A. Shukla, H.

Karki / Robotics and Autonomous Systems 75 (2016) 490507

it is 142 M boe/d (million barrels of oil equivalent per day), conventional resources of petroleum products are shrinking very fast and
remaining oil and gas fields are characterized by adjectives such as
arctic, deepw ater, cold, heavy, high in w ater content, high sulfur
content, to name but a few [5]. Further this scarcity of conventional
resources forcing oil companies to go for exploration of new nonconventional petroleum reserves such as heavy oil, tight gas, shale
gas and coalbed methane etc. Modern econom y is highly coupled
w ith oil prices therefore cost-effectiveness of the production is an
important key factor to the future of this industry and global economy. M ost of the production fields are not exploited fully because
after certain level of recovery, production cost does not justify further investments of stakeholders moneys. This increasing supply
demand and difficult oil fields have not only increased the cost of
production, but also compounded the risks related to human security and environmental safety.
Fossil fuels are hydrocarbons trapped in porous rocks and sediments in underground conditions [6]. Extraction of fuel from early
onshore resources only required shallow drilling in highly pressurized free flow ing w ells. Although there are many historical instances for usage of oil in ancient times but first professional oil
w ell in m odern time w as developed in 1859 by Edw in Drake in
northw estern Pennsylvania, USA [7]. Over the period of time as
natural pressure of crude oil reduces due to extraction its become
difficult to recover complete volume of fossil fuel due to technical and economical challenges. But w ith increasing prices and dependence on fossil fuels, many advance technology such as w ater
injection, gas lift and steam injection, as secondary and territory
recovery methods, have become economically viable to increase
recovery rate and extract more crude from non-free flow ing resources [8]. According to [6] w e now consume four barrels of
know n reserves for every barrel of new ly added reserve. Therefore, most of the giant oil fields (roughly 500 around the w orld)
w hich currently contributing 65%of the w orlds oil supply are almost now 50 years old and fast declining in production capacity keeping w orld oil production relatively flat [9]. With almost
stagnant conventional onshore production, compensation for deficiency betw een demand and supply coming from either increasing
offshore fields and non-conventional petroleum reserves such as
shale oil, shale gas, tight oil, heavy oil, tar sands coal bed m ethane
and m ethane hydrate etc. Revolution in production of these nonconventional resources of oil and gas in USA by new ly advanced
technology has generated an economic shock w ave in international
oil and gas market. For an example in the USA most part of unconventional fossil fuel supply is coming from horizontal drilling and
hydraulic fracturing of shale gas and tight oil [10] as show n in Fig. 1.
In the US alone production of shale gas has grow n from 12 million
m 3 in 2002 to 275 million m 3 in 2012, w hich accounts for almost
40%of the total natural gas production [11]. There are pros and cons
of production of shale gas as opponents w arn for w ater pollution,
earth quake and methane emission, supporters show reduction in
emission of green house gases as usage of coal gets reduced due to
usage of more cleaner natural gas in electricity production [10,11].
Since M ost of the new oil and gas fields are found in extreme environmental conditions such as harsh deserts, deep-w ater, frozen
arctic zones and deep below ground level. Extraction, processing
and transportation of petroleum products in adverse geographical
and environmental conditions pose serious challenges to environment, m arine lives and human health. M any serious accidents involving oil spills, leaks, fires, explosions, toxic emissions and w ater
solution has taken past in future. Recently happened terrible offshore accident of Deep Horizon oil spill in the Gulf of M exico [12]
has shocked the w orld and w arned us to be extremely careful of
HSErelated issues. Though this accident w as in offshore conditions
but many severe accidents has taken place in onshore facilities as
w ell and one such accident that shook USA has happened in July

491

Fi g. 1. Horizont al drilling and hydraulic fracturing [11].

2010 w hen a 30-inch carbon steel pipeline, operated by Enbridge

Energy Partners, Canadas largest crude oil transporter, ruptured
w hile carrying DilBit (short for Diluted-Bitumen extracted from oil
sand [13,14]). These accidents w ent unnoticed for next 17 h until it released almost 3 million liters of DilBit into a tributary of
the Kalamazoo River in Michigan. Whole clean-up operation costed
company till now around 1 billion USD but even after 2 years of
herculean efforts river is not completely clean w hile leaving permanent displacement of nearby communities [15]. Transportation
of DilBit in to conventional oil pipelines is a very dangerous idea
since it is by nature very corrosive, acidic and unstable m ixture of
volatile gases and thick raw bitumen [16]. And leak detection system across the industry is not yet reliable as mentioned in [17] that
oil spill data maintained by federal regulators show that over the
last 10 years, advanced leak detection systems identified only one
out of every 20 reported pipeline leaks [15]. In the w ake of several
terrible oil spill crisis in Europe, European Commission has already
funded several research projects w ith main objective of developing innovative intelligent robotic technologies for oil spill management [18,19].
2. Relat ed r esear ch
With increasing energy demand for ever grow ing urban
population of the w orld more pipeline netw orks are required
for transportation of fossil fuels in different stages of its
consumption cycle. Since petroleum products are sensitive and
found in extreme environmental conditions extra efforts are
required for continuous inspection and monitoring of the every
component involved its extraction, production, processing and
finally distribution. To w ork efficiently in such a intensive
environment w ith all hum an limitations, robotic assistance to
human cognition is extremely essential. Overall robotic assistive
strategy for condition monitoring, inspection and control strategy
of sensitive oil and gas industry can be break dow n in many
subproblems such as humanmachine interface [2,12,2022],
data-signal transmission [2326], resource allocation and task
scheduling [2730], navigation technologies [3133], localization
of the mobile robots and w orkspace-objects [3438], inspection
technologies [3945] and teleoperation [46] etc. Even after rightly
an efficiently solving all these subproblem s integration of all these
subsystems is an another challenge [47].

492

A. Shukla, H. Karki / Robotics and Autonomous Systems 75 (2016) 490507

Fossil fuels being main source of energy for economic activities in todays w orld, oil and gas industry has become fundamental
base for global economy. Our future highly depends on sustainability of this industry w hich in turn depend on safe, efficient and
cost-effective handling of conventional and non-conventional reserves. Unlike other industry, advancem ent of oil and gas industry is very critically linked to environmental safety and economic
grow th of the w orld at least in foreseeable future. There are many
scientific and industrial activity w hich have already greatly benefited by automatization such as space exploration and m anufacturing industry. Success of robotics in these areas can be an inspiration
for present oil and gas industry w hich is still in realm of manual operation or semi-automation at m axim um [48]. Improving
HSE standard w hile performing humanely impossible tasks w ith
high precision, such as in-pipe inspection or diving in ultra-deepw ater zones, is the main driving force behind robotization of oil
and gas industry in contrast to the manufacturing industry w here
increasing production and cost-efficiency are the prime goals
[46,49]. Unlike manufacturing industry w here design of production line, product and production environment are almost static or
w ell defined [50], in oil and gas industry, problem of producing explosive products is uniquely coupled w ith extreme environm ent.
Therefore, seeing the level risks involved, exactly same procedure
of m anufacturing automation cannot be replicated in oil and gas
industry, w ithout keeping the extra layer of cautions. For an example teleoperated intervention by trained operator in field operations from distantly located control center is now becoming
acceptable in oil and gas industry. In such cases trained human operator located at safe and comfortable location function as brain
of system w here as field robots become just an extended physical body of the operator. Various audio, visual and tactile sensors
become extended ear, eye and touch sensation for the operator
[33,48]. Biggest draw back of robots currently used in oil and gas
industry is there lack of desired autonomy, robustness and dependability. As of now most of these robots w ork in highly supervised short-term m issions w ith strong backing from researched
team [51].
Remote operation or teleoperation is mechanism w here master operator manipulate remotely located robot via communication channel. This remote operation can be of three categories,
firstly manual remote control, secondly semi-autonomous control
and thirdly completely autonom ous control. Since, materials involved in Oil and Gas industry are highly sensitive usage of fully autonomous robotic technology is risky and still farfetched solution.
Therefore, humanly supervised semi-autonomous robotic technology like teleoperation finds perfect m atch of expectations for
this industry. Intelligent drilling, under w ater exploration, smart
inspection and manipulation of pipes and tanks, and automated
operations for final production are some of the key areas w here
usage of teleoperated mechanism w ill not only increase safety
standards but also help to increase production as w ell. Keeping
all these issues in mind, present w ork describes a novel teleoperator m echanism to be used in oil and gas industry. Teleoperation technology has already produced exciting results in various
fields, such as offshore oil and gas explorations [5153], space explorations [5456], military [57], under-w ater exploration [58,59],
medical application [60,61], entertainment [62], and hazardous
environments [63,64]. Teleoperation, tele-manipulation, telerobotics and tele-inspection etc. are just different categories of
fundamentally same technology w ith some minor differences in
over architectural arrangement of system com ponents Agba:1995.
Teleinspection (sub-class of teleoperation) is related to usage of
mobile robots equipped w ith different kinds of sensors and devices [33,65] to perform the inspection and data collection [66].
In the current practices of surveillance, expensive sensors are installed only on critical areas of operations and non-critical areas are left out w ithout any regular supervision. Whereas, in the

case of mobile inspection, all the critical areas can be inspected

w ith higher frequency and non-critical areas w ith low er frequency,
leading to cost efficiency and optimally supervised plant.
Activities in the oil and gas industry is categorized mainly in
five stages, first stages is called as Exploration: it include seism ic
analysis and exploratory drilling, second stage is Upstream: this
refer to all the facilities related to production of oil and gas,
third stage is Midstream: broadly related to gas treatment, LNG
production regasification plants and oil and gas pipeline system,
fourth is Refining: w here oil and condensates are processed in
to marketable products involving chemical purification and tank
storage and fifth stage is producing by-products such as plastic,
fertilizer and other chemicals [67]. Now adays stage of production
heading for some degree of automatization and robotics assistance
to make it more efficient w hile having specific focus on facility
integrity management to im prove HSE parameters. There are tw o
broad categorization of the oil and gas facilities depending on
the site of exploration and production of fossil fuels. Underw ater
oil and gas fields are called as offshore fields w hich are further
divided in shallow and deep-w ater categories. Oil and gas fields
situated on land are generally called as onshore fields. Alm ost 67%
of the w orlds oil production comes from onshore oil fields [68].
Since many of the giant oil fields are either declining in production
or it had become economically unviable to continue further
extraction [9]. In such situations it has become essentially critical
to employ proven advanced robotic technology for higher, safer
and efficient onshore oil and gas production. Successful production
of oil and gas from a onshore reserve by a petroleum company is
generally described in follow ing basic steps:
Ini t i al explor at ion This step involves careful and extensive
risk analysis of the region found suitable for serious
exploration and production of fossil fuels by a company
before investing capital.
Leasi ng Once initial exploration establishes decent chances of
finding crude oil and gas at intended site w hich lead
to financial and energy benefits for all its stakeholders,
company goes for applying license to get legal rights for
further exploration.
Geogr aphi cal sur vey Once leasing process is completed company
goes ahead to develop a picture of geological structures
below the earth surface by seismic surveys. This process
also involves detailed mapping of region by collecting
and testing geological sam ples. Usage of vibroseis (source
of generating vibration), geophones, gravimeter and
magnetometer are quite common to generate accurate
patterns of rock formations under earth surface to know
w here oil and gas deposits have a good chance of being
found.
Dr il li ng Even after getting positive signals for presence of reservoir, final confirm ation and it commercial evaluation
comes from drilling exploratory w ells. After success of
initial exploratory w ell additional appraisal w ells are also
drilled to determine quality and quantity of hydrocarbons in the reservoir.
Well pr oduct i on Once exploratory drilling process confirms safe
and economic feasibility of the overall project, w ell
is developed for final production. In this phase all
the structures required for extraction, storage and
transportation of crude oil and gas are installed.
Ter m i nat i ng pr oduct ion Every reservoir goes to dow nhill in
terms of production after passing the peak of maximum
production. After some period w hen dow nhill production
becomes economically infeasible production is stopped,
site is cleaned by dismantling production infrastructures
and reserve pits are filled.

A. Shukla, H. Karki / Robotics and Autonomous Systems 75 (2016) 490507

493

not only inconvenient but also expensive. Nondestructive testing

(NDT) of these flaw s holds the key for any future development in
this field. In-pipe robots are proposed [71] as one of the solution
to this problem. These IPIRs are inserted in the pipe from a inlet
point and travel inside the pipe under external supervision. There
are various causes and kinds of corrosion, and for them there
are appropriate inspection techniques, such as visual inspection,
X-ray, eddy currents, acoustics and ultrasonics. These robots are
equipped w ith sonar and acoustics based leakage detector, HD
pantiltzoom camera and LED light panel etc. for the purpose of
inspection and sending critical data back to control center. Most
of the IPIR are teleoperated and connected by tethered cable to
the operator at remote location [72,73]. There are five essential
parameters to categorize IPIR:

Fi g. 2. W heel robot [86].

These steps of exploration and production are almost sim ilar in

both the onshore and offshore oil and gas facilities and differences
come from the infrastructure and technologies involved in the
overall process. Offshore oil and gas facilities being located in
sensitive and difficult environment poses even more serious
challenges to the safety of human lives than onshore. Innovative
technologies are required to deal these challenges w hile meeting
ever increasing demand of energy production. Since, it is
established by various researches that usage of robotics is now
almost compulsory for oil and gas industry therefore com ing
sections summarize the application of robotics in oil and gas
industry both at onshore and offshore facilities.
3. Robot i cs i n onshor e condi t i ons
M ost of the onshore oil and gas facilities, such as El M erk oil
and gas development in Algeria, lies in inhospitable deserts w here
the logistical constraints of deploying equipment, materials and
labor across multiple w ork fronts are parts critical challenges [69].
Oil and gas facilities have extensive usage of all kinds of
pipes and storage tanks during different stages of business
starting from exploration, extraction, transportation, processing
and distribution. Pipes and storage tanks need regular inspection
and maintenance especially those continuously used for long
distance transportation and long term storage. Humanly inspecting
these components are expensive and hazardous, so autom ated
inspection and manipulation for these components are very much
desired. M ost of the robotic research for onshore facilities, has been
dedicated for development of in-pipe inspection robots (IPIRs) and
tanks inspection robots.
3.1. Pipe inspection
At onshore oil and gas facilities, pipes are used as a tool for
transportation of oil, gas and other fluids, from production sites
to distribution sites. These pipes are m ostly laid dow n under
w ater or underground environment. In such environment pipes
are subjected to extreme w eather conditions such as extreme
temperature and pressure, humidity, dust and vibration. These
unfavorable conditions lead to m any troubles in pipes such as
corrosion, erosion, deposition, cracks, thermal cycling, pitting,
shock loading and joint-failure etc. [70]. Any kind of leakage of
petroleum products from pipes not only cause loss of revenue but
also invite ecological disaster. Therefore, regular inspection and
maintenance of transportation pipes are strongly demanded for
safe operation. Traditional w ay of digging and manually detecting
the temporal position of these flaw s for underground pipes are

