor at each foot and an angle sensor at each joint. Fitness was the net
distance a robot could travel. Evolving 300 controllers over 300 generations generated
numerous controllers that could make the machine move while keeping it upright.
These results may suggest that evolving a controller for a fixed morphology could be
too restrictive, while co-evolving both the controller and the morphology could yield
better results [46].
Karl Sims explored in simulation the idea of giving the evolutionary process more
freedom in designing both morphology and control using 3D cubes and oscillators as
building blocks [47]. Similarly, Lipson and Pollack explored physically-realizable
machines and started with lower-level building blocks, such as 1D elements and simple
neurons. The used design space was comprised of bars and actuators as building blocks
of structure and artificial neurons as building blocks of control.
8 Printable Robots
Existing rapid prototyping techniques; such as 3-D printing, are becoming increasingly
accessible due to their ability of achieving complex geometries. Therefore, printable
robots utilize these planar fabrication methods in order to create integrated
electro-mechanical laminates. Moreover, 3D Printing allows fabrication of low cost,
capable, agile, functional 3-D robots; such as Origami robots proposed by Onal et al. in
2014. Those robots can fold themselves into functional 3-D machines employing
origami-inspired techniques [48]. One of these robots is displayed in Fig. 7.
zamfira@unitbv.ro
Analyzing Modular Robotic Systems 1025
9 Automatic Manufacturing
10 Conclusion
In this paper, we have presented a comprehensive survey of modular robots that were
created to meet three main goals, versatility, robustness, and low cost. Also, modularity
offered a number of features that were used to differentiate types of modular robots from
self-assembly to self-repair. Self-assembly allows a number of modules to integrate and
form a robot. Self-reconfigurable robot is capable of changing its shape and locomotion
kinematics according to the task in hand. A robot that can fix itself by replacing dam-
aged modules with fresh ones is called self-repair. While the robot that can replicate
zamfira@unitbv.ro
1026 R. Alattas
itself is called self-reproducible. Since evolutionary robotics explore the design and
construction of robots using multiple modules, we covered it in this paper, followed by
printable robots and automatic manufacturing. Many representative works were selected
from the literature and some of the implemented prototypes were discussed.
References
1. Faíña, A., Bellas, F., López-Peña, F., Duro, R.J.: EDHMoR: evolutionary designer of
heterogeneous modular robots. Eng. Appl. Artif. Intell. 26(10), 2408–2423 (2013)
2. White, P., Zykov, V., Bongard, J.C., Lipson, H.: Three dimensional stochastic reconfig-
uration of modular robots. In: Robotics: Science and Systems, pp. 161–168 (2005)
3. Fukuda, T., Kawauchi, Y.: Cellular robotic system (CEBOT) as one of the realization of
self-organizing intelligent universal manipulator. In: IEEE International Conference on
Robotics and Automation (ICRA 1990), pp. 662–667 (1990)
4. Yim, M.: A reconfigurable modular robot with many modes of locomotion. In: JSME
International Conference on Advanced Mechatronics, Tokyo, Japan (1993)
5. Yim, M., Duff, D., Roufas, K.: PolyBot: a modular reconfigurable robot. In: IEEE
International Conference on Robotics and Automation (ICRA 2000), pp. 514–520 (2000)
6. Golovinsky, A., Yim, M., Zhang, Y., Eldershaw, C., Duff, D.: PolyBot and PolyKinetic™
system: a modular robotic platform for education. In: Robotics and Automation (ICRA
2004), vol. 2, pp. 1381–1386 (2004)
7. Chiang, C., Chirikjian, G.: Similarity metrics with applications in modular robot motion
planning. Auton. Robots (special issue on Modular Reconfigurable Robots) 10(1), 91–106
(2001)
8. Rus, D., Vona, M.: Crystalline robots: self-reconfiguration with compressible units modules.
Auton. Robots (special issue on Modular Reconfigurable Robots) 10(1), 107–124 (2001)
9. Kotay, K., Rus, D.: Motion synthesis for the self-reconfiguring molecule. In: IEEE/RSJ
International Conference on Intelligent Robots and Systems, pp. 843–851 (1998)
10. Kotay, K., Rus, D., Vona, M., McGray, C.: The self-reconfiguring robotics molecule. In:
IEEE International Conference on Robotics and Automation (ICRA 1998), pp. 424–431
(1998)
11. Suh, J.W., Homans, S.B., Yim, M.: Telecubes: mechanical design of a module for
self-reconfigurable robotics. In: Proceedings of IEEE International Conference on Robotics
and Automation, vol. 4, no. 5, pp. 4095–4101 (2002)
12. Støy, K., Shen, W.-M., Will, P.: On the use of sensors in self reconfigurable robots. In:
Hallam, B., Floreano, D., Hallam, J., Hayes, G., Meyer, J.-A. (eds.) Proceedings of the
Seventh International Conference on the Simulation of Adaptive Behavior, pp. 48–57 (2002)
13. Castano, A., Behan, A., Will, P.: The CONRO modules for self reconfigurable robots. IEEE
Trans. Mechatron. 7(4), 403–409 (2002)
14. Zykov, V., Chan, A., Lipson, H.: Molecubes: an open-source modular robotic kit. In:
IROS-2007 Self-Reconfigurable Robotics Workshop (2007)
15. Zykov, V., Phelps, W., Lassabe, N., Lipson, H.: Molecubes extended: diversifying
capabilities of open-source modular robotics. In: IROS-2008 Self-Reconfigurable Robotics
Workshop (2008)
16. Bishop, J., Burden, S., Klavins, E., Kreisberg, R., Malone, W., Napp, N., Nguyen, T.:
Programmable parts: a demonstration of the grammatical approach to self-organization. In:
Proceedings of IEEE/RSJ International Conference on Intelligent Robots and Systems
(IROS), pp. 3684–3691 (2005)
zamfira@unitbv.ro
Analyzing Modular Robotic Systems 1027
17. Jones,C., Mataric, M.J.: From local to global behavior in intelligent self-assembly. In:
Proceedings of IEEE International Conference on Robotics and Automation (ICRA),
pp. 721–726 (2003)
18. Kelly, J., Zhang, H.: Combinatorial optimization of sensing for rule based planar distributed
assembly. In: Proceedings of IEEE International Conference on Intelligent Robots and
Systems, pp. 3728–3734 (2006)
19. Werfel, J.: Anthills built to order: automating construction with artificial swarms. Ph.D.
dissertation, MIT (2006)
20. White, P., Kopanski, K., Lipson, H.: Stochastic self-reconfigurable cellular robotics. In:
Proceedings of IEEE Conference on Robotics and Automation, April 2004, pp. 2888–2893
(2004)
21. Tolley, M., Hiller, J., Lipson, H.: Evolutionary design and assembly planning for stochastic
modular robots. In: Proceedings of IEEE Conference on Intelligent Robotics and Systems
(IROS), October 2009, pp. 73–78 (2009)
22. Bishop, J., Burden, S., Klavins, E., Kreisberg, R., Malone, W., Napp, N., Nguyen, T.:
Programmable parts: a demonstration of the grammatical approach to self-organization. In:
Proceedings of IEEE/RSJ International Conference on Intelligent Robots and Systems
(IROS), August 2005, pp. 3684–3691 (2005)
23. Napp, N., Burden, S., Klavins, E.: The statistical dynamics of programmed self-assembly.
In: Proceedings 2006 IEEE International Conference on Robotics and Automation (ICRA
2006), pp. 1469–1476 (2006)
24. Wei, H., Cai, Y., Li, H., Li, D., Wang, T.: Sambot: a self-assembly modular robot for swarm
robot. In: Robotics and Automation (ICRA), pp. 66–71 (2010)
25. Yim, M., Zhang, Y., Duff, D.: Modular robots. IEEE Spectr. 39(2), 30–34 (2002)
26. Yim, M., Shen, W.M., Salemi, B., Rus, D., Moll, M., Lipson, H., Chirikjian, G.S.: Modular
self-reconfigurable robot systems (grand challenges of robotics). Robot. Autom. Magaz. 14
(1), 43–52 (2007)
27. Murata, S., Yoshida, E., Kamimura, A., Kurokawa, H., Tomita, K., Kokaji, S.: M-TRAN:
self-reconfigurable modular robotic system. IEEE/ASME Trans. Mechatron. 7(4), 431–441
(2002)
28. Kurokawa, H., Kamimura, A., Yoshida, E., Tomita, K., Kokaji, S., Murata, S.: M-TRAN II:
metamorphosis from a four-legged walker to a caterpillar. In: Intelligent Robots and Systems
(IROS 2003), vol. 3, pp. 2454–2459 (2003)
29. Kurokawa, H., Tomita, K., Kamimura, A., Kokaji, S., Hasuo, T., Murata, S.: Distributed
self-reconfiguration of M-TRAN III modular robotic system. Int. J. Robot. Res. 2(3–4), 373–
386 (2008)
30. Salemi, B., Moll, M., Shen, W.M.: SUPERBOT: a deployable, multi-functional, and
modular self-reconfigurable robotic system. In: Intelligent Robots and Systems, pp. 3636–
3641 (2006)
31. Østergaard, E.H., Kassow, K., Beck, R., Lund, H.H.: Design of the ATRON lattice-based
self-reconfigurable robot. Auton. Robots 21(2), 165–183 (2006)
32. Wieser, S.: Locomotion in Modular Robotics: Roombot Module. Semester project,
Biologically Inspired Robotic Group (2008)
33. Murata, S., Kurokawa, H., Kokaji, S.: Self-assembling machine. In: Proceedings of the 1994
IEEE International Conference on Robotics and Automation, San Diego (1994)
34. Yoshida, E., Murata, S., Tomita, K., Kurokawa, H., Kokaji, S.: An experimental study on a
self-repairing modular machine. Robot. Auton. Syst. 29, 79–89 (1999)
35. Fitch, R., Rus, D., Vona, M.: A basis for self-repair robots using self-reconfiguring crystal
modules. Intell. Auton. Syst. 6, 903–910 (2000)
zamfira@unitbv.ro
1028 R. Alattas
36. Lackner, S., Wendt, C.H.: Exponential growth of large self-reproducing machine systems.
Math. Comput. Model. 21(10), 55–81 (1995)
37. Ulam, S.: Random presses and transformations. In: Proceedings of the International
Congress of Mathematicians, vol. II, Cambridge, MA (1950)
38. von Neumann, J.: Theory of Self-Reproducing Automata. University of Illinois Press,
Urbana (1966). Edited and completed by A.W. Burks
39. Griffith, S., Goldwater, D., Jacobson, J.M.: Robotics: self-replication from random parts.
Nature 437, 636 (2005)
40. Zykov, V., Mytilinaios, E., Desnoyer, M., Lipson, H.: Evolved and designed
self-reproducing modular robotics. IEEE Trans. Robot. 23(2), 308–319 (2007)
41. Nolfi, S., Floreano, D.: Evolutionary Robotics – The Biology, Intelligence, and Technology
of Self-Organizing Machines. MIT Press, Cambridge (2000)
42. Lipson, H.: Evolutionary robotics and open-ended design automation. Biomimetics 17(9),
129–155 (2005)
43. Floreano, D., Husbands, P., Nolfi, S.: Evolutionary robotics. In: Siciliano, B., Khatib, O.
(eds.) Handbook of Robotics, pp. 1423–1451. Springer, Heidelberg (2007)
44. Bongard, J.C., Pfeifer, R.: Evolving complete agents using artificial ontogeny. In: Hara, F.,
Pfeifer, R. (eds.) Morpho-Functional Machines: The New Species, pp. 237–258. Springer,
Tokyo (2003)
45. Zykov, V., Bongard, J.C., Lipson, H.: Evolving dynamic gaits on a physical robot. In:
Proceedings of Genetic and Evolutionary Computation Conference (GECCO 2004) (2004)
46. Paul, C., Bongard, J.C.: The road less traveled: morphology in the optimization of biped
robot locomotion. In: Proceedings of the IEEE/RSJ International Conference on Intelligent
Robots and Systems (IROS 2001), Hawaii, USA (2001)
47. Sims, K.: Evolving 3D morphology and behavior by competition. In: Artificial Life IV,
pp. 28–39 (1994)
48. Onal, C.D., Tolley, M.T., Wood, R.J., Rus, D.: Origami-Inspired Printed Robots (2011)
49. Lipson, H., Pollack, J.B.: Automatic design and manufacture of robotic lifeforms. Nature
406(6799), 974–978 (2000)
zamfira@unitbv.ro
An Educational Physics Laboratory in Mobile
Versus Room Scale Virtual Reality
- A Comparative Study
1 Introduction
The improvement of science education is still a topic under frequent discussion
in the world today. In physics education in particular in, the situation is two-
fold: many teachers are challenged in teaching concepts to an increasing number
of students, who in turn often face issues themselves in trying to understand
the concepts taught while linking theoretical formulas to natural phenomena.