3.1.1. Shape and size of robots

There are pipes of various shapes (straight line, elbow -shape
and T-shape etc.) and sizes (different radii) involved in the
oil and gas industries according to usage and flow conditions.
Therefore, shape and size of pipes are on of the major parameter
to affect the design of IPIR for example micro-robots [7480] are
required for sm aller size pipes. There are mainly tw o type of
mechanism for adaptation of varying radii of the pipes namely
active linkage type and passive linkage type [81]. In active linkage
type, separate actuators are installed to apply norm al force to
generate required traction pow er [82]. Therefore, this m echanism
requires more space and expensive from m anufacturing point of
view . Whereas, passive linkage mechanism is designed merely
w ith elastic components such spring. This arrangement leads to
simplicity of control mechanism and inexpensive manufacturing
of robot. Inspection robots described in [8386] have used simple
spring on the main axis of robot.
3.1.2. Steering mechanism
M ost of the robots are designed to pass through horizontal
pipe structures but urban gas pipe lines have complicated structures due to complex distribution netw orks. Therefore, successful navigation through these transportation pipes requires in-pipe
robots to steer through complex shapes such as vertical, elbow and
branched. There are mainly tw o categories of steering m echanism
firstly articulated type and secondly differential type. Articulated
steering is a mechanism w hich allow s a robot to take turn in restricted space around a pivot located on robot body by splitting the
total body of robot in front half and back half. This kind of steering mechanism is directly inspired by movement of the snake and
the annelid anim al in nature. There are several kind of articulated
steering mechanism depending how steering is activated around
pivot point such as steering joint [8789], double active universal
joint [90,91], and rubber gas actuated joint [92]. There are several
successful inspection robots based on differential steering mechanism as w ell [9399]. Differential drive is mechanism w hich allow s the turning of vehicle by modulating the speeds of the w heels
depending on the desired direction for turning [93,95]. Since, inside the pipe traction surface is three-dimensional curved rather
than planer, hence it requires sophisticated speed control mechanism to avoid slipping of w heels inside the pipe [86]. These requirements for know ledge of pipe geometry and locus of contact point
make this mechanism complicated problem from control point of
view . Fig. 2 show s an active universal joint steering m echanism in
a w heel type IPIR along w ith its different components.
3.1.3. Propelling mechanism
Every IPIR must have some propelling mechanism to navigate
inside the pipe apart from steering mechanism . There are many

494

A. Shukla, H. Karki / Robotics and Autonomous Systems 75 (2016) 490507

Fi g. 3. Classificat ion of in-pipe robot s. (a) Pig t ype. (b) Wheel t ype. (c) Cat erpillar
type. (d) W all-press. (e) W alking type. (f) Inchw orm t ype. (g) Screw t ype [86].

different kinds of navigation technologies used by in-pipe robots

depending on the requirements of the operations such as Pig
(Pipeline Inspection Gauge) type [100,101], Wheel type [73,76,
83,86,90,91,93,95,102110], Snake type [111114], Legged mobile
type or Walking type [115118], Caterpillar type [70,81,119],
Wall-press type [90,105,119], Inchw orm type [74,75,77,78,92,
120130], and Screw type [74,125,126,131,132] etc. Fig. 3 show s
classification of IPIRs depending on their propulsion mechanisms.
Pig type robot is collection of capsule type metallic body,
rubber disk for support from pipe w alls, ultra-sonic transducers
for detection, and odom eter for distance calculation. Rubber disk
not only get support from the w all but also blocks the inpipe fluid to get passively driven by it [100]. Draw back of this
mechanism is that, robot keep on rotating around its central
axis. Passively driven in-pipe robots perform poor on various
occasions such as w hen fluid pressure inside pipes are low and
transportation pipes have complex shapes (T-shape and L-shape).
Wheel type inspection robots are proposed to overcom e these
shortcomings [83]. Wheel based robots have many advantages
such as easy speed, direction control and higher energy efficiency
but suffer from complex steering mechanism and instability during
navigation. To overcome this instability, springs are used to press
the in-pipe w alls against w heels of robot [105]. This pressing of
w alls gives extra friction force for better slip-less operation and
adaptability of robot to different sizes of transportation pipes. [90]
describes a mechanism in w hich legs of the robot, w hich contain
w heels, are stretched and contracted radially to generate w all
pressure. This pressure gives all the above mentioned benefits and
saves body of robot from distortion forces w hen it crosses over
obstacles such as steps, reducers, protrusions inside the pipes. It
can be observed that most of these w heel based inspection robots
generate traction pow er by pressing the pipe w all passively or
actively. These w heel based robots get stuck inside the pipe w hen
there are sharp corners, steps, sudden big change in pipe crosssection and variable surface roughness. As a alternative to w heel
based locomotion, legged type robots has been proposed in [115].
The higher flexibility of legs, improves the performance of robot
crossing the obstacle and turning around joints and corners [118].
Design of in-pipe robots for the inspection of pipes of small
diam eters by using w heel based locom otion mechanism is really
a challenge. Wheel based robots require heavy motors and
gear systems, and usage of other kind of actuators such as

piezoactuators suffer from low er traction forces [131]. Therefore,

tw o locomotion mechanisms free from electrical machines are
proposed firstly inch-w arm type robots and secondly screw type
robots, for inspection of small diameter pipes. Earth w orm is a
creature w hich moves forw ard by sw elling its bodily rings and
then propagating it in forw ard direction [74]. This locom otion
mechanism has inspired researchers to design inch-w arm robots,
w hich do not require electrical motors and w heels to move
forw ard. Apart from being cheaper and light in w eight absence
of electrical machines makes inch-w orm robots safer and suitable
for w orking of in explosion risk areas [128]. These robots have
flexible structure w hich is divided mainly in three parts backw ard
clamp, middle part and forw ard clamp. First backw ard part expand
radially to generate support from w all w hile middle part get
elongated longitudinally to move forw ard and then backw ard
clap get contracted and forw ard clamp get expanded radially to
generate support [77,130]. How ever, many of these inch-w arm
robots, such as [92,128], w ere not effective because of their low
speed and poor reliability. Screw -type robots can freely move
backw ard and forw ard, inside the pipe of smaller diameter, based
on the principle of screw . [131] propose a robot w hich has tw o
units, separated by a micro electromagnetic motor, and one unit
is fixed to rotating shaft and another to casing. This arrangement
forces shaft and casing to rotate in opposite direction so that
robot m oves forw ard. This kind of robot is mainly used for precise
motion inside smaller diameter pipes w ith very low payload and
unsuitable for complex pipe structures. Flexibility of the robot
body is a key factor for its ability to pass through complex shape
of pipe. Many robots using flexibility of snake kind body has
been proposed in literature [111,112]. The snake type robots are
designed by serial interconnection of many identical modules on
both sides w ith rotational joints. Robot can move horizontally
much like normal w heel based motion. Whereas, vertical m otion
is achieved by using flexibility of pushing interconnected joints
against the tw o opposing side of the pipe at the same time, w hile
moving upw ard by forw ard motion of w heels [112]. Since, snakelike robots use a number of active joints or m odules; thus, their
development cost is expensive and they need more energy for
operation [81].
3.1.4. Detection technology
M ost of the research w orks published in literature are
concerned w ith navigation of robot inside the pipe. The reason
is that due to availability of various ranges of pipes and their
complexity navigation becomes challenging task. These in-pipe
robots are equipped w ith various sensory devices each related
to specific kind of job. Visual cameras such as CCD [71,78,83,
86,90,91,93,103,105,107109,125,127,129], micro CCD [76,127],
CM OS [110] and micro CMOS [133] cam eras are used for the
purpose of navigation and inspection. CM OS cameras require
lesser components than CCD, w hich makes them comparatively
more energy efficient and cheaper than CCD. Since, CCD cameras
generally tend to have higher resolution than CM OS, therefore
choice of camera requires trade-off betw een cost and clarity of
image. M ost of the time these cameras are mounted on the
front head of the robot from w here these devices continuously
send to the operator for clear navigation. Sometimes IPIR rotates
around their longitudinal axis during passing through complex
geometries, such as elbow and branch, but for clear navigation
camera has to keep its medial axis aligned w ith robots axis. Not
only that sometimes cameras has to deliberately rotate 360 around
its ow n axis for the purpose of overall inspection of circular pipes.
In [86] researchers have proposed a CCD assembly, w hich includes
a CCD w heel rotating circumferentially and eight couples of sub
w heels w hich can rotate circumferentially along w ith around their
ow n axis. Most of the commercially available CCD cam eras are

A. Shukla, H. Karki / Robotics and Autonomous Systems 75 (2016) 490507

very big and cannot be directly installed on the micro-robots.

Therefore, [76,127] have developed their ow n specific micro CCD
cameras suitable for inspection of small diameter pipes. Micro CCD
camera developed in [76] carries 41K color pixels and w ith this
resolution it can locate micro cracks of 25 in the internal surface
of the pipe.
One of the major problems for the pipes in oil and gas industry
is corrosion due to unw anted electro-chemical reactions. Role of
NDT becomes critical to detect early stages of corrosion so that corrective measures can be taken before severe damage takes place.
Early stages of corrosion is not directly visible or detectable by
ordinary surface measurement techniques. How ever, early detection of corrosion is possible by estimation of shirking pipe w all
thickness and structural discontinuities. At present there are several advance techniques available for corrosion detection such as
X-ray [45,134136], ultrasonic [44,100,106,137143], Eddy current [120,122,125,127,144146], and magnetic flux leakage (MFL)
[137,147152]. In [45] have presented a novel vision system w here
images of gas pipes are obtained on radiographic films and then
various image processing and computer vision algorithms techniques are applied to identify w elding defects. X-ray based detection techniques enjoy edge over other NDT of its being capable of
pictorial representation. There are many other advantages of X-ray
NDT such as ability to identify material, estimating material density, and capable of inspecting almost all kind of material, but suffers from requirement of skilled analyzer, expensive, inability to
detect closed cracks and above of all safety concerns related to its
operation [153].
When magnetic field is applied to the internal w all of the pipe
uneven surface of corrosion affected w all distorts the resulting
magnetic field. These distortions in m agnetic field are measured
by magnetic sensors but overall measurement result is only
qualitative in nature. This magnetic flux leakage technique (MFL)
is only suitable for small pipes, and sometimes w hen pipe material
have impurities results of MFL are spurious in nature [143].
There are tw o kinds of M LF technique firstly circumferential
and secondly axial. Circumferential M FL is more popular and
effective than axial one [150]. M LF technique is sensitive to pipe
material under inspection w hereas ultrasonic in free from this flaw
and more accurate in predictions [149]. In ultrasonic detection
method, high frequency acoustic w aves are transmitted from
transducers w hich are reflected by internal surface of the pipes.
Some parts of the propagated ultrasonic w aves penetrate through
w all thickness of pipes and get reflected at outer surface of the
pipe. These reflected and refractedreflected w aves are received
back by the transducers. By using know n distance betw een
transducers and internal w all of the pipe and time delay in
receiving first reflected w ave, corrosion depth on the internal w all
of the pipe is calculated. Similar calculation is also done betw een
internally and externally reflected (first refracted through internal
surface and then reflected from outer surface) ultrasonic w aves
to calculate pipe thickness, w hich allow s estim ation of external
corrosion [100]. Ultrasonic technique (UT) has many advantages
such as high penetration depth, high accuracy, high sensitivity,
rapid testing, portable, safe and finally it can test all kinds of
materials and their properties [153]. But this technique suffers
from expensive training for expert operator, necessity of contact of
the transducers w ith the surface of material and inability to detect
crack along the line of w ave travel [153].
Eddy current technique is also one of the most popular NDT
methods used for crack detection, material thickness estimation and conductivity measurements [154]. When magnetic field,
generated through usage of electrical coil, is applied on nearby
surface circulating eddy is generated in that surface. This eddy
current in turn affect the loading of the coil and its inductance [155]. If constant current is applied across the coil, change

495

in overall impendence can be detected by continuously measuring

the voltage across the coil. Quantitative variation of impendence
leads to detection of crack in the pipe and estimation of other parameters mentioned above [156]. Remote field eddy current (RFEC)
NDT is another variation of above m entioned conventional eddy
current technique [71,157,158]. In this technique low frequency AC
current is applied in excitation coil and an array of circumferential
detector, w hich is placed coaxial to excitation coil but in axially
displaced manner, is used as a pickup coil [159]. Phase change in
the flux received at pickup coil has linear relation w ith pipe thickness [160,161]. According to [159] RFEC is the only detection technique w hich can be used for defect detection of many concentric
tubes [162]. Conventional single coil based eddy current testing
takes very long tim e especially for long and complex pipe structure, therefore [163] proposes eddy current array probe for faster
and efficient detection of defects in oil pipes of aircraft. M ost of
the corrosion detection techniques fails at very high temperature,
such as 175 inside deeper core of the earth, and [164] proposed
a new dc electrom agnetic induction using eddy current testing
method. Eddy current testing techniques is sensitive to large number of parameters related to m agnetic conductivity, permeability
and geometry. It has many other advantages such as w ider temperature range for operation, smaller sizes for probes to be utilized
for smaller diameter pipes, light w eight and portable to be installed
on micro robots, and relatively low er in cost, but suffers from liftoff
effect and its inability to be used for non-m etal pipes [153]. Ultrasonic NDT is more sensitive and has better spatial resolution
than eddy current based techniques, w hereas eddy current techniques are more suitable for inspection of multilayered structures
[161,165].
The MFL method has superior capability of detecting even
smaller metal loss geometries in pipe w alls because generated
magnetic pattern and its distortion has larger dimension in
comparison to anomaly w hich caused it. Whereas the UT m ethod is
excellent in giving precise measurement of the w all thickness and
detecting large corroded w all points. UT also occasionally suffers
from blind spots of echo loss but then it can be compensated
by measurement data provided by MFL sensors w hich do
not suffer from this problem. Essentially these tw o methods
strongly complement each others by compensating draw backs
and enhancing detection capabilities. Extended inspection range,
w here only one method is able to detect fault, and better
accuracy of fault size and its nature, w here both the methods
are able to detect fault, are the tw o main benefits of combining
M FL and UT. Therefore, recently in [166168] researchers have
proposed novel detection mechanism w here M FL and UT both
are combined together so that best of both the techniques such
higher probability of corrosion detection and precise measurement
of metal loss can be achieved in one inspection tool itself as show n
in Fig. 4. In [169] researchers have investigated the improvement
in detection quality by combining the of eddy-current and UT
inspection techniques. In [170] researchers have proposed a neural
netw ork based geometric transformation algorithms w hich can
be used to fused data obtained from multiple NDT sensors. This
techniques uses radial basis function artificial neural netw ork and
allow s user to define complimentary and redundant information
present in inspection data.
Apart from defect inspection, sensors in-pipe robots are
equipped w ith many other kind of sensors for different purposes,
such as gravitational sensor [91], temperature sensor, humidity
sensors and tactile sensors [104,115,171] to successfully navigate
through geometries like elbow and branch. Tactile sensors detect
these geometries of the pipes by the virtue of estimating the force
upon contact w ith hard surface [104]. Researchers in [172] have
used laser projection technique to identify the special geometries
of the pipe w hich helps them in localization of the autonomous
robot, and in [173] laser sensors are utilized for the pose estimation
of robot inside the pipe.

496

A. Shukla, H. Karki / Robotics and Autonomous Systems 75 (2016) 490507

Fi g. 4. UT and MFL are combined in one NDT unit on a IPIR [166].