Engaging and interesting students in this increasingly relevant issue in our edu-
cational system is thus matter of the utmost importance. Emerging technologies,
such as virtual simulations and laboratories in VR provide novel ways to engage
and interest students in class while at the same time also giving educators more
possibilities to create and improve classroom experiences.
Simulations and dynamic visualizations can be used to make invisible con-
cepts visible, stretch time and space, and conduct dangerous or even impossible
experiments [2,14]. While earlier studies suggest that the use of simulations can
c Springer International Publishing AG 2018
M.E. Auer and D.G. Zutin (eds.), Online Engineering & Internet of Things,
Lecture Notes in Networks and Systems 22, DOI 10.1007/978-3-319-64352-6 95
zamfira@unitbv.ro
1030 J. Pirker et al.
In the following sections we will first discuss related work on STEM education
with focus on virtual reality experiences. After that we shortly describe Maroon,
the virtual laboratory developed for the experiments. In Sect. 4 two different
user studies on mobile and room-scale experiences in VR are presented.
2 Related Work
Designing STEM education in an interesting and engaging manner still repre-
sents a challenge. One successful pedagogical method for teaching practice in
regular classrooms is “active learning”. In this method, students not only listen
passively to the concepts, but they are also directly involved in the learning
process. This has been shown to be an effective strategy for increasing the stu-
dents performance compared to traditional methods [10,15]. In physics education
zamfira@unitbv.ro
An Educational Physics Laboratory 1031
zamfira@unitbv.ro
1032 J. Pirker et al.
main concept to the learning experience could be used to add new ways to
create professional and interesting working and learning environments. The use
of virtual reality headsets and technologies is a promising way to create a more
immersive, engaging, and interactive environment. With the current efforts to
produce VR headsets which are affordable for private users (e.g. PlayStation
VR, Samsung Gear VR, HTC Vive), VR is also becoming more attractive as
a tool to enhance classroom experiences. Several studies have looked into the
potential of virtual reality (VR) for educational scenarios.
In this paper, we introduce Maroon, an interactive immersive physics labo-
ratory, integrated with (1) the interactive virtual reality technology HTC Vive,
supporting in-room movement and a two controller setup and (2) a mobile setup
with the Samsung Gear VR.
zamfira@unitbv.ro
An Educational Physics Laboratory 1033
including the Samsung Gear VR. For the HTC Vive, the official SteamVR2 plu-
gin and framework was used. This lab prototype was the design basis for the two
VR variants. From originally six implemented simulations in the context of elec-
tromagnetism for this setup, only two were integrated in this initial prototype
for the study.
In our setup, users can experience VR in two distinct ways on two conceptu-
ally different devices: either through a mobile, more light-weight setup (Samsung
Gear VR, using the Samsung Galaxy S6) or a more graphically rich, advanced
room scale system tracking both HMD and controllers (HTC Vive using two con-
trollers). In particular for user interaction, navigation, manipulation and selec-
tion of UI elements with the virtual world, two different design approaches were
chosen, considering various limitations and the different design of these two VR
devices.
The version of the immersive physics lab Maroon as introduced is designed
to support both mobile VR systems such as Google Cardboard or Samsung Gear
VR3 running on mobile phones as well as more advanced setups with roomscale
VR such as the HTC Vive.4 The designed interaction with the environment
and the experiments is mostly performed through gaze for the Samsung Gear
VR and via controllers for the HTC Vive. Samsung Gear VR additionally pro-
vides possibilities to interact through touch and slide input, whereas the HTC
Vive benefits from several buttons on both its tracked controllers which can be
specifically programmed and also visually adapted for individual user actions.
The navigation designs for the two VR alternatives are discussed in more detail
in the following.
Navigation Design in Mobile VR. Given the Samsung Gear VR system with
the smartphone inserted into a head-mounted gear, a real-life like user experience
is achieved through a combination of eye gaze, a virtual avatar and a touchpad
2
http://store.steampowered.com/steamvr.
3
http://www.samsung.com/global/galaxy/gear-vr/.
4
https://www.htcvive.com/.
zamfira@unitbv.ro
1034 J. Pirker et al.
mounted on the side of the device, with user actions such as double tap, long
press and swipe to rotate. Here, the user controls are mostly designed for gaze
and tap interactions. An avatar (see Fig. 1) is controlled with a gaze point to
move through the laboratory. The avatar is always placed on the gaze point -
the center of the screen - and can be moved by moving the gaze. Simulations
can be started by moving the gaze cursor to the interaction button. Movement is
designed as teleporting the avatar to different locations. Sliding (only supported
by Samsung Gear VR) can be used optionally to rotate the character or to move
specific controls (sliders) of experiments.
Navigation Design in Room Scale VR. In contrast, the HTC Vive sys-
tem consists of a larger HMD connected to the PC as well as two additional
controllers, which include a highly-sensitive touchpad and individually program-
mable buttons with haptic feedback for improved user interaction within virtual
worlds. Each hardware element in the Vive setup is tracked by two base stations
named lighthouses, thus eliminating the need for an avatar and further enabling
the user to move around freely for a more immersive room-scale VR experience.
Simulations are started by entering a portal-like object through button press on
the controllers. Movement as in teleporting is achieved by pressing the touchpad
on one of the controllers, which in turn acts like a pointer, as the user aims at
the preferred target and displays a precise colored beam for visual orientation.
Concerning the experimental setup, the main difference between the imple-
mentation for Samsung Gear VR and HTC Vive was the addition of interactable
objects in the HTC Vive version and its lack of a virtual avatar which was instead
implemented in the Samsung Gear version for better usability. By using several
programmable controller buttons as well as touchpad press, HTC Vive users are
able to benefit from further real-life like interaction possibilities. The necessity
of a virtual avatar was not given for these since users carry both HMD and
controllers which are being tracked by the lighthouse system.
zamfira@unitbv.ro
An Educational Physics Laboratory 1035
visualized (see Fig. 2) and (3) a whiteboard with information and labeled pic-
tures to explain the theory behind the Van De Graaff experiments. In order to
showcase the manifold possibilities of user interaction with virtual objects using
controller mechanisms, the HTC Vive version of this station additionally features
an interactive playground with different textured objects such as throwable and
grabbable cubes and metal balls. (4-only HTC) A triboelectric experiment with
two rods and one balloon as well as a miniature version of the previous Van
de Graaff experiment, however, this was only fully implemented for the HTC
Vive test setup. Hence, to achieve more diversity in our experimental setting,
this specific station was replaced by another station on the Samsung Gear VR
version where it features a laptop with an interactive, feedback-supported quiz
session in order to test the theoretical knowledge users should have gained with
their practical hands-on walk-through of Maroon Mobile VR. (5-optional) Addi-
tionally, an accurate model of a Tesla transformator can be found by users as a
hidden “easter egg” by further exploring the virtual laboratory world.
In our research, these two conceptually different VR setups provide the frame
for our implementation of the interactive immersive physics laboratory. Ulti-
mately, the goal in developing these simulations is to let users act more or less
the same way they were would if placed in a real-life physics laboratory. As of
now, users are - to some extent - able to immerse themselves into this world while
being shielded from (visual) influences of their actual physical surrounding. As
such, immersive 3D has shown to be a beneficial aid to present difficult concepts
in physics, such as the effect of switching a Van De Graaff generator on and off.
4 User Studies
We performed two preliminary user studies with a total of 17 participants to
evaluate the system and the experience. In a first study (with 9 participants) we
focused on testing the Maroon with the mobile setup only. In the second study
(with 8 participants) we focused on evaluating (1) engagement, (2) immersion,
(3) learning experience, (4) virtual reality experience, and (5) usability and user
experience in comparison to a more interactive VR experience with the HTC
VIVE.
The VR setup for Samsung Gear VR consists of the following hardware compo-
nents: mobile HMD and smartphone Samsung Galaxy S6. Figure 3a shows the
Samsung Gear VR with the attached mobile phone. The setup for HTC Vive
contains the HMD itself, cables and two base stations as well as two controllers.
For a room-scale setup setting, we provided an area of about 2 m × 2 m. Further-
more, a powerful high-end hardware PC is necessary. A mobile VR setup was
chosen in order to support a widely accessible and cost-effective way to interact
with the laboratory, which could be used in classroom environments (e.g. guided
by an instructor), or for self-regulated learning at home.
zamfira@unitbv.ro
1036 J. Pirker et al.
(a) (b)
For the first study with Samsung Gear VR, we first asked the participants to fill
out a pre-questionnaire. The pre-questionnaire was used to get information about
the participants experience with virtual experiences, and VR technologies, and
their expertise in physics. They were then were introduced briefly to the system.
After this they were asked to use the Maroon Mobile VR with the Samsung Gear
VR. After the experience, the participants shortly described their impressions in
form of an open dialog. Finally, they were asked to complete a post-questionnaire
with 10 open-ended question on the experience and 20 single-choice questions
with ratings on a Likert scale between 1 (fully disagree) and 7 (fully agree).
In the other extended study with both devices, participants were required to
fill out a short pre-questionnaire with standard personal background information,
followed by a brief introduction to the experimental setup. The main goal was then
to complete consecutive tasks in the immersive lab Maroon, which were announced
by the study moderator during the test run. Since we examine the differences and
similarities of both devices, our eight test subjects were divided into two sepa-
rate groups of four persons each for the purpose of AB/BA testing where users
test both devices in reverse order. (Specifically, four users tested the Vive first;
whereas the other four tested the Samsung Gear VR first.) After each single run,
users completed a corresponding post-questionnaire containing 19 standardized
questions from the Game Engagement Questionnaire (GEQ, [5]) to measure the
level of engagement based on absorption, flow, presence, and immersion, as well
as ten open-ended questions on the experience and 20 single-choice questions with
ratings on a Likert scale between 1 (fully disagree) and 7 (fully agree). For a com-
parative evaluation, all subjects had to complete a “combined” post-questionnaire
with open-ended questions about their experience on both devices at the end of
the experiment (Fig. 4).
zamfira@unitbv.ro
zamfira@unitbv.ro
Fig. 4. Survey results of experience with Maroon Mobile VR between 1 (not at all) and 7 (fully agree) in GEAR VR
An Educational Physics Laboratory
1037
1038 J. Pirker et al.
4.3 Participants
Experiment 2. In the second study 8 (1f) participants were asked to test the
mobile (Maroon Mobile VR) and the interactive physics lab (Maroon Room
Scale VR). 7 are very experienced in the use of computers (AVG = 4.38;1.41),
only 2 in the usage of video-games (AVG = 3;1.2), and only 1 in VR (AVG =
2.25;1.39). 4 have used a mobile VR setup before, nobody the HTC VIVE. 7
rated their physics knowledge a 3 or below (AVG = 2.63;0.92).
In the following sections we discuss different aspects of the outcomes of the
post-questionnaires and the interviews. The individual aspects will be mainly
described by including outcomes of the questionnaire and direct quotes describ-
ing the students impressions and experiences. An overview of the results can also
be found in Fig. 1.
Most of the participants said they find learning in this manner more engaging
(AVG = 6.67; SD = 0.82) and fun (AVG = 6.33;0.82). When being asked if
they find it engaging and motivating, most of them agreed: “very motivating
way of demonstrating stuff ”. The lack of content and variety was mentioned as a
drawback here: “Not yet, but I can see how the concept would be engaging once
more variety exists.” When asked what they liked about the system, immersive
and three-dimension characteristics were mentioned in particular: “Immersion
makes me remember stuff better”. The VR experience was received very positive
and described as very immersive. In the second part of the study we compared
presence, absorption, flow, and immersion between interactive VR experience
(with the HTC Vive setup and the mobile setup. As seen in Fig. 6 the interactive
version achieves only slightly better results in all 4 categories.
In the first part of the study, on a Likert scale between 1 (not at all) and 7
(fully agree) most of the people questioned said they would like to learn with
Maroon Mobile VR (AVG = 5.33; SD = 1.51) and feel that the content is easier
zamfira@unitbv.ro
An Educational Physics Laboratory 1039
zamfira@unitbv.ro
1040 J. Pirker et al.
for learning include the topic choice (“It’s good for demonstrating something,
maybe not as good for learning facts etc., because you can’t for example take
notes etc.”) and additional overhead. The VR aspect was very well received for
learning. Participants thought it was engaging to see the physics simulations
with the VR glasses (AVG = 6.5,0.55) and also a bit more engaging than with-
out VR (AVG = 5.83;0.98) “learning with VR is gonna be awesome and I never
thought about what happens to a balloon if we place him between a Tesla-coil and
a grounder Funny”.