3.1.5. Control mechanism

Most of the IPIR are controlled and pow ered by tethered cable
or umbilical cord connected betw een operator system and in-pipe
robots [73]. Tethered cables are com posed of pow er cable and
optical fiber w ires for the purpose of video communication and
data transmission [90]. These tethered cable serves many purposes
such as [70]:

Supplying pow er to robot mechanism s and installed devices

such as locomotive motors, video cameras, lighting, defect
detection and other kinds of sensors, in the form of AC/DC,
pneumatic or hydraulic pow er.
Control signals to navigation (w heels and legs), manipulation
(manipulators installed on robots for the purpose of w elding,
cutting and drilling) and inspection mechanism (defect detection sensors and pan, tilt and zoom of camera) of robots.
Used as a safety rope to pull out robot from the pipe
or inspection site in situation of accidental pow er loss or
completion of inspection task [174].
The biggest draw back of using tethered cable w ith IPIR lies in
continuously increasing friction force w ith increasingly traveled
distance by robot inside the pipe [106,108,174]. In such situations
slip-less travel of in-pipe robots require more and more traction
force [73,108] w ith higher and higher dem and of kinetic energy. All
round 360 inspection in circular pipes lead to tw isting of tether
cable w hich may at the end damage the cable itself [175]. Along
w ith all these limitations, total distance traveled by in-pipe robots
are physically restricted w ith usage of tether cable [106,121,174].
Freedom from inconvenient tether cable is the biggest impetus for using w ireless technology for inspection robots. There has
been various un-tethered and w ireless inspection robots proposed
in literature to overcom e problems related to conventionally tether
cable controlled inspection robots [72,74,78,89,98,99,106,125,127,
176178]. Tether less robots can be picked up from their exit point
itself in the pipe rather than coming back to the base point for
their collection. All the w ireless inspection robots carry, pack of
rechargeable batteries for on-board pow er supply and communication system w ith higher bandw idth for data transmission [174].
For micro robots, w here space is already a constrained and usage of heavy inspection devices such as camera [175] and thick
hard w ire w ill further lim it its mobility inside the pipe [74,78].
Therefore, [125,127,175] have proposed micro robots w ith compact and w ireless energy supply, communication, and navigation
mechanism. The biggest draw back of w ireless technology comes
from their inability to penetrate m etallic pipes due to high electrical conductivity of materials such as steel [179]. To make w ireless
technology even to w ork for metallic pipelines, problem of electromagnetic shielding has to be solved first. And for this purpose [98]
has proposed a emitter w ith extrem ely low frequency of Electromagnetic w ave, w hich can pass through metallic w all as w ell.

Overall control architecture of the inspection robots highly depend on their degrees of autonomy starting from manually operated, semi-autonom ous and fully-autonomous [47,180]. Visual
feedback from the remote location to the operator and in return
command signal from the operator to the machine completes the
loop in manual remote control process [181]. Sensors and manipulators installed on remote device (or site) w orks as eyes, ears and
hands of operator at operation site. This kind of automation requires w ell trained person, w ho has to decide every single step
of process such as safe navigation of the remote device, sw itch-on
and off of the inspection sensors, and manipulation operation etc.
[33,181]. Mostly all the commercially available inspection robots
belong to this category [176]. In semi-autonomous robots, in principal operator has to define only initial and final conditions of the
operations. Sensory data received by robot is interpreted either locally by processor installed on the robot or by computer at operator side [176]. Overall in this process, role of the operator is greatly
reduced by inclusion of intelligent navigation and collision avoidance techniques [174]. In fully automated operations, robotic devices can be preprogrammed for complete operation and there is
no need of human interventions at all especially during the operation [177,181]. Since, oil and gas industry is extremely sensitive
to the environment and requires continuous human supervision,
therefore it uses semi-autonomous human controlled robots such
as [182185].
Fully autonomous IPIR do not require any kind of intervention
from remote operator during the operations. KANTARO [186] and
M AKRO [177] present tw o fully-autonomous, module based, untethered and self-steering IPIR, but these IPIRs are only suitable
for navigation in straight pipelines or at maximum in curved
pipe. These IPIR suffer through w heel slip error and are unable
to climb in vertical sections of pipelines. In [119] researchers
have presented a fully autonomous IPIR, named FAM PER, w ith
excellent mobility it can navigate through any spatial conditions
of the pipeline (horizontal and vertical both) having complex
geometries (T-branches, Y-branches, and L-joints) using its craw ler
and extendable link system. The biggest push for development
of completely independent IPIR stems from the fact that general
w ireless communication is not possible for most of the metallic
pipes and w ired communication is severely lim iting the mobility
and navigation of robot inside the pipe. With these reasons in
mind [172] has proposed autonomous navigation inside pipelines
w ith the help of laser based landmark detection technique. For
autonomous inspection robots, all the inspection data has to be
stored locally w hich can be analyzed later on after exit of robot
from the pipeline. Apart from local data storage, autonomous
robots must possess some local intelligence as w ell for the purpose
of navigation [95] and defect detection [179]. For example, [106]
contains craw ler unit, drive unit, central controller unit, battery
unit and ultrasonic inspect unit. In this IPIR, drive control system
receives motion com mand from central controller via CAN bus.

A. Shukla, H. Karki / Robotics and Autonomous Systems 75 (2016) 490507

497

Fi g. 6. Nept une in-t ank craw ler [209].

Fi g. 5. Met allic oil t ank climbing t he inspection robot [191].

And, all these units are installed on robot itself and there is no need
of external operator.
Navigation control, specifically steering through elbow and
branches, is one of the most important control aspect of all the
IPIR described in literature. For example [86] has proposed a
sophisticated control algorithm for steering of IPIR from various
complex geometries of pipes based on modulation of w heel speeds.
In [114], researchers have developed a kinematic m otion control
strategy for passive-joint active-w heel snake robot, to coordinate
betw een different modules for higher mobility and low er demand
of traction force. In [187], researchers have proposed a new IPIR,
called National Taiw an University (NTU) Navigator, w hich have a
stable and smooth fuzzy steering controller for operating inside the
pipes of varying diameters. Most of the IPIRresearch focuses on the
drive control mechanism but ignore global planning for automatic
yet comprehensive inspection of complex pipelines. Therefore,
in [181] researchers have presented an efficient 3-D guarding
algorithm that can cover a given complicated environment using
as few as possible points.
3.2. Tank inspection
Huge metallic tanks are used for storing the crude oil and gas
both at offshore and onshore production plants. These tanks have
many w elded seams along w elded plates and these seams are
prone to leakage due to corrosion and w ear. Continuous storage of
crude oil and other such products inside the metallic tanks generates many corrosive by-products such as iron sulfide and hydrogen sulfide. Bubbling H 2 S does more damage to the roof than
the bottom of the tank. Bottom of the tank, mostly used in offshore facilities, is mainly damaged by collection of large quantities of sludge material containing heterotrophic microorganisms.
Although bottom also contains many corrosion pits created due to
internal reactions of these products [188]. Since, humanly inspection requires completely emptying the tank and stopping all the
production for few w eeks [189], hence, this process is lengthy, expensive and hazardous from safety point of view [190]. Autom ated
inspection, w hile tanks are full, w ith continued operation of the
plant is the motivation for the research for mobile robots for intank inspection.
M ain criteria of categorizing the tank inspection robots is based
on the principle of clim bing the tank under inspection [192]. There
are tw o broad categories of climbing techniques, firstly based on
adhesion m echanism and secondly on the locomotion principle.
The most common adhesion mechanisms are magnetism [71,
146,191,193202], vacuum suction [203205], specific attachment

devices such as rails [206] or legs and grippers/clamps [207].

The locomotion category can be further divided in four subgroups [199], such as w heels [71,146,193196], tracks [204],
legs [197,208] and actuator or arm s [199] based devices. Fig. 5
show s general functioning of permanent magnet adhesion based
locomotion principle during out tank inspection of above ground
tank.
There are some outstanding tank inspection robots such as
Neptune, Maverick and Scavenger. Neptune, as show n in Fig. 6,
is a mobile robot designed by Robotics Institute at Carnegie
M ellon University for in-tank inspection [209]. This robot aims
for remote inspection of above-ground tanks, w hile they are
filled w ith petroleum products. This robot is made for detecting
corrosion sites at floor and side w alls of the tank w ithout emptying
it. Neptune is equipped w ith various components designed for
specific purposes such as vehicle craw ler to move up and inside
the tank, ultrasonic sensors for detecting leakage, HD cameras
to take the visual feedback and acoustic sensors for position
locationing inside the tank etc. Neptune provides the visual records
of each w elded seam and thickness-contour maps for the floor
of the tank using ultrasonic thickness measurement sensor [209].
M ost of the robots mentioned above carries almost similar kinds
of equipments for the purpose of inspection but their main
differences come from climbing and control mechanism s. Osaka
gas has also proposed a robot w ith a novel detection technique
w hich w orks on a principle that ultrasonic w ave diffracts only
w hen there is flaw . This technique is much faster and cheaper than
conventional ultrasonic pulse reflection methodology [194].
There are various w ays in w hich control system for navigation
of the inspection robots can be designed. [210] has presented a
fuzzy CM AC algorithm along w ith neural netw orks to establish
the tracking control system for improving the performance of the
robot navigation. In [211] a novel client/server architecture has
been proposed for autonomous operation of the inspection robot.
Here client program is run on inspection root locally w hich is
related to safe climbing and navigation of robot, w hereas server
program is run on operator side at control room. At server side,
programs are concerned w ith capturing the visual data, detection
of leakage and some manipulation w ork. Fuel tanks installed on
the ship have thin metallic w alls so normally used heavy climbing
robot cannot be used due to tendency of deformation of surface. A
novel architecture called mother/child has been proposed in [212],
to deal w ith such a situation. Where mother robot is a normal
heavy climbing robot w ith high mobility used for climbing through
strong track available on the tank, and child robot, w hich is very
light in w eight, carries only required detectors for inspection.
All the locom otive and adhesive devices are installed on mother
robot only, and child robot is only a extended inspecting device in

498

A. Shukla, H. Karki / Robotics and Autonomous Systems 75 (2016) 490507

this system. Most of these robots are sem i-autonomous, w here a

operator at remote location supervising the w hole mission. In [191,
213] a novel control mechanism has been presented for inspection
robot w hich can be operated in fully autonomous mode w ith
w ireless control. After analyzing all these automation model it can
be safely said that somehow or other most of these automation are
broadly based on the principle of teleoperation.
3.3. Automated gas sampling
To determine composition and quality of hydrocarbon, w hich
is essential to know for chemical processing and price of fuel in
international market, there is a routine process of collecting gas
from production site. Normally this job is done manually w here
an operator collects desired amount of the gas in a container by
connecting it to supply valve. This process involves certain set
of protocol to collect fully representative sample but it has been
found out that many of the times this lead to sample variations
depending on individual operators responses. Though sample
variation may be cause of error but HSE is the far greater concern
in this w hole operation. Therefore, Norw egian company Statoil has
developed automated gas sampling station w ith usage of standard
6-DOF serial manipulator for this repetitive process [50].
3.4. Aviation robotics for inspection
Normally industry standard intelligent pigs are used for
inspection of corrosion and other such integrity issues in
transportation pipes but such internal inspection alw ays interferes
w ith overall routine operation of the plant hence external
inspection of pipes are preferable. At present only external
inspection for transportation pipes is through a process of manual
inspection, w here a group of w orkers, driving a vehicle along the
transportation pipes, perform visual inspection of the pipes for
detection of leakage or any other kind of damage. But such external
manual inspection processes, currently em ployed in industry, are
highly inefficient, expensive and hazardous as w ell. Therefore,
usage of unm anned aerial vehicle (UAV) w ith appropriate sensors
has been considered as a suitable option for external automated
inspection of the pipelines [214]. The first UAV w as tested
in 1950 by Ryan Aeronautical for military reconnaissance and
since then this technology has come long w ay in service of
humanity [215]. UAVs are remotely controlled devices operating in
semiautonomous or autonomous architecture via command center
at ground. A human pilot can be engaged in flying for not more than
five hour at a stretch w hereas an UAV can remain in air up to 30 h
making it suitable to be used in long inspection operations [216].
Absence of human pilot from flight in UAVs, not only makes it
safer but also reduces cost of dangerous operations w ith higher
performance efficiency in m ost cases [217].
As a matter of record now adays UAVs are extensively used for
surveillance and reconnaissance in military missions [218,219],
rescue missions [220], remote sensing and exploration [221], domestic policing [222], disaster relief [223], scientific research [224],
archeology [225], film or photo shooting [226], forest fire detection and fire-fighting [227,228], pesticide spraying and geophysical survey [229231], logistics and payload transport, and ad-hoc
communication gatew ay [232,233], to name just few [217]. Such
inspection mechanism are not only cheaper but also more efficient
and robust, supplying inspection data round the clock w ithout any
interruption due to fatigue [234]. In regard to oil and gas industry, apart from inspection of the pipelines, such UAVs equipped
w ith relevant sensors and data transmission system, can be w idely
used for surveillance of production plants, flare stacks, refineries
and transportation system of petroleum products. Exploration of
oil and gas in inhospitable and difficult remote areas is another

Fi g. 7. Quadrot or UAV used for surveillance of pipelines [238].

challenging task w here aviation robotics can be of great use [235].

In [236] researchers have argued for the usage of UAVs in arctic
zone for deploying tools for ice advisory system. First fully functional UAV to inspect plant assets, such as flare stacks, w ere at the
UK onshore oil refineries in year 2010. With this technique, for
the first time operators got the chance to understand condition of
their equipments before any kind of shutdow n, and w ithout exposing w ork personals to risky environment. This system has enabled
engineers to continuously monitor critical components located at
difficult height in plant structures, such as chimney stacks, ducting, pipe racks, and vents, w hile allow ing them to prioritize m aintenance, relocate budgets, defer shutdow n, reduce the time for
turnaround and order replacement parts before turnaround [237].
This success has motivated many operators across the w orld to
use UAV technology for cost effective and time efficient inspection
than expensive and dangerous conventional techniques. For an example in year 2012 British Petroleum (BP) has setup a research
team to develop technology suitable for usage of UAVs for inspection of their oil pipelines in Prudhoe, Alaska, USA [238] as show n
in Fig. 7. In last four years this technology has become standard
operational practice for being used extensively in routine remote
surveillance over both onshore and offshore platforms [237]. Not
only in case of surveillance but also in the cases of oil spill [239] and
other such disasters w here direct human intervention is not only
hazardous but also inefficient, UAV technologies can be of great
help [233,240]. Belgium has deployed UAVs of B-Hunter class to
monitor marine oil spill in its portion of north sea [241]. This is a
big UAV flying at high attitude and carrying various kind of sensors
including visual and IRcameras for monitoring [242]. When gusher
happens in oil field, spread of poisonous H 2 S gas poses acute danger to human life along w ith loss of revenue. Since, know ledge of
speed of spread and density of gases are tw o immediately required
parameters to control situation from further deterioration, therefore, in such critical situations usage of UAVs equipped w ith suitable sensors to collect data is one of the safest and reliable option
now adays [243].
Like other robotic technologies in oil and gas industry, UAVs
also w ork on the principle of teleoperation, w here flying m achine
as a remote system remains in autopilot mode, w ith programmed
navigation system equipped w ith GPS and inertial sensors, w hile
transmitting and receiving information signals from ground level
operator [46]. Whole UAV system is composed of four major
components w hich is described as follow s [240,244]:
3.4.1. Hardware platform
The first and foremost important part of a UAV is its body
frame along w ith w ell defined geometrical and dynamics related
parameters. There is no single UAV fulfilling all the needs to
accomplish all the desired missions, therefore, there is a long
range of purpose-specific UAVs according to requirement of
particular mission. Different parameters are used for classification

A. Shukla, H. Karki / Robotics and Autonomous Systems 75 (2016) 490507

of UAVs such as desired tasks, w eight, range, speed, control

mechanism and physical structure etc. and depending on these
parameters, overall navigation control technology and physical
structure changes for different UAVs [245]. There are total six
categories of UAVs depending on their functional aspects such as
(1) Target and decoyproviding ground and aerial gunnery a target
that simulates an enemy aircraft or missile, (2) Reconnaissance
providing battlefield intelligence, (3) Com batproviding attack
capability for high-risk missions (see Unm anned combat air
vehicle) (4) LogisticsUAVs specifically designed for cargo and
logistics operation (5) Research and developmentused to further
develop UAV technologies to be integrated into field deployed
UAV aircraft, (6) Civil and Commercial UAVsUAVs specifically
designed for civil and commercial applications. Weight (W ) and
maximum height of operation or flight altitude (FA) are tw o
another w idely used criteria for categorization of UAVs and
there are total seven categoriesof UAVs based on these tw o
parameters [246], (1) Nano-UAVs(W < 0.1 kg and FA < 100 ft),
(2) Micro-UAVs(0.1 < W < 1.0 kg and FA < 500 ft), (3) M iniUAVs(1 < W < 20 kg and FA < 5000 ft), (4) Sm all-UAVs
(20 < W < 120 kg and FA < 20 000 ft), (5) Tactical-UAVs(120 <
W < 1000 kg and FA < 30 000 ft), (6) M ALE-UAVs(1000 < W <
6000 kg and FA < 45 000 ft) and (7) HALE-UAVs(W > 6000 kg
and FA < 65 000 ft) [245,246].
There are another tw o broad categories for UAVs depending on the thrust generation mechanism [247], first fixedw ing UAVs [221,225,228,243,248254] and second rotary-w ing
UAVs [255266] and both of these UAVs have their ow n benefits
and limitations, w hich have to be looked into detail before used for
any particular mission. Fixed-w ing UAVs are simple architecture
w ith efficient aerodynam ics, suitable for higher speed and longer
duration of flight covering relatively w ide territory for mission.
Biggest limitation of such UAVs is that it require a launcher [249]
or longer runw ay for their take-off or landing. Whereas, rotaryw ing UAVs have complicated aerodynamics making it suitable for
shorter duration of flight w ith low er speed. But the benefit, of using such UAVs, is their capability of vertical take-off and landing (VTOL) w hile having additional capabilities of hovering over
target objects and able to perform agile maneuvering [267]. Such
UAVs are suitable for inspection of complicated oil and gas facilities
both at onshore and offshore, w here longer duration is required
for making optim al inspection and very little space is available for
take-off and landing [268,269].
3.4.2. Ground control
Every UAV requires a ground control station (GCS) and the
important responsibilities of this system is to supervise flight route
planning and uploading, flight posture parameters dow nloading,
sw itching betw een remote control and autonomous navigation,
real time monitoring and control, etc. GCS require a pilot w ho
does maneuvering of the UAV by tele-command and a payload
operator w ho gather surveillance data to forw ard it to other end
users of UAV [270]. Communication betw een UAV and GCS is
bi-directional as signals received from UAV to GCS are called as
dow nload link and this data is processed to generate command
signal back to the UAV, w hich is called as upload link, completing
control loop [271]. GCS has three parts, first hardw are such as
computers and communication systems, second GCS softw are
running on the hardw are and third human operatorcognitive
authority in the overall system. For bigger UAVs, hardw are includes
pow erful computers and communication antennas installed at
fixed ground station, w hich are generally expensive. Whereas for
sm all UAVs, tablets or similar mobile devices can be used as both
processing-cum-display and communication unit carried in hand
by human operator w hich is quite cheaper in comparison to fixed
station [270]. Softw are running on GCS computers are integral to

499

w hole system and are of tw o kinds, real-time and non-real time.

Real time softw are is used to send and receive real time signals for
operation of UAV w hile non-real time softw are is used for postprocessing of the data already received by UAV [272]. In [273,274]
researchers have used reconfigurable data interchange format such
as LINUX to enable GCS real-tim e softw are to perform multi-type
large volume bi-directional communication, w hich for smooth
functioning require high transferring speed of the data and high
reliability for safe operation of the UAV [271].
3.4.3. Flight control
The flight control system (FCS) is a seam less integration of
softw are and hardw are as one system required for navigation
and control of the UAVs such as inertial navigation system (INS),
global positioning system (GPS), air-data dead reckoning system,
radio navigation system and doppler heading reference system
[275,276]. In [277] presents technique to combine GPS and vision
data for precise target detection, tracking and navigation by UAV.
In [278] researchers have implemented a code delay algorithm
based on GPS delay mapping scheme, w here only direct and reflected GPS signals are processed to generate localization points
and height measurement for the UAV. Usage of only INS provides
position, speed and attitude of the UAV but suffers through continuously increasing bias error therefore GPS is used as an additional
sensor to filter out bias error and improving overall estim ation.
Data generated from the GPS [218,245,279] and INS sensors are
together used to track the accurate location of the UAV and this
online data is fed to FCS for navigation guidance. The major functionalities of FCS are: (a) Determination of attitude angles, coordinates and motion parameters (angular rates, accelerations); (b)
UAV guidance and flight control w hen perform ing programmed
automatic flight; (c) Pitch and roll angles stabilizing; (d) Navigation and attitude telemetry data output; (e) Programmed payload
control.
M ost commonly Kalman filter is used to fuse these navigation
data [275,280282] but the UAVs have extremely non-linear dynamics, therefore, extended Kalman filter (EKF) [283] has been
adopted to linearize model around each current state [284286].
Assuring stability of the UAV is one of the m ost critical task in
the UAV research, therefore, in [254] researchers have proposed
a PID controller in conjunction w ith EKF to calculate attitude ensuring stability of the flight even w ith uncertainty and perturbed
conditions. In [243] researchers have used adaptive Kalman filter
w ith radial basis function neural netw ork to estimate attitude angle. But in general nonlinearity of UAVs are so high that estimations results even by using EKFs are not accurate enough, therefore
in [250] a novel unscented Kalman filters (UKF) has been used
for determination of attitude angles. When UAVs are operated at
high speed, effect of non-linearity becomes extremely significant
then nonlinear controllers are implementer. Feedback linearization is the most common technique to design controller for nonlinear systems but unfortunately UAVs are non-minimum phase
systems and direct feedback linearization can lead to instability of
system. Therefore, in [287] researchers have proposed recursive
nonlinear design technique called backstepping algorithm w hich
is inherently stable due to usage of Lyapunovs stability method.
To im prove flight performance tackling of uncertainties and external disturbances are compulsory, therefore, in [288] a combination of stable backstepping and neural netw orks based adaptive
controller has been proposed to compensate for above mentioned
uncertainties and disturbances in overall aerodynamic modeling of
system. In [289] researchers have draw n comparison betw een four
controllers w ith different combination of backstepping and sliding
mode controller. In [290] researchers have used adaptive backstepping technique [291] to design tw o level advance controller, w here,

500

A. Shukla, H. Karki / Robotics and Autonomous Systems 75 (2016) 490507

internal attitude control loop tracks attitude reference generated

by outer position control loop.
Most of the above mentioned controllers are used for sophisticated UAVs of bigger sizes w hich are not applicable for very small
size cheaper UAVs therefore in [292,293] researchers have come
up w ith M EMS based low -cost, light w eight, flight control system
specifically designed for small UAVs w ith suitable sensors, controllers and w ireless transmission system. With decreasing size of
UAVs, they can be used for flying in complex space constraint areas as w ell such as inside plant buildings. But flying in such areas is
challenging form navigation point of view as quality of GPSsignals
is degraded due to multi-path effects and frequent dropouts for
prolonged period of time [294]. Therefore, [295] presents visionaided navigation, w here onboard vision camera is used to detect,
track and geo-locate 3-D landmark w hile GPS is available and,
w hen GPS is denied temporarily, system combines this estimated
3-D map w ith new ly observed landmarks to both navigate and localize UAV in a process called boot-strapping. Apart from navigation, autonomous landing of UAV is a task of critical importance
for safety of routine UAV operations. In [296] researchers have
fused data from three sensors GPS, compass and vision camera get
safe autonomous landing w ith maximum deviation of 0.5 m. Without GPS data, especially for small UAVs, navigation approaches
based on GPS and inertial sensors lead to unacceptable high drifts,
therefore, [297] proposes new attitude control based on monocular
vision for autonomous landing of small UAVs in GPS-denied scenarios. The biggest problem of using visual data based navigation
is that generally it requires higher computation pow er w hich is difficult for small UAVs to possess [276].
In some cases w here territory is too w ide to cover in short
time then rather than single UAV, strategy of using multiple UAVs
in a cooperative w ay is more effective [228,230,233,253,298,299].
When multiple UAVs are engaged for a common mission, cooperation betw een different UAVs can be achieved successfully only
by flaw less coordination betw een UAVs and their communication betw een them selves and command center [300]. In [301] improved version of particle sw arm optimization algorithm has been
used for optimal task allocation among the different UAVs by considering task precedence, time constraints, trajectory limitations
and capability of UAV etc. In [302] researchers have proposed an
improved cooperative control technique for task allocation even
in the situation of imperfect or delayed communication betw een
UAVs. In [217] researchers have presented a novel haptic teleoperation based control strategy for group of UAVs having three layers
of control and overall masterslave configuration.
3.4.4. Sensors
Sensors are essential part of any autonomous inspection
system . There are at least four kinds of sensors installed on a
UAV, navigation sensors for position detection (accelerom eters or
inclinometers, global positioning system (GPS), inertial navigation
sensors (INS), visual cameras and gyroscopes) [303,304], the
remote sensors (com munication channels), optical sensors (to
get visual images) [214] and detection sensors (detection of
defects such gas leakage from pipes [305] or normal operational
parameters such as temperature). Generally INS is combination of
three orthogonal accelerom eter and three orthogonal gyroscopes
to track position, orientation and velocity of moving body
w ith respect to a know n reference point. Usually INS suffer
from integration drift w hich is compensated by integrating GPS
data. In [306308] researchers have presented methodologies of
combining GPS/INS data to im prove precision of navigation.
A major challenge in commercializing UAVs is to develop
cheaper and reliable w ireless comm unication w ith other UAVs
and ground control center. For UAV to ground communication
involves line-of-sight long rage communication such as radio

modems, satcom, microw ave links or combination of these at

different levels [309311]. For example in [312314] radio signal
based different w ireless netw ork standards, such as 802.11b,
802.15.4 and 802.11a respectively, are used from establishing
communication betw een UAV to ground control operator. In
the detection category, small and light w eight sensors w hich
can be used for non-contact data collection are installed on
the UAVs for aerial inspection w ithout endangering human
lives. In [315] the Federal Institute for M aterials Research and
Testing (BAM ), in cooperation w ith the AirRobot GmbH & Co.
KG company has developed a mini UAV equipped w ith modular
gas detection sensors w hich can act as flying mobile device to
be used for estimating risk potential in gas accidents scenarios,
emission control, detecting gas leakage in transportation pipes
and monitoring critical areas. UAVs equipped w ith infrared and
optical sensors can be used to detect and monitor humans trapped
in hazardous areas under critical rescue m issions as proposed
in [316]. UAVs equipped w ith thermal, visual, multispectral
imaging and oil detection sensors are used for remote sensing
operations [317] in the oil and gas industry suchas oil spill
detection [239,318323]. To increase precision of sensor data,
unlike maned aircraft, UAVs can fly very close to ground facilities
even in night time, w hen environment is calm in absence of usual
disturbances, w ithout risking the life of pilot [216,234].
3.5. Wireless sensor network
Pipelines are most inexpensive w ay to transport all kinds of
fluids especially oil and gas, and these pipelines are need to
be continuously monitored for leaking and damage by regular
patrolling group of technicians. As it has been already mentioned
that such process of inspection is not only inefficient and expensive
but also hazardous therefore as an alternative, sensors are installed
at regular interval along the pipe netw ork itself. Now these
installed sensors can be w ired for their pow er supply and data
transmission but this technique suffers from unreliability in many
w ays especially in the case of any damage to w ire, w hich makes
w hole inspection netw ork completely dysfunctional [324,325].
Therefore, in [326] researchers have proposed an integrated
netw ork w here multiple battery pow ered w ireless sensors are
fitted in w ired sensor netw ork to improve overall reliability. In this
case even if w ired sensor netw ork do not function, still inspection
process for detection of leakage and damage remain unperturbed
supported by w ireless sensors. In [325] researchers have presented
review of many important WSN techniques currently used for
inspection of overground, underground and underw ater pipelines.
In [327329] researchers have described in detail all the technical
requirements and possible applications of the WSNs in the oil and
gas facilities. Though W SN techniques are more reliable than w ired
sensors netw ork but it suffers from energy hot spot problem
and short life span for WSN, due to unequal energy dissipation at
different nodes. To overcome this problem [330] have proposed a
novel data collection algorithm depending on the WSN data fusion
strategy w hich eventually improves the netw ork performances
on delay and energy. This strategy not only ensures the effective
propagation of urgent data but also lead to prolong life span of the
overall netw ork. To mitigate the hot spot problem and optimally
utilize energy of the w ireless sensors, in [331] researchers have
proposed a sleepw ake cycle w hich enables each sensor in the
netw ork to w ake up for only 10 s in a 10 min time interval saving
significant energy.
Not only leakage but also sand production in oil w ells is a
major problem for petroleum industry because excessive sand
production can cause severe erosion of transportation pipelines,
valves and separators. The most common solution for sand
production is stopping or reduction of production [332] therefore