In Fig. 5 we compare the above mentioned results with the results of the
second part of the study. Again, we can see that the interactive VR experiences
achieved slightly better results compared to the mobile experience. However, in
the first experiment with the mobile device only, the results were slightly better
for the mobile setup compared to the second experiment. This could explained
by a bias through the interaction with the HTC Vive setup.
zamfira@unitbv.ro
An Educational Physics Laboratory 1041
experience was mentioned as a drawback by some students. “If the target group
is Cardboard users then theres not much to improve graphics wise I think. Maybe
having a narrator voice explaining things or physic concepts to the user would
be nice.” Several participants mentioned that they would like to see more exper-
iments and simulations in the world and that the lab still looks very empty “I
think one can learn a lot, however, many experiments or models are required for
that”, “It was nice, a bit empty, not very realistic looking, but nice.” “the VR
technology itself needs to be improved. Higher resolution and lenses will make
a huge difference. The game it self was, except of some teething troubles, well
done. The controls are good, maybe improvable with a controller. But all in all I
liked it.”
The study was designed to get insights to improve the current prototype with
focus on engagement, immersion, and learning outcomes. The first study only
focuses on testing the mobile experience and was also used to evaluate the VR
study design for the second study. Based on these finding, the prototype will be
updated and a large-scale study with more participants is designed.
zamfira@unitbv.ro
1042 J. Pirker et al.
study results. To fully explore the potential of such environments we are cur-
rently extending the laboratory with other forms of simulations with different
educational goals. Additionally, we are planning to study further the effects on
learning of the VR experience of the laboratory, also in comparison to the same
desktop experience.
Acknowledgment. We would like to thank John Winston Belcher from the Depart-
ment of Physics Massachusetts Institute of Technology for supporting this research.
The Maroon projects is a research project at Graz University of Technology:
gamelabgraz.com/maroon/. We thank all people who are and were involved in the
development process.
References
1. Adams, W.K., Reid, S., LeMaster, R., McKagan, S.B., Perkins, K.K., Dubson, M.,
Wieman, C.E.: A study of educational simulations part 1-engagement and learning.
J. Interact. Learn. Res. 19(3), 397 (2008)
2. Aldrich, C.: Learning Online with Games, Simulations, and Virtual Worlds: Strate-
gies for Online Instruction, vol. 23. Wiley, San Francisco (2009)
3. Bell, R.L., Smetana, L.K.: Using computer simulations to enhance science teaching
and learning. Natl. Sci. Teachers Assoc. 3, 23–32 (2008)
4. Bonde, M.T., Makransky, G., Wandall, J., Larsen, M.V., Morsing, M., Jarmer, H.,
Sommer, M.O.: Improving biotech education through gamified laboratory simula-
tions. Nat. Biotechnol. 32(7), 694–697 (2014)
5. Brockmyer, J.H., Fox, C.M., Curtiss, K.A., McBroom, E., Burkhart, K.M.,
Pidruzny, J.N.: The development of the game engagement questionnaire: a measure
of engagement in video game-playing. J. Exp. Soc. Psychol. 45(4), 624–634 (2009)
6. Corter, J.E., Nickerson, J.V., Esche, S.K., Chassapis, C., Im, S., Ma, J.: Construct-
ing reality: a study of remote, hands-on, and simulated laboratories. ACM Trans.
Comput.-Hum. Interact. (TOCHI) 14(2), 7 (2007)
7. Csikszentmihalyi, M., Csikszentmihalyi, I.S.: Optimal Experience: Psychological
Studies of Flow in Consciousness. Cambridge University Press, Cambridge (1992)
8. Dori, Y.J., Belcher, J.: How does technology-enabled active learning affect under-
graduate students’ understanding of electromagnetism concepts? J. Learn. Sci.
14(2), 243–279 (2005)
9. Dori, Y.J., Hult, E., Breslow, L., Belcher, J.W.: How much have they retained?
Making unseen concepts seen in a freshman electromagnetism course at MIT. J.
Sci. Educ. Technol. 16(4), 299–323 (2007)
10. Freeman, S., Eddy, S.L., McDonough, M., Smith, M.K., Okoroafor, N., Jordt, H.,
Wenderoth, M.P.: Active learning increases student performance in science, engi-
neering, and mathematics. Proc. Natl. Acad. Sci. 111(23), 8410–8415 (2014)
11. OMICS International: Van de Graaff Generator (2014). http://research.
omicsgroup.org/index.php/Van de Graaff generator
12. Lindsay, E., Good, M.: Virtual and distance experiments: pedagogical alternatives,
not logistical alternatives. In: American Society for Engineering Education, pp. 19–
21 (2006)
13. Lowe, D., Murray, S., Lindsay, E., Liu, D., Bright, C.: Reflecting professional reality
in remote laboratory experiences. In: Proceedings of International Conference on
Remote Engineering and Virtual Instrumentation (REV 2008) (2008)
zamfira@unitbv.ro
An Educational Physics Laboratory 1043
14. Lunce, L.M.: Simulations: bringing the benefits of situated learning to the tradi-
tional classroom. J. Appl. Educ. Technol. 3(1), 37–45 (2006)
15. Olson, S., Riordan, D.G.: Engage to excel: producing one million additional col-
lege graduates with degrees in science, technology, engineering, and mathematics.
Report to the president, Executive Office of the President (2012)
16. Pirker, J., Berger, S., Guetl, C., Belcher, J., Bailey, P.H.: Understanding physical
concepts using an immersive virtual learning environment. In: Proceedings of the
2nd European Immersive Education Summit, Paris, pp. 183–191 (2012)
17. Pirker, J., Gütl, C.: Educational gamified science simulations. In: Gamification in
Education and Business, pp. 253–275. Springer (2015)
18. Pirker, J., Gütl, C., Belcher, J.W., Bailey, P.H.: Design and evaluation of a learner-
centric immersive virtual learning environment for physics education. In: Human
factors in computing and informatics, pp. 551–561. Springer (2013)
19. Settgast, V., Pirker, J., Lontschar, S., Maggale, S., Gütl, C.: Evaluating experiences
in different virtual reality setups. In: International Conference on Entertainment
Computing, pp. 115–125. Springer (2016)
20. Wieman, C., Perkins, K.: Transforming physics education. Phys. Today 58(11), 36
(2005)
21. Windschitl, M.A.: Using computer simulations to enhance conceptual change: the
roles of constructivist instruction and student epistemological beliefs (1995)
zamfira@unitbv.ro
Human Interaction Lab: All-Encompassing Computing
Applied to Emotions in Education
1 Introduction
Today, working with human behavior is vitally important, especially if we consider the
impact neuroscience has had in our understanding of learning processes. The idea is to
discover the part of behavior that is essential in organizing learning that occurs in the
brain, where emotions become involved. In this paper we analyze emotions in relation
to the opinions students have about the work teachers give them. Then, we try to deter‐
mine if that work along with tests are the appropriate tools in evaluating output or if they
are simply something they have to comply with as part of their curriculum.