A. Shukla, H. Karki / Robotics and Autonomous Systems 75 (2016) 490507

in [333] researchers have come up w ith sand monitoring system

using w ireless sensor netw ork (WSN). Aim of this system is to
develop a mechanism w hich can be used for sand detection and
sand production measurements, thus allow ing operators to place
sand removal machinery in w ell advance before any extreme
situation emerges [333]. Apart from detection of sand, leakage
and damage, WSN techniques can be used in designing anti-theft
mechanism for oil and gas pipes as proposed in [334]. In [335337]
researchers have proposed cost-effective, efficient and reliable
w ays of monitoring the oil and gas pipelines by combining
standard WSN technique w ith unmanned aerial vehicle, in-pipe
inspection legged robot and autonomous underw ater vehicle.
Leakage in pipelines are mainly due to corrosion, therefore rather
than detection leakage as reactive measures, in [338] researchers
have proposed usage of W SN to detect early stages of corrosion
itself as proactive measures to stop final damage.
Generally due to the linear nature of structure to be monitored
(e.g. oil and gas pipelines) sensors are placed linearly along the
structure, therefore such netw orks are also called as linear sensor
netw ork (LSN). There are four kinds of LSNs depending on the node
hierarchy and its increasing capacity of m onitored areas from sm all
to medium to large to very large [339]. In first sm all range LSNs
only Sensor Nodes (SNs) are used for generating the sense data
and also directly communicating it to the Netw ork Control Center
(NCC). Obviously such netw orks cannot be used for long, complex
and uncertain environm ent as sensors generally have limited range
of communication. Sensing and com munication both w ill drain
huge energy reducing overall life cycle of such LSNs. Second kind
of LSNs have both array of SNs and Relay Nodes (RNs). These RNs
dedicatedly used for aggregating sense data sent by SNs and then
relaying it to the base station (BS) [340]. In this case since SNs
are need not to communicate directly to the base station saves lot
of energy and bring sim plicity in the design of SNs. Such W SNs
are suitable for medium range of inspection area. Third kind of
LSNs use three kind of nodes, SNs, RNs and Dissemination Nodes
(DNs). In this case RNs need not to comm unicate directly w ith base
stations rather there w ill be dissemination nodes located nearby
RNs to collect total data over the period of time and sending it to
the base station. Such WSNs are suitable for inspection of large
areas and complex netw orks of pipes. In [341], researchers have
used a UAV as a mobile DN to receive and transmit the inspection
data. In fourth kind of LSNs there are four nodes SNs, RNs, DNs
and Sink Nodes. Overall W SN generally has tw o sink nodes, one at
the starting of the netw ork and one at the end of the netw ork and
a mobile DN like UAV travels back and fourth along the netw ork
to transm it data collected from RNs to sink nodes. Sink nodes are
places w here all the data of a particular netw ork is collected for
finally to be sent to the NCCfor last stage of processing [341]. W hen
pipelines under inspection travel through hundreds of kilometers,
overall netw ork is broken in many small segments of fourth kind
of LSNs for efficient handling and processing of the bulk of the
received data.
These w ireless sensors can be installed on the undergroundpipe surface as w ell, to be called as under-ground w ireless sensor
netw ork (UWSN), for timely detection of any kind of unw anted
damage and leaks by sensing multiple flow parameters such
as pressure and temperature. M any of the sensors used for
measuring temperature and pressure are required to be installed
in invasive w ays [342] therefore, [343] proposes M EMS based
sensors deployed on pipelines non-invasively. Wireless sensors
alw ays have problem of coverage range due to limitations on
energy supply therefore in [330] researchers have studied three
dimensional sensing coverage range of SNs and its relation
w ith pipe radius. There are many acoustic or vibration sensors,
claiming to give accurate measurem ents of leakage by using cross
correlation methods [344], such as used in [344347]. In [346]

501

researchers have used ultrasonic sensor for detecting leakage in

multiphase pipelines (i.e., pipelines w hich carry simultaneously
more than one fluid). Biggest problem w ith such sensors is that
they require very high sampling rate leading to huge energy
consumption at SN level itself w hich is highly undesirable for
energy limited w ireless SNs. Another problem is that such sensors
send huge data for processing at RNs w hich unnecessary burdens
every stage of data transfer and final computation as w ell [348]. To
overcome this, multiple SNs, located on pipe surface, are connected
to a master sensor (w hich is actually RN) by RF comm unication
for transmission of collected sense data and from there these data
sets are uploaded to above ground cloud or base station (BS) [348].
In [336] researchers have proposed a in-pipe inspection mobile
robot carrying SN inside the pipe and there are multiple RNs
located above ground and one BS. Transferring data signals from
moving SN to above ground RNs and limited energy capacity of SNs
are some of the major challenges to deal w ith in this process. A path
loss m odel is designed to characterize com munication betw een
one SN and m ultiple RNs w hile optimizing energy consumption of
SN depending on the pipeline coverage requirement and the SNs
transmission pow er limit [336].
4. Conclusion
This paper is aim ed tow ards fulfilling the dire need of a
brief technical survey summarizing all the current applications of
robotics in the onshore oil and gas industry. There are five main
steps involved in the global business of oil and gas, starting from
exploration, drilling, production, refining and to finally transportation to consumer end. Every step involves som e degree of automation but robotics inspection of metallic pipelines and tanks,
used for transportation and storage of oil and gas, are most critical and popular. Therefore, in the onshore oil and gas facilities
most of the robotic solutions are available in the form of In-Pipe
Inspection Robots and Tank Inspection Robots. This paper has summarized various steering mechanisms, propulsion technologies,
defect-detection techniques and control mechanisms for these inpipe and tank inspection robots. Most of the commercially available robots, used for pipe inspection are still in the form of Pigs due
to long experience of its usage and reliability. Whereas m ost of the
tank inspection robots, are remotely operated vehicles carrying appropriate sensors for detection and communication. M ost of these
robots are still heavily depending on skilled operators for their operations but trends of successful im plementation of reliable teleoperated sem i-autonomous robots are also fast becoming viable
solutions. Inspection of underground metallic pipes by Intelligent
Pig is not only hazardous and inefficient but also interfere w ith
routine production operations of the facilities. Therefore, w ith advancing aviation technology usage of UAVs equipped w ith suitable
sensors for non-contact inspection has become a new trend. This
trend is being inspired by successful usage of UAVs in many other
different fields such as military, geographical survey and oil spill
confrontation etc. Operation of UAVs are also based on the principle of teleoperation and four main components of overall UAV
operation such as, hardw are platform, ground control, flight control and sensors are also summarized here. Continuous inspection
is the key for safety of sensitive oil and gas facilities therefore rather
than usage of mobile inspection spatially fixed sensor netw orks
such as w ired and w ireless are also w idely used in oil and gas industry. Thus This paper summarizes the state of art robotic solutions available for the onshore facilities of oil and gas industry.
Ref er ences
[1] Schlumberger, Schlumberger chairman and ceo andrew gould addressed t he
oil and gas investm ent community at t he 33rd Annual How ard W eil Energy
Conference on April 4, 2005 in New Orleans, Louisiana, USA, 2005.

502

A. Shukla, H. Karki / Robotics and Autonomous Systems 75 (2016) 490507

[2] Y. Bai, Q. Bai, Subsea Engineering Handbook, Gulf Professional Publishing,

2010.
[3] C.A. M iller, Energy resources and policy: Vulnerability of energy resources
and resource availabilityfossil fuels (Oil, Coal, Natural Gas, Oil Shale).
[4] US EPA (2014) inventory of US greenhouse gas em issions and sinks
(19902012).
[5] H. Devold, S. Taylor, Oil and gas-the energy w e love to hate.
[6] G.G.B. David Evans, A. Bloodw orth, J. Cow ley, Onshore oil and gas, 2011.
[7] H. Devold, Oil and Gas Product ion HandbookAn Int roduct ion to Oil and Gas
Production, Transport , Refining and Pet rochemical Indust ry, 2013.
[8] A. Muggeridge, A. Cockin, K. Webb, H. Frampt on, I. Collins, T. Moulds, P. Salino,
Recovery rat es, enhanced oil recovery and t echnological limit s, Phil. Trans. R.
Soc. A 372 (2006) (2014) 20120320.
[9] M . Hook, R. Hirsch, K. Aleklet t , Giant oil field decline rat es and t heir influence
on w orld oil product ion, Energy Policy 37 (6) (2009) 22622272.
[10] M . Bazilian, A. Brandt , L. Billman, G. Heat h, J. Logan, M. Mann, M. Melaina, P.
St at w ick, D. Arent, S. Benson, Ensuring benefits from nort h American shale
gas development : Tow ards a research agenda, J. Unconv. Oil Gas Resour.
[11] G. Erbach, Unconventional gas and oil in nort h america: The im pact of shale
gas and t ight oil on t he US and canadian econom ies and on global energy
flow , June 2014.
[12] D.A. Anisi, E. Persson, C. Heyer, Real-w orld demonst rat ion of sensor-based
robot ic aut om at ion in oil and gas facilities, in: IEEE/RSJ Int ernational
Conference on Intelligent Robot s and Syst ems, IROS, 2011, pp. 235240.
[13] Argonne Nat ional Laborat ory, USA, Tar Sands Basics (2012). URL:
ht t p://ost seis.anl.gov/guide/index.cfm .
[14] Center for Sust ainable Syst ems, Universit y of M ichigan, Unconventional
Fossil Fuels Fact sheet . Pub. No. CSS13-19 (Oct. 2014).
[15] E. M cGow an, L. Song, M . M arshall, K. River, The dilbit disast er: Inside the
biggest oil spill youve never heard of, Part 1, InsideClim at e New s, June 26.
[16] A. Sw ift , E. Shope, S. Casey-Lefkow it z, S. Club, Tar sands pipelines safet y
risks, Nat ural Resources Defense Council W ashingt on, DC, 2011.
[17] D. Sassoon, Crude, dirty and dangerous, August 2012. URL:
ht t p://w w w .nyt imes.com/2012/08/21/opinion/t he-dangers-of-dilut edbit um en-oil.ht ml?_r=0.
[18] H.N. Psaraft is, N.P. Vent ikos, An int elligent robot syst em t o respond t o oil
spills: the EU-MOP project , in: In The Proceedings of t he INTERSPILL, 2006,
pp. 110.
[19] D. Fritsch, K. Wegener, R.D. Schraft, Cont rol of a robot ic sw arm for the elimination of marine oil pollutions, in: IEEE Sw arm Int elligence Sym posium,
2007, pp. 2936.
[20] T.B. Sheridan, Hum ans and Aut omat ion: Syst em Design and Research Issues,
John W iley and Sons, 2002.
[21] B. Robins, K. Daut enhahn, R. Boerkhorst , A. Billard, Robot s as assistive
t echnologydoes appearance mat ter? in: IEEE Internat ional W orkshop on
Robot and Hum an Interact ive Comm unicat ion, 2004. ROM AN 2004. 13th,
2004, pp. 277282.
[22] M .A. Goodrich, A.C. Schult z, Humanrobot int eract ion: A survey, Found.
Trends Hum.-Comput . Int eract. 1 (2007) 203275.
[23] E.J. Rodriguez-Seda, D. Lee, M.W. Spong, Experimental comparison st udy of
cont rol archit ect ures for bilateral t eleoperat ors, IEEE Trans. Robot . 25 (6)
(2009) 13041318.
[24] E. Delgado, M. Diaz-Cacho, D. Bust elo, A. Barreiro, Generic approach to
stability under t im e-varying delay in teleoperat ion: Applicat ion t o the
position-error control of a gantry crane, IEEE/ASM E18 (5) (2013) 15811591.
[25] Y. Ye, Y.-J. Pan, T. Hilliard, Bilat eral teleoperat ion w it h t im e-varying delay:
A comm unicat ion channel passificat ion approach, IEEE/ASM E 18 (4) (2013)
14311434.
[26] C.-C. Hua, X.P. Liu, A new coordinat ed slave t orque feedback control
algorithm for netw ork-based t eleoperat ion syst ems, IEEE/ASM E18 (2) (2013)
764774.
[27] R. Haupt, A survey of priority rule-based scheduling, Oper. Res. Spekt rum 11
(1989) 316.
[28] H. Li, K. Womer, Scheduling projects w it h mult i-skilled personnel by a hybrid
M ILP/CP benders decomposit ion algorithm, J. Sched. 12 (2009) 281298.
[29] T.S. Dahl, M . Mat aric, G.S. Sukhat me, M ulti-robot task allocation t hrough
vacancy chain scheduling, Robot. Auton. Syst . 57 (6) (2009) 674687.
[30] T. Zheng, J. Li, Mult i-robot t ask allocat ion and scheduling based on fish sw arm
algorithm, in: Int elligent Control and Aut omat ion, 2010, pp. 66816685.
[31] B. Graf, K. Pfeiffer, H. St aab, Mobile robot s for offshore inspect ion and
manipulat ion, in: In Proc. of the Int. Petroleum Tech. Conf., Dubai, 2007,
pp. 16.
[32] B. Graf, K. Pfeiffer, M obile robot ics for offshore automation, in: Proceedings of
t he EURON/IARP Int ernational Workshop on Robot ics for Risky Int ervent ions
and Surveillance of the Environment , Benicassim, Spain, 2008.
[33] M . Bengel, K. Pfeiffer, B. Graf, A. Bubeck, A. Verl, M obile robot s for offshore
inspect ion and m anipulat ion, in: IEEE Conf. on Intelligent Robots and Sys.,
2009, pp. 33173322.
[34] L. M arce, M . Julliere, Dynamic localizat ion of a mobile robot t hrough range
measurement s, in: Cambridge Symposium_Int elligent Robot ics Syst ems, Vol.
0727, Int ernat ional Society for Opt ics and Photonics, 1987, pp. 274281.
[35] B. Jalving, J.E. Faugst admo, K. Vestgard, O. Hegrenaes, O. Engelhardt sen, B.
Hyland, Payload sensors, navigat ion and risk reduction for AUV under ice
surveys, in: Aut onom ous Underw at er Vehicles, AUV 2008, IEEE/OES, 2008,
pp. 18.

[36] A. Transeth, O. Skotheim , H. Schumann-Olsen, G. Johansen, J. Thielem ann,

E. Kyrkjebo, A robot ic concept for remote maintenance operat ions: A robust
3D object det ection and pose est imation met hod and a novel robot t ool, in:
Int elligent Robot s and Syst ems, IROS, 2010 IEEE/RSJInternat ional Conference
on, 2010, pp. 50995106.
[37] I. Schjolberg, Developm ent of robot ic solut ions for oil/gas, alum inum and
m anufacturing indust ry, in: Emerging Tech. Fact ory Aut o., ETFA, 2012,
pp. 14.
[38] F. Albring Gut h, L. Silveira, M. Am aral, S. Bot elho, P. Drew s, Underw at er
visual 3D SLAM using a bio-inspired syst em, in: Com put ing and Aut om at ion
for Offshore Shipbuilding, NAVCOM P, 2013 Symposium on, 2013, pp. 8792.
[39] R. Ditchburn, S. Burke, C. Scala, Ndt of w elds: stat e of t he art , NDT & E Int . 29
(2) (1996) 111117.
[40] J. Makar, N. Chagnon, Inspect ing syst ems for leaks, pit s, and corrosion,
J. Amer. W at er W orks Assoc. 91 (7) (1999) 3646.
[41] R. Bickerst aff, M . Vaughn, G. St oker, M. Hassard, M . Garrett , Review of sensor
t echnologies for in-line inspect ion of natural gas pipelines, Sandia Nat ional
Lab, Albuquerque, NM .
[42] J. Rose, A baseline and vision of ult rasonic guided w ave inspection potent ial,
J. Press. Vessel Technol. 124 (3) (2002) 273282.
[43] J. Jiao, C. He, B. Wu, R. Fei, X. W ang, Applicat ion of ult rasonic guided w aves in
pipes NDT, J. Exp. Mech. 1 (2002).
[44] K. Reber, M . Beller, H. W illems, O. Barbian, A new generat ion of ult rasonic inline inspect ion t ools for detecting, sizing and locating met al loss and cracks
in t ransmission pipelines, in: Ult rasonics Symposium, Vol. 1, IEEE, 2002,
pp. 665671.
[45] H. Shafeek, E. Gadelmaw la, A. Abdel-Shafy, I. Elew a, Assessment of w elding
defect s for gas pipeline radiographs using computer vision, NDT & E Int . 37
(4) (2004) 291299.
[46] J. Pretlove, C. Skourup, J. Gunnar, A. Ryry, Robot ics for int egrat ed remote
operat ions, in: Society of Petroleum Engineers, 2010, p. 1.
[47] D.A. Anisi, C. Skourup, A st ep-w ise approach t o oil and gas robotics, in: IFAC
W orkshop on Automatic Control in Offshore Oil and Gas Product ion, 2012,
pp. 4752.
[48] C. Skourup, J. Pret love, Rem ot e inspection and int ervent ion-remot e robot ics
at w ork in harsh oil and gas environment s, January 2012.
[49] H.A.F. Almurib, H.F. Al-Qrimli, N. Kumar, A review of applicat ion indust rial robot ic design, in: Ninth Int . Conf. on ICT and Know ledge Eng., 2012,
pp. 105112.
[50] V. Kongezos, B. Anisi, R. Krieger, T. Lillehagen, Hum anrobot cooperat ion for
error recovery in oil and gas applications, in: IET Conference on Cont rol and
Aut omat ion 2013: Unit ing Problems and Solut ions, 2013, pp. 16.
[51] J. Albiez, M . Hildebrand, T. Vogele, S. Joyeux, F. Kirchner, Robust robot s for
arct ic explorat ion, in: OTC Arct ic Technology Conference, 2011.
[52] D. M cLeod, Emerging capabilities for aut onomous inspect ion repair and
m aint enance, in: OCEAN, 2010, pp. 14.
[53] K. Pfeiffer, M. Bengel, A. Bubeck, Offshore robot icssurvey, implement at ion,
out look, in: IEEE/RSJ Internat ional Conference on Int elligent Robot s and
System s, 2011, pp. 241246.
[54] J. Cui, S. Tosunoglu, R. Roberts, C. Moore, D.W. Repperger, A review of
t eleoperation system cont rol, in: Proceedings of the Florida Conference on
Recent Advances in Robotics.
[55] P.C. Leger, A. Trebi-Ollennu, J.R. Wright, S.A. M axw ell, R.G. Bonit z, J.J.
Biesiadecki, F.R. Hart man, B.K. Cooper, E.T. Baum gart ner, M .W . M aimone,
M ars exploration rover surface operat ions: driving spirit at gusev crat er,
in: IEEE Internat ional Conference on Syst ems, M an and Cybernet ics, 2005,
pp. 18151822.
[56] S. St ram igioli, R. M ahony, P. Corke, A novel approach to haptic t ele-operat ion
of aerial robot vehicles, in: IEEE Int ernat ional Conference on Robotics and
Aut omat ion, ICRA.
[57] P. Wells, D. Deguire, TALON: A universal unmanned ground vehicle plat form,
enabling the mission t o be the focus, in: Proc. of SPIE, Unmanned Ground
Vehicle Technology, 2005.
[58] E.I. Agba, SeaM ast er: an ROV-manipulat or system simulator, IEEE Comput.
Graph. Appl. 15 (1) (1995) 2431.
[59] C.P. Sayers, R.P. Paul, L.L. W hit com b, D.R. Yoerger, Teleprogram ming for
subsea teleoperat ion using acoustic comm unicat ion, IEEE J. Ocean. Eng. 23
(1) (1998) 6071.
[60] H. Flem mer, B. Eriksson, J. W ikander, Cont rol design and st abilit y analysis of
a surgical teleoperator, M echatronics 9 (7) (1999) 843866.
[61] A. Halm e, J. Suoela, M . Savela, Applying telepresence and augment ed realit y
t o t eloperat e field robots, Robot. Aut on. Syst . (1999) 117125.
[62] S. All, I.R. Nourbakhsh, Insect t elepresence: Using robotic tele-embodiment
t o bring insects face t o face w it h hum ans, in: Aut omous Robots, Special Issue
on Personal Robotics, 2001.
[63] P. Milgram, J. Ballant yne, Real w ord teleperat ion via virt ual environm ent modelling, in: Internat ional Conference on Art ificial Realit y and
Tele-existence.
[64] S.-G. Honga, B.S. Kim, S. Kim, J.-J. Lee, Art ificial force reflect ion control for
t eleoperated mobile robot s, M echatronics 8 (6) (1998) 707717.
[65] S. Soldan, G. Bonow , A. Kroll, Robogasinspectora mobile robotic syst em
for rem ot e leak sensing and localization in large indust rial environments:
Overview and first results, in: Proceedings of the 2012 IFAC Workshop on
Aut omat ic Control in Offshore Oil and Gas Production, 2012, pp. 3338.
[66] C. Heyer, Humanrobot interact ion and future indust rial robot ics applicat ions, in: IEEE/RSJ, 2010, pp. 47494754.

A. Shukla, H. Karki / Robotics and Autonomous Systems 75 (2016) 490507

[67] H. Devold, Oil and gas product ion handbook an introduction t o oil and gas
product ion, transport , refining and petrochem ical indust ry, ABB Oil and Gas.
[68] J. Ferent inos, Global offshore oil and gas out look, 2013.
[69] S. Casat i, In the dept hs of the desert :t he challenges of building oil and gas
infrast ruct ure in a remote and inhospit able part of the sahara desert.
[70] H.T. Rom an, B.A. Pellegrino, W.R. Sigrist , Pipe craw ling inspect ion robots: an
overview , IEEE Trans. Energy Convers. 8 (3) (1993) 576583.
[71] Y. Kaw aguchi, I. Yoshida, H. Kurum at ani, T. Kikut a, Y. Yamada, Int ernal pipe
inspect ion robot , in: IEEE Conf on Robot ics and Aut omat ion, vol. 1, 1995,
pp. 857862.
[72] H. Roth, K. Schilling, S. Fut terknecht, U. W eigele, M . Reisch, Inspect ion and
repair robot s for w ast e w ater pipes a challenge to sensorics and locomotion,
in: IEEE Int . Symp. on Industrial Electronics, 1998, pp. 476478.
[73] S. gon Roh, H. Choi, St rat egy for navigation inside pipelines w it h differentialdrive inpipe robot, in: Proceedings of IEEE Int ernat ional Conference on
Robotics and Autom at ion, 2002, vol. 3, 2002, pp. 25752580.
[74] I. Hayashi, N. Iw at uki, M icro moving robot ics, in: Proceedings of t he Internat ional Symposium on Microm echat ronics and Human Science, 1998, pp.
4150.
[75] L. Sun, P. Sun, L. Lian, X. Qin, Z. Gong, Improvement of characterist ics of inpipe micro robot, in: Int . Symp. on M icromechatronics and Human Science,
1999, pp. 153156.
[76] K. Suzum ori, T. M iyagaw a, M . Kimura, Y. Hasegaw a, M icro inspection robot
for 1-in pipes, IEEE/ASME Trans. Mechat ronics 4 (3) (1999) 286292.
[77] C. Ant hierens, C. Libersa, M . Touaibia, M . Betemps, M. Arsicault , N. Chaillet ,
Micro robots dedicat ed to small diamet er canalizat ion explorat ion, in:
IEEE/RSJ, 2000, pp. 480485.
[78] N. Mit umot o, K. Tsurut a, T. Shibata, N. Kaw ahara, Wireless link syst em for
communication and energy transmission of microrobot, in: Proceedings of
Int ernat ional Sym posium on M icromechatronics and Human Science, 2001,
pp. 5762.
[79] E. Dertien, S. St ramigioli, K. Pulles, Development of an inspect ion robot for
sm all diamet er gas distribution mains, in: IEEE Int ernat ional Conference on
Robotics and Autom at ion, 2011, pp. 50445049.
[80] E. Dert ien, S. Stramigioli, Basic maneuvers for an inspect ion robot for small
diamet er gas dist ribut ion mains, in: IEEE Conf on Robotics and Automation,
2011.
[81] Y.-S. Kw on, B.-J. Yi, Design and motion planning of a t w o-m odule
collaborat ive indoor pipeline inspect ion robot , IEEE Trans. Robot . 28 (3)
(2012) 681696.
[82] J. Park, T. Kim , H. Yang, Development of an act ively adapt able in-pipe robot ,
in: IEEE Int ernat ional Conference on M echatronics, 2009, pp. 15.
[83] S. Hirose, H. Ohno, T. M it sui, K. Suyama, Design of in-pipe inspection vehicles
for 25 , 50 , 150 pipes, in: IEEE Conf. on Robot ics and Aut om at ion, 1999,
pp. 23092314.
[84] C. Jun, Z. Deng, S. Jiang, St udy of locomot ion cont rol charact erist ics for six
w heels driven in-pipe robot, in: IEEE Internat ional Conference on Robot ics
and Biomimet ics, 2004. ROBIO 2004, 2004, pp. 119124.
[85] T. Oya, T. Okada, Development of a st eerable, w heel-t ype, in-pipe robot and
it s pat h planning, Adv. Robot . 19 (6) (2005) 635650.
[86] S. gon Roh, H.R. Choi, Differential-drive in-pipe robot for m oving inside urban
gas pipelines, IEEE Trans. Robot. 21 (1) (2005) 117.
[87] B.B. Gam ble, R.M . W iesman, Tethered mouse syst em for inspect ion of gas
dist ribut ion mains, Tech. Rep., Gas Research Instit ut e, Foster-Miller, Inc.,
Waltham , M A, 1996.
[88] K.U. Scholl, V. Kepplin, K. Berns, R. Dillmann, Controlling a m ult i-joint robot
for autonom ous sew er inspection, in: IEEEConf on Robotics and Automation,
2000, pp. 17011706.
[89] H. Schem pf, E. Mut schler, V. Goltsberg, G. Skopt sov, A. Gavaert, G. Vradis,
Explorer: Unt ethered real-time gas main assessment robot system , in:
Proceedings of Internat ional W orkshop on Advances in Service Robot ics,
ASER03, 2003.
[90] S.M. Ryew , S.H. Baik, S.W . Ryu, K.M . Jung, S.G. Roh, H.R. Choi, In-pipe inspect ion robot system w ith active steering m echanism, in: IEEE/RSJConf on
Int elligent Robots and Systems, vol. 3, 2000, pp. 16521657.
[91] H.R. Choi, S.M. Ryew , Robotic system w ith active steering capabilit y for
internal inspection of urban gas pipelines, M echatronics 12 (5) (2002)
713736.
[92] T. Fukuda, H. Hosokai, M . Uemura, Rubber gas act uat or driven by hydrogen
st orage alloy for in-pipe inspection mobile robot w ith flexible st ruct ure,
in: IEEE Int ernat ional Conference on Robot ics and Automation, 1989,
pp. 18471852.
[93] S.G. Roh, S.M . Ryew , J.H. Yang, H.R. Choi, Actively steerable in-pipe inspection robot s for underground urban gas pipelines, in: IEEE Int Conference on
Robotics and Autom at ion, 2001, pp. 761766.
[94] Z. Deng, H. Chen, S. Jiang, D. Sun, Traction robot driven by six independent
w heels for inspection inside pipeline, Chin. M ech. Eng. J. 41 (9) (2005) 6771.
[95] S. gon Roh, D.W . Kim , J.-S. Lee, H. Moon, H.R. Choi, Modularized in-pipe robot
capable of select ive navigation inside of pipelines, in: IEEE/RSJInternat ional
Conference on Int elligent Robots and Syst ems, 2008, pp. 17241729.
[96] S.G. Roh, D.W. Kim, J.S. Lee, H. M oon, H.R. Choi, In-pipe robot based on
selective drive mechanism, Mech. M ach. Theory 7 (1) (2009) 105112.
[97] H. Qi, X. Zhang, H. Chen, J. Ye, Tracing and localization syst em for pipeline
robot , Mechat ronics 19 (1) (2009) 7684.
[98] H. Qi, J. Ye, X. Zhang, H. Chen, W ireless t racking and locat ing system for inpipe robot, Sensors Act uators A 159 (1) (2010) 117125.

503

[99] X. Gao, S. Jiang, X. Jiang, H. Sun, Analyzing of t ri-axis different ial in-pipe robot
w it h adapt ive differential capability for moving inside pipelines, in: Conf on
M echatronic Science, Elect ric Engineering and Com put er, 2011, pp. 564567.
[100] J. Okamot o Jr., J.C. Adamow ski, M.S.G. Tsuzuki, F. Buiochi, C.S. Camerini,
Aut onomous syst em for oil pipelines inspect ion, Mechat ronics 9 (7) (1999)
731743.
[101] J.M. Lopez, S. Sadovnychiy, Small pig for inspect ion pipeline, in: Elect ronics,
Robot ics and Aut om ot ive M echanics Conference, 2007. CERM A 2007, 2007,
pp. 585590.
[102] T. Okada, T. Kanade, A three-w heeled self-adjust ing vehicle in a pipe, FERRET1, J. Robot . 6 (4) (1987) 6075.
[103] T. Okada, T. Sanem ori, M ogrer: A vehicle st udy and realization for in-pipe
inspection tasks, IEEE J. Robot . Autom. 3 (6) (1987) 573582.
[104] K. Taguchi, N. Kaw arazaki, Development of in-pipe locomot ion robot , in:
Conf. on Advanced Robotics, Robot s in Unst ruct ured Environments, 1991,
pp. 297302.
[105] M . Muramatsu, N. Namiki, R. Koyama, Y. Suga, Autonomous mobile robot
in pipe for piping operat ions, in: Proceedings of IEEE/RSJ Int ernat ional
Conference on Int elligent Robot s and System s, 2000, IROS2000, vol. 3, 2000,
pp. 21662171.
[106] H.J. Chen, B.T. Gao, X.H. Zhang, Z.Q. Deng, Drive control syst em for pipeline
craw l robot based on can bus, in: Journal of Physics, Series Int ernat ional
Symposium on Instrum ent ation Science and Technology, vol. 48, 2006,
pp. 12331237.
[107] M .M . M oghaddam, A. Hadi, Cont rol and guidance of a pipe inspection craw ler
(PIC), in: 22nd Internat ional Sym posium on Aut om at ion and Robot ics in
Const ruct ion, 2005, pp. 15.
[108] Y. Zhang, G. Yan, In-pipe inspection robot w ith act ive pipe-diam eter
adaptability and automatic t ract ive force adjust ing, Mech. M ach. Theory 42
(12) (2007) 16181631.
[109] P. Li, S. Ma, B. Li, Y. W ang, Development of an adaptive mobile robot for
in-pipe inspect ion t ask, in: Int . Conference on M echatronics and Aut om at ion,
2007, pp. 36223627.
[110] P. Li, S. Ma, B. Li, Y. W ang, Design of a mobile mechanism possessing driving
ability and det ecting funct ion for in-pipe inspect ion, in: IEEE Int ernat ional Conference on Robotics and Automation, 2008. ICRA 2008, 2008,
pp. 39923997.
[111] T. Aoki, S. Hirose, Study on t he brake operat ion of bridle bellow s,
in: Int elligent Robots and Sys., IEEE, 2007, pp. 4045.
[112] S.A. Fjerdingen, P. Liljeback, A.A. Transeth, A snake-like robot for int ernal
inspection of complex pipe st ruct ures, in: Int elligent Robots and System s,
IEEE, 2009, pp. 56655671.
[113] T. M aneew arn, B. Maneechai, Design of pipe craw ling gaits for a snake robot,
in: IEEE Int ernational Conference on Robot ics and Biomimet ics, 2008. ROBIO
2008, 2009, pp. 16.
[114] Y. Kim, M .A. M inor, Dist ributed kinemat ic m ot ion cont rol of mult i-robot
coordination subject t o physical constraints, Int. J. Robot . Res. 29 (1) (2010)
92109.
[115] W . Neubauer, A spider-like robot t hat climbs vert ically in duct s or pipes,
in: IEEE/RSJ/GI Int ernational Conference on Int elligent Robots and System s,
1994, pp. 11781185.
[116] F. Nickols, D. Ho, S.O. Harrold, R.T. Bradbeer, L. Yeung, An ultrasonically cont rolled robot submarine for pipe inspection, in: M echatronics and Machine
Vision, 1997, pp. 142147.
[117] G.V. Kostin, F.L. Chernousko, N.N. Bolotnik, F. Pfeiffer, Regular m ot ions of
a t ube-craw ling robot: simulat ion and optimizat ion, in: Robot Motion and
Cont rol, 1999, pp. 4550.
[118] F. Pfeiffer, T. Rossmann, K. Loffler, Cont rol of a tube craw ling machine, in:
Conf. on Control of Oscillat ions and Chaos, vol. 3, 2000, pp. 586591.
[119] J.-H. Kim, G. Sharma, S.S. Iyengar, FAM PER: a fully aut onom ous m obile
robot for pipeline explorat ion, in: IEEE Conf. on Indust rial Technology, 2010,
pp. 517523.
[120] T. Fukuda, H. Hosokai, M. Ot suka, Aut onomous pipeline inspection and
m aint enance robot w ith inch w orm mobile mechanism , in: Robot ics and
Aut omat ion. Proceedings. 1987 IEEE Int ernational Conference on, vol. 4,
1987, pp. 539544.
[121] S.E. Landsberger, B.F. M art in, The design of a pipe craw ling robot for cont rol
of zebra mussel infestat ions, in: M ast ering t he Oceans Through Technology,
Vol. 2, 1992, pp. 819824.
[122] M . Takahashi, I. Hayashi, N. Iw atsuki, K. Suzumori, N. Ohki, The development
of an in-pipe microrobot applying t he mot ion of an earthw orm , in: Int.
Symposium on Micro M achine and Human Science, 1994, p. 35.
[123] Y. Kondoh, S. Yokota, M icro in-pipe m obile machines by m aking use of an
electro-rheological fluid, in: IEEE/RSJ Conf. on Intelligent Robot s and Sys.,
vol. 3, 1997, pp. 16721677.
[124] T. Shibat a, T. Sasaya, N. Kaw ahara, Microw ave energy supply system for
in-pipe m icromachine, in: Int . Symp. on Microm echat ronics and Human
Science, 1998, pp. 237242.
[125] H. Nishikaw a, T. Sasaya, T. Shibata, T. Kaneko, N. M it umoto, S. Kaw akit a,
N. Kaw ahara, In-pipe w ireless micro locom ot ive system , in: Proceedings of
1999 Internat ional Sym posium on Microm echat ronics and Human Science,
1999, pp. 141147.
[126] C. Ant hierens, A. Ciftci, M. Bet emps, Design of an electro pneumatic micro
robot for in-pipe inspect ion, in: Sym posium on Indust rial Elect ronics, IEEE,
1999, pp. 968972.