We center our study on facial expression and writing as they are deposits of positive
and negative emotions. In fact, we can firmly say that a camera, depending on the algo‐
rithm used, can capture more information than the human eye. By following a specific
process, a camera directed to a student’s face can help us identify what students think
about the work they have at hand. This is where the neuroscience lab comes into play.
It is a place that is dedicated to identify emotional human behavior through the signals
received on cameras that are located in the Faculty of Human Sciences and Education.
The scientific background that supports our work includes the following:
zamfira@unitbv.ro
Human Interaction Lab 1045
The following theoretical entries allowed us to determine the efficiency of open educa‐
tion resources OER1 used by a teacher in a specific subject area. We were able to use
them to measure their emotional impact and polarity right when they were being imple‐
mented in the classroom. This allowed us to obtain original answers from the students
with regards to the material handed out by the teacher. There is a direct estimated distri‐
bution for teacher satisfaction regarding the academic performance of the students
(Marcenaro-Gutierrez et al. 2015). In fact, the distribution of academic performance can
be checked by contrasting the teacher’s interest with his students’ opinions about race
(Hanushek and Rivkin 2009). We maintain that this proposal is feasible if we want to
determine whether teaching and learning performance indicators are appropriate in
measuring the academic performance of a student. There are several methodologies that
utilize traditional techniques for Natural Language Processing (NLP) together with
sentiment analysis processes and Web Semantic technologies. Their main objectives are
to improve results for opinion mining, which are based on specific characteristics, e.g.
by employing ontologies for the selection of techniques as well as providing a new
1
OER.- <www.oercommons.org/>.
zamfira@unitbv.ro
1046 H.F. Gomez Alvarado et al.
method for sentiment analysis that is based on vector analysis. It is comprised of four
main modules: the Natural Language Processing module (NLP), the module for the
identification of characteristics based on ontologies, the module for the identification of
polarity, and the module for opinion-based data mining (Peñalver Martínez et al. 2011).
In (Pang and Lee 2004), Pao and Lee propose a novel machine learning method that
applies techniques for the categorization of text- namely its subjective contents. The
extraction of these subjective parts is achieved by using efficient techniques to determine
minimum reductions (cessations) in the networks- thereby facilitating the incorporation
of contextual sentences. Therefore, a methodology has been proposed to facilitate the
usage of a subjectivity detector, which determines whether each phrase is subjective or
not- thus highlighting the objectives and creating an extract of critical and subjective
content for the classification of a predetermined polarity. We developed an application
that would enable us to identify emotions while a person was writing a text. All this
information is recorded in a video for each participant. In our study, however, we did
not need to focus so much on the details of the photo image seeing that we had to correlate
the sentiment analysis with the emotion that the person is displaying. For this reason,
the results that we obtained during the experimental stage are promising. Below are
details of the methodology that was used and the experimental phase that was carried
out by this research team (Gomez et al. 2015).
3 Methodology
First, we installed the lab to collect the data. Cameras that centered on participants’ faces
were placed in hallways and classrooms throughout the Faculty to identify emotion
patterns. We also took into account safety as the cameras can also be used to identify
criminal acts. Figure 1 shows one of the cameras located at one of the halls of the Faculty.
Fig. 1. Camera installed at the hall of the Faculty of Human Sciences and Education.
The monitor that can be seen in the Fig. 1 gives us a broad view of the cameras. It
also allows us to see if the software is running correctly. There are appropriate research
backup cameras not only in the Faculty in Ambato. The camera signals are processed
by the prototype software found at the Human Faculty on the Behavior Lab. The general
distribution of the cameras in the Faculty’s Lab is shown in the following diagram. The
zamfira@unitbv.ro
Human Interaction Lab 1047
methodology being used is the development of prototypes that can provide us with soft‐
ware that can used to obtain camera signals which, after being processed, can give us
emotion analysis data. The following is an example of a face captured by the prototype
(Fig. 2).
4 Experimentation
For the indicators of academic performance, we worked with students that took class
during the 2015 semesters. The selection process and the classes involved do not require
a special statistical process since the proposed method is based on 4000 emotional
training faces that were previously included in the database. During the training phase
people were allowed to try out the prototype software to identify their emotions. Based
on that data we were able to train the prototype software and measure its validity in
recognizing emotions in students’ faces. The following table shows the results obtained,
with 10 cameras in classrooms, respect at students emotions recognition:
Figure 3 show that recall are better that precisión with 0.83 with maximal value. The
precision has 0.73 of maximal value.
zamfira@unitbv.ro
1048 H.F. Gomez Alvarado et al.
5 Conclusions
It’s imperative to take into account the opinion of students in the learning process. In
this sense, we developed tools applied to education in its Human Behavior Lab. These
can be used to accurately understand if the materials and the academic performance
indicators are the ones to be used in the process. Thus, students and teachers are the
direct beneficiaries of this project because it contributes to the follow-up process that is
part of teaching and learning. The universities in the study have advancing scientific
research through projects such as the one described here. The Faculty of Human Sciences
and Education has taken advantage of the mechanisms of intervention provided by the
security system put in place, but, in this case, as it is being applied in education. The
recycling of electronic devices benefits the University, Science and the research that is
being carried out in the country. New areas of research are being created which open up
new areas of study, because the values of precision and recall are aceptables in relation
with (Arunnehu and Kalaiselvi 2016). This time it is being done with the participation
of students who we consider to be the scientific community par excellence. That is, new
research topics, with a focus on neurosciences as well as Alzheimer (Arias Tapia et al.
2016), are being born at the Human Behavior Lab, with a special emphasis on Education
(Gomez et al. 2015). The beneficiary of these new projects is the population considered
to be vulnerable in the countries.
Acknowledgments. We thank UTA for supporting the Research Project Human Performance
Laboratory.