504

A. Shukla, H. Karki / Robotics and Autonomous Systems 75 (2016) 490507

[127] K. Tsuruta, T. Sasaya, T. Shibat a, N. Kaw ahara, Cont rol circuit in an in-pipe
w ireless micro inspection robot , in: Int. Symp. on M icromechatronics and
Hum an Science, 2000, pp. 5964.
[128] A.M . Bert ett o, M . Ruggiu, In-pipe inch-w orm pneum at ic flexible robot , in:
IEEE/ASME Int . Conf. on Advanced Int elligent M echatronics, vol. 2, 2001,
pp. 12261231.
[129] A. M enciassi, J.H. Park, S. Lee, S. Gorini, P. Dario, J.-O. Park, Robotic solutions and m echanism s for a semi-autonom ous endoscope, in: IEEE/RSJ
Internat ional Conference on Int elligent Robots and Syst ems, vol. 2, 2002,
pp. 13791384.
[130] J. Lim , H. Park, J. An, Y.-S. Hong, B. Kim, B. Yi, One pneumat ic line based
inchw orm-like micro robot for half-inch pipe inspect ion, M echatronics 18
(7) (2008) 315322.
[131] S. Iw ashina, I. Hayashi, N. Iw at suki, K. Nakamura, Developm ent of in-pipe
operation micro robot s, in: 5t h Int . Symp. on M icro M achine and Human Sci.,
1994, p. 41.
[132] I. Hayashi, N. Iw atsuki, S. Iw ashina, The running charact eristics of a screw principle microrobot in a small bent pipe, in: Micro M achine and Hum an
Science, 1995, pp. 225228.
[133] Y. Kw on, B. Lee, I.-C. W hang, B. Yi, A pipeline inspect ion robot w ith a linkage
t ype mechanical clutch, in: IEEE Int. Conf. on Intelligent Robot s and Syst ems,
2010, pp. 28502855.
[134] P. Ong, W . Anderson, B. Cook, R. Subramanyan, A novel x-ray technique for
inspect ion of steel pipes, J. Nondest ruct . Eval. 13 (4) (1994) 165173.
[135] H. Chen, J. Li, X. Zhang, Z. Deng, Application of visual servoing t o an x-ray
based w elding inspect ion robot , in: Control and Aut omat ion, 2005. ICCA05.
Internat ional Conference on, Vol. 2, IEEE, 2005, pp. 977982.
[136] J. Tabary, P. Hugonnard, F. Mat hy, Sindbad: a realistic m ulti-purpose and
scalable x-ray simulation t ool for ndt applicat ions, in: Int . Symp. on DIR and
CT, Lyon, vol. 1, 2007, pp. 110.
[137] J.N.H. Tiratsoo, Pipeline Pigging and Inspection Technology, 1992.
[138] J. Pei, M .I. Yousuf, F.L. Degert ekin, B.V. Honein, B.T. Khuri-Yakub, Lam b w ave
t omography and it s applicat ion in pipe erosion/corrosion monit oring, in:
Proceedings of IEEE on Ult rasonics Symposium, vol. 1, 1995, pp. 795798.
[139] J.L. Rose, Guided w ave nuances for ult rasonic nondestructive evaluation, IEEE
Trans. Ult rason. Ferroelectr. Freq. Cont rol 47 (3) (2000) 575583.
[140] S. Li-ying, Y. Xiao-dong, L. Yi-bo, Research on t ransducer and frequency
of ult rasonic guided w aves in urban pipe inspect ion, in: 4th IEEE Conference on Industrial Elect ronics and Applications, 2009, ICIEA 2009, 2009,
pp. 27082711.
[141] F. Kojima, J.i. Tsubota, Development of guided w ave sim ulat or and applicat ion t o sizing met hodologies of pipe w all t hinning, in: ICROS-SICE Int . Joint
Conf., 2009, pp. 23272332.
[142] W . Ang, Z. Juan-hua, H. Fu-yun, H. Jian-dong, W . Ling, Measuring syst em for
t he w all t hickness of pipe based on ult rasonic multisensor, in: Int . Conf. on
Elect ronic M easurem ent Inst rument s, vol. 1, 2009, pp. 641644.
[143] S. Qi, J. Liu, G. Jia, Study of submarine pipeline corrosion based on ult rasonic
det ect ion and w avelet analysis, in: Int ernational Conference on Comput er
Application and Syst em M odeling, ICCASM , 2010, vol. 12, 2010, pp. 440444.
[144] T. Kikut a, K. Tomita, K. Yasui, Developm ent of remote field eddy current
inspect ion syst em for gas dist ribut ion pipes, in: Int . Gas Research Conf., vol.
1, 1986, pp. 122130.
[145] T. Idogaki, H. Kanayam a, N. Ohya, H. Suzuki, T. Hat t ori, Characterist ics
of piezoelectric locomotive mechanism for an in-pipe micro inspection
machine, in: 6t h Int . Symp. on Micro Machine and Human Sci., 1995,
pp. 193198.
[146] F. Tache, W . Fischer, R. Moser, F. M ondada, R. Siegw art , Adapt ed magnet ic
w heel unit for compact robot s inspecting complex shaped pipe structures,
in: IEEE/ASM E Int . Conf. on Adv. Intelligent M echatronics, 2007, pp. 16.
[147] G. Dobmann, G. W alle, P. Haller, M agnet ic leakage flux t esting w it h probes:
physical principles and restrict ions for applicat ion, NDT Int . 20 (2) (1987)
101104.
[148] A.C. Bruno, R. Schifini, G.S. Khner, C.H. Barbosa, J.P. Wiksw o Jr., C.S. Camerini,
New magnet ic t echniques for inspection and met al-loss assessm ent of oil
pipelines, J. M agn. Magn. M ater. 226230 (Part 2) (2001) 20612062.
[149] J. Tao, Q. Peiw en, T. Zhengsu, Development of m agnetic flux leakage pipe
inspect ion robot using hall sensors, in: Micro-Nanomechatronics and Hum an
Sci., 2004, pp. 325329.
[150] Z. Guoguang, L. Jing, Finit e element modelling of circumferent ial magnet ic
flux leakage inspection in pipeline, in: Intelligent Comput ation Tech. and
Automation, 2010, pp. 327330.
[151] Z. Zakaria, M.S.B. Mansor, A.H. Jahidin, M .S.Z. Azlan, R.A. Rahim, Simulation of
magnet ic flux leakage (M FL) analysis using fem m soft w are, in: IEEE Sym posium on Industrial Elect ronics Applications, ISIEA, 2010, 2010, pp. 481486.
[152] S. LiYing, L. YiBo, S. LiBo, L. LingGe, Comparison of magnetic flux leakage
(M FL) and acoust ic emission (AE) t echniques in corrosion inspect ion for
pressure pipelines, in: 31st Chinese Cont rol Conference, CCC, 2012, 2012,
pp. 53755378.
[153] P.J. Shull, Nondestructive Evaluation: Theory, Techniques, and Applicat ions,
M echanical Eng. Ref. Books, 2002.
[154] A. Rerkratn, T. Cheypoca, A. Kaew poonsuk, Syst em development for tube
inspect ion based on eddy current t echnique, in: SICE Annual Conf., 2012,
pp. 69.
[155] Y. Hamasaki, T. Ide, Fabricat ion of m ult i-layer eddy current micro sensors
for non-dest ruct ive inspect ion of sm all diamet er pipes, in: IEEE on M icro
Elect ro Mechanical System s, 1995.

[156] Y. Hamasaki, T. Ide, A m ult i-layer eddy current micro sensor for nondest ruct ive inspect ion of small diam eter pipes, in: Solid-St at e Sensors and
Act uat ors, and Eurosensors, 1995.
[157] R. Palanisamy, Theoretical predict ion of remote field eddy current response
for t he nondestructive evaluat ion of met allic t ubes, J. Appl. Phys. 61 (8) (1987)
32023204.
[158] Y. Zhao, Y. Wang, Three-dim ensional finit e element comput ation of elect rom agnet ics for nondest ruct ive t est ing, in: Ant ennas and Propagat ion Societ y,
Vol. 1, 1996, pp. 166169.
[159] S.M. Haugland, Fundam ent al analysis of t he rem ot e-field eddy-current effect,
IEEE Trans. M agn. 32 (4) (1996) 31953211.
[160] D. At herton, W. Czura, Finit e element poynt ing vector calculat ions for remote
field eddy current inspection of tubes w ith circum ferential slots, IEEE Trans.
M agn. 27 (5) (1991) 39203922.
[161] G.A. M at zkanin, H.T. Yolken, Nondest ruct ive evaluation techniques for
hidden corrosion, 2012.
[162] B. Dennis, Casing corrosion evaluat ion using w ireline techniques, J. Can. Pet.
Technol. (1990).
[163] L. Bo, L. Feilu, J. Zhongqing, L. Jiali, Eddy current array inst rument and
probe for crack detect ion of aircraft tubes, in: Internat ional Conference on
Int elligent Com put at ion Technology and Automation, ICICTA, 2010, vol. 2,
2010, pp. 177180.
[164] J. Yin, M . Lu, J.P. de Gyvez, Full-signature real-t ime corrosion det ect ion of
underground casing pipes, IEEETrans. Inst rum. Meas. 49 (1) (2000) 120128.
[165] Z. Xingfu, W . Yingxiang, L. Qiang, Y. Xiw ei, M ulti-pipe st ring elect romagnetic
det ection t ool and its applicat ions, in: Conf. on Elect ronic Measurement and
Instrum ent s, 2007, pp. 423427.
[166] T. Beuker, J. Palm er, M. Quack, In-line inspection using com bined t echnologiesmagnet ic flux leakage and ult rasonic test ing and their advant ages, in:
4t h Pipeline Tech. Conf., 2009.
[167] A. Pople, P. Int egrit y, M agnet ic flux leakage pigs or ult rasonic pigs? The
case for com bined int elligent pig inspect ions, in: Pipeline Rehabilitation and
M aintenance, 2003, pp. 610.
[168] X. Song, X. W u, Y. Kang, An inspect ion robot for boiler tube using magnetic
flux leakage and ultrasonic m ethods, Insight -Non-Dest r. Test. Cond. Monit.
46 (5) (2004) 275278.
[169] D. Horn, W . Mayo, Nde reliability gains from com bining eddy-current and
ult rasonic test ing, NDT & E Int. 33 (6) (2000) 351362.
[170] J. Oagaro, S. M andayam , M ult i-sensor dat a fusion using geom etric transformations for gas transmission pipeline inspection, in: IMTC, IEEE, 2008,
pp. 17341737.
[171] R.R. Costa, N.R.S. dos Reis, L. Hsu, A.J. Peixoto, P.C.S. Gomes, Robot ized syst em
for in-pipe inspect ion, in: IEEE Internat ional Sym posium on Indust rial
Electronics, 2003.
[172] D. Lee, H. M oon, H. Choi, Landmark detect ion of in-pipe w orking robot using
line-laser beam project ion, in: Int . Conf. on Cont rol Automation and System s,
2010, pp. 611615.
[173] Y. Hu, Z. Song, J. Zhu, Est imating t he post ure of pipeline inspect ion robot w it h
a 2D laser rang finder, in: Mult isensor Fusion and Integration for Intelligent
Sys., 2012, pp. 401406.
[174] J.K. Ong, D. Kerr, K. Bouazza-M arouf, In-pipe mult i-robot syst em: M odular
configurable co-operative semi-aut onom ous robot ic units, in: Int. Conf. on
Gas Research, 2001, pp. 4752.
[175] T. Izumikaw a, H. Yaguchi, Movement of a cableless in-piping magnetic
actuat or w ith a new propulsion module, IEEE Trans. M agn. 48 (11) (2012)
41964199.
[176] A.A.F. Nassiraei, Y. Kaw amura, A. Ahrary, Y. M ikuriya, K. Ishii, A new approach
t o the sew er pipe inspection: Fully autonomous mobile robot KANTARO,
in: IEEE Indust rial Elect ronics, 2006, pp. 40884093.
[177] A.A.F. Nassiraei, Y. Kaw am ura, A. Ahrary, Y. Mikuriya, K. Ishii, Concept
and design of a fully aut onomous sew er pipe inspection mobile robot
KANTARO, in: IEEE Conf. on Robot ics and Aut omat ion, 2007, pp. 136143.
[178] A.A.F. Nassiraei, M . Honda, K. Ishii, New approach to t he self-localizat ion of
aut onomous sew er inspection robots, in: W orld Aut omat ion Congress, W AC,
2010, 2010, pp. 17.
[179] A. Bouchalkha, M .S. Hamad, K. Al-Beloushi, M . Al-Qayedi, K. Al-Hammadi,
Design of an oil pipe inner surface inspect ion system , in: GCC Conference
and Exhibit ion, 2011, pp. 2932.
[180] D.A. Anisi, J. Gunnar, T. Lillehagen, C. Skourup, Robot aut omation in oil
and gas facilit ies: Indoor and onsit e dem onst rat ions, in: IEEE/RSJ Int ernational Conference on Int elligent Robot s and Systems, IROS, 2010, 2010,
pp. 47294734.
[181] X. Li, W. Yu, X. Lin, S.S. Iyengar, On opt imizing aut onomous pipeline
inspect ion, IEEE Trans. Robot . 28 (1) (2012) 223233.
[182] G. Cambell, K. Rogers, J. Gilbert , Pirat a syst em for quant itative sew er
assessment, in: Proceedings of t he 12t h Int ernational No-Dig Conference,
Ham burg, 1995, pp. 455462.
[183] R.M. Gooch, T.A. Clarke, T.J. Ellis, A semi-autonom ous sew er surveillance and
inspect ion vehicle, in: IEEE Int elligent Vehicles Symp., 1996, pp. 6469.
[184] R. Kirkham , P.D. Kearney, K.J. Rogers, Pirata syst em for quant it at ive sew er
assessment, in: Int . Conf. Field and Service Robotics, CM U, Pit sburgh, 1999,
pp. 712.
[185] J.Y. Choi, H. Lim, B.-J. Yi, Semi-aut om at ic pipeline inspection robot system s,
in: SICE-ICASE, 2006. Int ernat ional Joint Conference, 2006, pp. 22662269.
[186] E. Rome, J. Hert zberg, F. Kirchner, U. Licht , T. Christaller, Tow ards aut onom ous sew er robots: t he makro project, in: IEEEInt ernat ional Conference
on Industrial Technology, 1999.