References
Peñalver Martínez, I., Valencia García, R., García Sánchez, F.: Minería de opiniones basada en
características guiada por ontologías. Sociedad Española para el Procesamiento del Lenguaje
Natural
Arias Tapia, S., Martínez-Tomás, R., Gómez, H., Del Salto, V., Guerrero, J., Mocha-Bonilla, J.,
et al.: The dissociation between polarity, semantic orientation, and emotional tone as an early
indicator of cognitive impairment. Frontiers in Computational Neuroscience (2016)
Arunnehu, J., Kalaiselvi, G.: Automatic human emotion recognition in surveillance video. En
Intelligent Techniques in Signal Processing for Multimedia Secutiry, pp. 321–342. Springer
(2016)
Gomez, A.H., Arias, T.S., Torres, P., Sanchez, J., Hernandez, V.: Emotions analysis techniques:
their application in the identification of criteria for selecting suitable Open Educational
Resources. In: International Conference on Interactive Collaborative and Blended Learning.
Mexico (2015)
Hanushek, E., Rivkin, S.: Harming the best: how schools affect the blackwhite achievement gap.
Natl. Bur. Econ. Res. 28, 366–393 (2009)
Liu, B.: Web Data Mining Exploring Hyperlinks Contents and Usage Data. University of Illinois,
Chicago (2011)
Marcenaro-Gutierrez, O., Luque-Gallego, M., Lopez-Aguado, L.: Teacher’s satisfaction as
indicator of education system performance. In: EAEE 2015, Alicante (2015)
zamfira@unitbv.ro
Human Interaction Lab 1049
Pang, B., Lee, L.: A sentimental education: Sentiment analysis using subjectivity summarization
based on minimum cuts. In: Proceeding ACL, New York (2004)
Park, E., Lee, Y.: Emotion-based image retrieval using multiple queries and consistency feedback.
In: Conference on Industrial Informatics, New York (2008)
Rodriguez, P., Ortigosa, A., Carro, R.: Detecting and making use of emotions to enhance student
motivation in e-learning environments. Int. J. Continuing Eng. Educ. Life Long Learn. 24(2),
168–183 (2014)
zamfira@unitbv.ro
Distance Learning System Application for Maritime
Specialists Preparing and Corresponding Challenges
Analyzing
Vladlen Shapo ✉
( )
1 Introduction
Last years Ukrainian universities implement learning management systems (LMS) very
actively. In connection with impossibility to buy quite expensive software the choice
usually stops on free and open source LMS Moodle. It renews regularly, it has conven‐
ient interface and a lot of possibilities to work for students, teachers, teaching and
methodical materials (TMM) developers [1, 2]. Bit by bit some books appear in different
languages where procedures of distance courses creation, TMM elaboration and LMS
administration are described with different levels of specification [3–5].
Significant difference of National University “Odessa Maritime Academy” (NU
“OMA”) [6] from a lot of different another universities is in presence of remote struc‐
tured subdivisions in different parts of Ukraine: in the city of Izmail (Izmail faculty, IF)
and in the city of Mariupol (Azov maritime institute, AMI). Moreover thousands of full
time students, students by correspondence, postgraduate students, advanced training
courses students and trainees every year pass many month naval training on different type
of commercial vessels and need to have access to TMMs and teacher’s consultations,
being far from the home and NU “OMA”. Before these possibilities could be realized by
passing paper or electronic TMMs to vessels during crew members change (unexpect‐
edly and unreliable), sending electronic TMMs via e-mail or downloading electronic
TMMs from NU “OMA”’s web site [6] (these ways were realized during 13 years without
any statistics, registration, efficiency analyzing, etc.). Using LMS Moodle (in test mode
it has begun in September 2009, in full mode it has begun in September 2010) has signif‐
icantly allowed to enhance quality of TMMs preparing, and also theirs quantity and
zamfira@unitbv.ro
Distance Learning System Application 1051
actuality, and to prepare teachers and students step by step for regular using of Moodle
possibilities wide spectrum.
Scheme of distance information interaction between participants of educational
process which is realized in NU “OMA” is presented on Fig. 1.
zamfira@unitbv.ro
1052 V. Shapo
It’s known that in most cases students begin studying of disciplines from school
books, which are created by universities’ teachers, where they study. Precisely these
school books much simpler (in comparing with books of another authors and franchisors)
to place in LMS because by this way it’s possible to minimize problem of content piracy
and another juridical aspects.
At present time distance learning system of NU “OMA” is based on Moodle 1.9.6 [5]
software and provides work of more than 8400 users (including 320 teachers). For fast
registration of big number of users special additional software is created. Using of LMS
allows to get additional positive results in studying of interested students [1, 2]. Activity
and interest of LMS using by students of different studying forms are growing permanently.
2 Methodology
In any information system can happen some fault or failure that’s why it’s necessary to
store backup copy of LMS folders structure with created distance courses and another
materials and database of users. When moving LMS Moodle from one physical server
to another one there is a possibility to export all data from previous server in SQL format
(database dump) and to import all these data to new server. But during LMS exploitation
zamfira@unitbv.ro
Distance Learning System Application 1053
server will store new distance courses with a lot of different methodical materials and
tests in the frame of hundreds or even more than thousand disciplines, archive files which
contain backup copies of each course will be added, new users will be added as well,
thousands of users will upload different files, pictures and will take part in forums and
so on. As a result the volume of stored data will grow very fast, the volume of export
SQL file will grow very fast as well and will take place the problem of importing database
dump at new server even from administrator’s console on server computer without LAN
using. Time of import will exceed network (browser) time out value and process of
importing will be interrupted itself without finishing. Concurrently level of hardware
utilization of quite modern typical computer configuration is very high (50–94% of 2-
cores CPU resources and 700–1200 MBytes of RAM). Also LMS Moodle uses CRON
scenario (system job) which works in endless cycle checking and processing all new
events in LMS and adds significant CPU and RAM loading.
Graphics of computer resources utilization with small, middle and high loading are
shown at Figs. 2, 3 and 4 accordingly. These graphics contain following information:
CPU utilization (50, 76 and 94% accordingly) and chronology of CPU utilization in the
top part of graphics; swap file size (696, 941 MBytes and 1.21 GBytes accordingly) and
chronology of swap file using.
zamfira@unitbv.ro
1054 V. Shapo
In the same time repeated manual data input which is necessary to restore LMS is
quite primitive, very boring, long and laborious process which can take between some
days and some weeks of LMS administrator work depends on his qualification and
experience. Herewith it’s undesirable to attract any staff for execution of these works
inside LMS in connection with necessity of full administrator rights granting to addi‐
tional people and unpredictable results of further work because of human factor (fully
unworking LMS in unexpected moment, appearance of faked users, plagiarism of
different learning materials an so on).
Some tasks and problems which have to be solved in university information systems
are described in [7].
It’s necessary to formulate some recommendations for LMS Moodle administrators
based on own 6 years long experience.
1. Not to add LMS users one by one manually even when number of these users is not
too big. It’s much preferably to create external text files with full list of user’s data
using special template, to add new users in the end of this file and to execute auto‐
matic procedure of importing by means of LMS Moodle, because file exported from
LMS which contains the user’s data will not contain user’s passwords and in the
case of LMS fault requires some additional manual processing of exported files
zamfira@unitbv.ro
Distance Learning System Application 1055
which requires additional time. Also this way involves the necessity of new pass‐
words generation and user’s informing about this situation with naturally arising
dissatisfaction and mess, or searching and restoring of previous versions of pass‐
words.
2. Before exporting of full database dump it’s necessary to delete beforehand all archive
copies of distance courses from the server, because presence of these files signifi‐
cantly increases database dump (sometimes even in times).
3. It’s necessary to restrict file sizes (especially for graphical files), which will be
uploaded on the server by users.
4. Not to import database dump to the server by LAN. Generally the most of Ukrainian
universities use Fast Ethernet with 100 Mbps bandwidth as LANs. Just LAN data
transfer speed will be as bottle neck because even bandwidth of old hard drive
connection interface IDE is 100 MBps (theoretically 8 times faster), and modern
SATA interface has at least 150 MBps. At the same time the same LAN will be used
by another users which will take some part of bandwidth as well. So, during
importing of database to the server directly from administrator console LAN will
not become as restricting factor.
5. To store archive copies of each distance course on external hard, optical, flash drives,
streamers and so on. It will make procedure of data restoring for separate distance
courses or whole database much simpler and faster.
6. During database restoring on new server at first it’s necessary to restore only cate‐
gories and subcategories structure and to create empty distance courses without
restoring of their content, and to make database dump. Such dump will have rela‐
tively small size and can be simple and fast restored, being the skeleton of LMS.
7. Before importing of big database dump it’s necessary to enlarge values of following
parameters: max_execution_time, max_input_time, memory_limit,
default_socket_timeout, mysql.connect_timeout, session.gc_maxlifetime in configura‐
tion file php.ini.
Last 10–15 years it’s become absolutely clear that Life long learning (LLL) concep‐
tion is necessary to be realized by any engineer, developer, valuable specialist working
in the field of automation, industry, transport, data transfer and control systems, etc. One
of the facilities to get new knowledge, practical skill, experience, etc. is using of training
equipment.
Main lack of any training equipment is big or huge cost because modern training
equipment consists of high performance computers, network equipment, very expensive
touch panels with big diagonals (at least 40 in) and high resolution and so on. Typically
training equipment may cost several tens, hundreds or even millions dollars depending
on functionality and sphere of application. In the same time even super modern equip‐
ment becomes morally outdated very fast because of appearance of new hardware, soft‐
ware, technologies, network protocols, concepts, algorithms and so on. That’s why from
the economical point of view such training equipment has to be used 24 h per day. Thus
it’s necessary to choose optimal hardware (memory volume and performance, disk
subsystem type and performance, network and graphical interfaces bandwidth, central
and graphic processors productivity, network technology and data transfer rate) config‐
uration taking into consideration cost/productivity ratio.
zamfira@unitbv.ro
1056 V. Shapo
Additionally it’s very complex to realize training equipment loading during whole
day in real life. Moreover very often trainees, which have to pass corresponding training,
live in different cities and even countries, have different level of language speaking,
work in different companies (for example, in maritime branch). These reasons quite
often don’t allow to get all them together. That’s why very important property of training
equipment is possibility to work with trainee and to be controlled by administrator and/
or trainee remotely.
One more task to be solved is Internet channel bandwidth optimal choosing and
analyzing and calculating of additional loading in corporate (campus) computer
network. Transferring of uncompressed graphical data will be reason for unstable
network condition. Fast Ethernet network technology with 100 Mbit/s data transfer rate,
which is most popular in Ukrainian campus networks, will become the bottle neck. There
are 2 evident ways to solve this problem.
1. Using of additional real time data compression boards which will contain own CPU
and memory.
2. Upgrading of network equipment (switches) from Fast Ethernet to Gigabit Ethernet
or even to 10 Gigabit Ethernet for some network segments and structured cabling
system or some it’s segments from category 5 twisted pair to newer twisted pair
category or fiber optics.
3 Conclusions
In this paper some challenges and problems connected with modern university learning
management systems software and hardware are touched on and some recommendations
on exploitation of such systems are proposed. Currently learning management system
is a part of more complex enterprise information systems, it can be combined with
another software like ERP-, CRM-, BI-systems, complex computerized training systems
for seafarers within the same computer network and hardware using virtualization tech‐
nologies in particular. Inasmuch in present time learning and self-learning is life long
process in any field of human activity and is very actual and will become much more
dynamic, these recommendations may be useful not only for specialists in education
system but for specialists which create and exploit different information systems as well.
References
1. Vinnikov V., Shapo V.: Qualified personnel preparing in logistics field with distance learning
systems using. In: INCEL 08: E-learning in Higher School – Problems and Perspectives: Works
of International Conference. Odessa National Polytechnic University – National Technical
University, Kharkiv Polytechnic Institute (2008). http://cde.kpi.kharkov.ua/tempus/incel/
2. Vinnikov V., Shapo V.: Specialists preparing in logistics field with distance learning
application. In: Strategy of Quality in Industry and Education: Works of 4th International
Conference, V.2, pp. 536–539. Technical University, Varna (2008)
3. Miasnikova, T.S., Miasnikov, S.A.: Distance Learning System MOODLE. Publishing House
of Sheinina E, Kharkiv (2008)
zamfira@unitbv.ro
Distance Learning System Application 1057
4. Anisimov, A.: Working in Distance Learning System Moodle. Kharkiv National Academy of
City Management, Kharkiv (2009)
5. MOODLE project. https://moodle.org/mod/data/view.php?id=7246
6. National University “Odessa Maritime Academy”. http://www.onma.edu.ua
7. Shapo V.: Building of complex university information system. In: Problems of Information
Society Development: Materials of VI International Scientific-Practical Conference,
INFORMATIO-2009: Electronic Information Resources: Creating, Using, Access and XII
International Scientific-Practical Conference, Building of Information Society: Resources and
Technologies, Kiev, pp. 120–123 (2009)
zamfira@unitbv.ro
Author Index
zamfira@unitbv.ro
1060 Author Index
zamfira@unitbv.ro
Author Index 1061
zamfira@unitbv.ro
1062 Author Index
zamfira@unitbv.ro