A. Shukla, H. Karki / Robotics and Autonomous Systems 75 (2016) 490507

[187] H.-P. Huang, J.-L. Yan, T.-H. Cheng, Development and fuzzy cont rol of a pipe
inspect ion robot, IEEE Trans. Ind. Elect ron. 57 (3) (2010) 10881095.
[188] A. Groysman, Corrosion for Everybody, Springer, Dordrecht , Heidelberg,
London, New York, 2010.
[189] A.S.A. Abdulla, A.M .A. Khalil, A. Shehhi, H. Karki, S.C. Fok, Localizat ion of a submersible mobile inspect ion platform in an oil storage t ank, in: 7t h Int ernational Symposium on Mechat ronics and it s Applications, ISM A, 2010, pp. 16.
[190] A.C. da Cruz, M.S. Ribeiro, Robtank inspecin service robotized inspection
tool for hazardous product s st orage t anks, Ind. Robot : Int. J. 32 (2) (2005)
157162.
[191] L. Kalra, J. Gu, M. M eng, A w all climbing robot for oil t ank inspection, in:
Robotics and Biomimet ics, 2006, pp. 15231528.
[192] W. Yan, L. Shuliang, X. Dianguo, Z. Yanzheng, S. Hao, G. Xueshan, Development and applicat ion of w all-climbing robot s, in: IEEE Int . Conf. on Robot ics
and Aut om at ion, 1999.
[193] H. Slocum, S. Aw t ar, J. Hart, Magnebots: M agnetic w heels based overhead
transport at ion concept , in: IFAC M echat ronics.
[194] T. Sogi, Y. Kaw aguchi, T. Morisaki, K. Ohkaw a, N. Kai, H. Hayakaw a, Inspection robot for spherical st orage tanks, in: IEEE Indust rial Electronics Society,
Vol. 1, 2000, pp. 393398.
[195] Anon., Tripoda revolution in remote t ransportat ion, November 2012.
[196] T. Yukaw a, H. Okano, S. Komatsubara, M echanisms for the m ovement of
piping inspection robot w ith magnet ic element s, in: Int . Conf. Robotics and
Applications, 2005.
[197] J.C. Grieco, M. Priet o, M .A.P.G. de Sant os, A six-legged climbing robot for high
payloads, in: IEEEInt . Conf. on Control Applications, vol. 1, 1998, pp. 446450.
[198] S. Hirose, H. Tsut sumitake, Disk Rover: A w all-climbing robot using permanent magnet ic discs, in: lEEE Int . Conf. on Int elligent Robot s and Syst ems,
vol. 3, 1992, pp. 20742079.
[199] W. Shen, J. Gu, Y. Shen, Proposed w all climbing robot w ith permanent
magnet ic t racks for inspect ing oil tanks, in: IEEE Int. Conf. on M echatronics
and Aut om at ion, vol. 4, 2005.
[200] S. Kit ai, K. Tsuru, S. Hirose, The proposal of sw arm type w all clim bing robot
syst em anchor clim berdesign and examination of the adhering mobile unit ,
in: IEEE/RSJ Int ernational Conference on Intelligent Robot s and Syst ems,
2005.
[201] M. Eich, T. Vogele, Design and cont rol of a lightw eight m agnet ic clim bing
robot for vessel inspect ion, in: M editerranean Conf. on Control Automation,
MED, 2011, pp. 12001205.
[202] A.L.C. Oliveira, M .F. Silva, R.S. Barbosa, Architecture of an w heeled clim bing
robot w it h dynamic adjust ment of t he adhesion system , in: 8t h Internat ional
Symposium on Intelligent Systems and Inform at ics, SISY, 2010, pp. 127132.
[203] J. Xiao, A. Sadegh, M . Elliott , A. Calle, V. Persad, H.M . Chiu, Design of mobile
robot s w it h w all climbing capabilit y (overlook on pneumatic robot s), in:
IEEE/ASME Internat ional Conference on Advanced Int elligent M echatronics,
2005, pp. 438443.
[204] R.T. Pack, J.L. Christ opher Jr., K. Kaw amura, A rubbertuat or-based st ruct ureclim bing inspection robot, in: IEEE Int ernat ional Conference on Robotics and
Aut om at ion, vol. 3, 1997, pp. 18691874.
[205] D. Longo, G. Muscat o, The Alicia3 clim bing robot : a three-m odule robot for
automatic w all inspection, IEEE Robot . Aut om. M ag. 13 (1) (2006) 4250.
[206] S. Yam am ot o, Developm ent of inspection robot for nuclear pow er plant, IEEE
Robot. Autom. 2 (1992) 15591566.
[207] R. Aracil, R.J. Saltaren, O. Reinoso, A climbing parallel robot : a robot t o clim b
along tubular and m etallic structures, IEEE Robot . Autom. M ag. 13 (1) (2006)
1622.
[208] H.R. Choi, S.M. Ryew , T. Kang, J. Lee, H.M . Kim, A w all climbing robot w ith
closed link mechanism, in: IEEE Conf. on Intelligent Robots and Sys., vol. 3,
2000, pp. 20062011.
[209] H. Schempf, B. Chemel, N. Everet t, Nept une: above-ground st orage t ank
inspect ion robot system , IEEE Robot. Aut om. M ag. 2 (2) (1995) 915.
[210] H. W ang, X. Huang, R. Hong, C. Fang, A new inspect ion robot syst em for
st orage tank, in: 7th W orld Congress on Int elligent Control and Automation,
2008, pp. 74277431.
[211] R.F. Andndez, E.G. andlez, V. Feliu, A. Rodri andguez, A w all climbing robot for
tank inspect ion. an aut onomous prot ot ype, in: IEEE Indust rial Electronics,
2010, pp. 14241429.
[212] W. Fischer, F. Tache, R. Siegw art , Inspection syst em for very thin and fragile
surfaces, based on a pair of w all climbing robot s w it h m agnet ic w heels, in:
IEEE/RSJInt . Conf. on Int elligent Robots and Sys., 2007, pp. 12161221.
[213] L.P. Kalra, W. Shen, J. Gu, A w all climbing robotic system for non dest ruct ive
inspect ion of above ground t anks, in: Canadian Conf. on Elect rical and
Com put er Eng., 2006, pp. 402405.
[214] J. Gleason, A. Nefian, X. Bouyssounousse, T. Fong, G. Bebis, Vehicle det ection
from aerial imagery, in: IEEE Conf. on Robot ics and Aut omat ion, ICRA, 2011,
pp. 20652070.
[215] J.M . Sullivan, Revolut ion or evolution? The rise of the UAVs, in: Technology
and Society, 2005. W eapons and W ires: Prevent ion and Safet y in a Time of
Fear. ISTAS2005. Proceedings. 2005 Int ernat ional Symposium on, IEEE, 2005,
pp. 94101.
[216] J.A. Barnard, The use of unmanned aircraft in oil, gas and mineral E & P
activities, in: Explorat ion Geophysicist s, 2008.
[217] D. Lee, A. Franchi, H.I. Son, H. Changsu, H. Bult hoff, P. Giordano, Semiautonomous haptic teleoperat ion control archit ect ure of multiple unmanned
aerial vehicles, IEEE/ASM E Trans. M echatronics 18 (4) (2013) 13341345.
[218] D. Lyon, A milit ary perspective on small unmanned aerial vehicles, IEEE
Inst rum. Meas. Mag. 7 (3) (2004) 2731.

505

[219] S. Davies, UAVs in t he firing line, Eng. Technol. 6 (8) (2011) 3436.
[220] M . Bernard, K. Kondak, I. M aza, A. Ollero, Aut onomous t ransport ation and
deployment w ith aerial robot s for search and rescue missions, J. Field Robot.
28 (6) (2011) 914931.
[221] J. Tisdale, Z. Kim, J. Hedrick, Aut onomous UAV pat h planning and est im at ion,
IEEE Robot. Aut om. M ag. 16 (2) (2009) 3542.
[222] A. M olnar, The geo-historical legacies of urban security governance and t he
vancouver 2010 olym pics, Geograph. J. (3) (2015) 235241.
[223] R.R. M urphy, K.L. Dreger, S. New some, J. Rodocker, B. Slaughter, R. Sm it h,
E. Steimle, T. Kim ura, K. M akabe, K. Kon, H. M izum ot o, M . Hatayama, F.
M atsuno, S. Tadokoro, O. Kaw ase, Marine heterogeneous multirobot system s
at t he great East ern Japan Tsunami recovery, J. Field Robot . 29 (5) (2012)
819831.
[224] P.J. Hardin, T.J. Hardin, Small-scale rem ot ely piloted vehicles in environment al research, Geography Compass 4 (9) (2010) 12971311.
[225] A. Lin, A. Novo, S. Har-Noy, N. Ricklin, K. St am at iou, Combining geoeye1 satellit e remot e sensing, UAV aerial imaging, and geophysical surveys in
anomaly detect ion applied to archaeology, IEEE J. Sel. Top. Appl. Earth Obs.
Remot e Sens. 4 (4) (2011) 870876.
[226] H. Eisenbeiss, M . Sauerbier, Invest igat ion of UAV syst ems and flight modes for
phot ogrammet ric applicat ions, Phot ogram m. Rec. 26 (136) (2011) 400421.
[227] L. M erino, F. Caballero, J.M . de Dios, J. Ferruz, A. Ollero, A cooperat ive
perception syst em for multiple UAVs: Applicat ion t o automatic det ect ion of
forest fires, J. Field Robot . 23 (34) (2006) 165184.
[228] R. Beard, T. McLain, D. Nelson, D. Kingst on, D. Johanson, Decentralized
cooperat ive aerial surveillance using fixed-w ing miniat ure UAVs, Proc. IEEE
94 (7) (2006) 13061324.
[229] F. Lipez-Granados, W eed det ect ion for site-specific w eed managem ent :
m apping and real-t ime approaches, W eed Research 51 (1) (2011) 111.
[230] A. Barrientos, J. Colorado, J.d. Cerro, A. Martinez, C. Rossi, D. Sanz, J. Valente,
Aerial remote sensing in agricult ure: A pract ical approach to area coverage
and path planning for fleet s of mini aerial robot s, J. Field Robot . 28 (5) (2011)
667689.
[231] J. Rasmussen, J. Nielsen, F. Garcia-Ruiz, S. Christ ensen, J.C. Streibig, Potential
uses of small unmanned aircraft syst ems (uas) in w eed research, W eed Res.
53 (4) (2013) 242248.
[232] P. Zhan, K. Yu, A. Sw indlehurst, W ireless relay communications w it h
unmanned aerial vehicles: Performance and opt im izat ion, IEEE Trans.
Aerosp. Electron. Syst . 47 (3) (2011) 20682085.
[233] N. Goddemeier, K. Daniel, C. W iet feld, Role-based connect ivity management
w it h realist ic air-t o-ground channels for cooperat ive UAVs, IEEE J. Sel. Areas
Commun. 30 (5) (2012) 951963.
[234] J. M acnae, Design specificat ions for a geophysical unm anned air vehicle
assem bly (GUAVAS), in: Societ y of Explorat ion Geophysicist s, 1995.
[235] Anon., Ingrobotics aviat ion, htt p://ingrobotic.com /sect ors/oil-gas/ (accessed
14.08.14).
[236] S. Tiffin, R. Pilkington, C. Hill, M. Debicki, D. M cGonigal, W . Jolles, A decisionsupport system for ice/iceberg surveillance, advisory and management
act ivit ies in offshore pet roleum operat ions, in: Offshore Tech. Conf., 2014.
[237] M . Connolly, P. Buchan, The use of m ult i rot or rem ot ely operated aerial
vehicles (ROAVs) as a m ethod of close visual inspection (CVI) of live and
difficult t o access assets on offshore plat form s, in: Societ y of Pet roleum
Engineers, 2013.
[238] Anon., BP Alaska: Unmanned aerial vehicle (UAV) pilot t est ing, 2012.
[239] J. Allen, B. W alsh, Enhanced oil spill surveillance, det ection and monit oring
t hrough t he applied technology of unmanned air syst ems, in: Int . Oil Spill
Conf., No. 1, 2008, pp. 113120.
[240] J. Rao, Z. Gong, J. Luo, S. Xie, Unmanned airships for em ergency management,
in: Safet y, Securit y and Rescue Robot ics, Workshop, 2005 IEEE Int ernat ional,
2005, pp. 125130.
[241] E. Donnay, Use of unmanned aerial vehicle (UAV) for t he detect ion and
surveillance of marine oil spills in t he belgian part of t he nort h sea, Vol. 2,
2009.
[242] M . Fingas, Oil Spill Science and Technology, Gulf Professional Publishing,
Elsevier Inc., 2010.
[243] T.-M . Wang, X.-S. Lei, J.-H. Linag, B.-Q. Pei, A small unmanned aerial vehicle
for oil-gas field surveillance, in: Machine Learning and Cybernetics, 2008
Int ernational Conference on, vol. 4, 2008, pp. 18401846.
[244] J. Lin, H. Tao, Y. Wang, Z. Huang, Practical applicat ion of unmanned aerial
vehicles for m ountain hazards survey, 2010, pp. 15.
[245] D. W ang, J. Xu, R. Yao, R. M iao, Simulation system of telemet ering and
t elecont rol for unmanned aerial vehicle, IEEE Aerosp. Elect ron. Syst . M ag. 21
(9) (2006) 35.
[246] S. Tsach, A. Tatievsky, L. London, Unmanned Aerial Vehicles (UAVs), John
W iley Sons, Lt d., 2010.
[247] W . Khan, M . Nahon, Tow ard an accurate physics-based UAV thrust er model,
IEEE/ASM E Trans. M echat ronics 18 (4) (2013) 12691279.
[248] Y. Kang, J. Hedrick, Linear t racking for a fixed-w ing UAV using nonlinear
m odel predictive control, IEEE Trans. Cont rol Syst. Technol. 17 (5) (2009)
12021210.
[249] Y. Luo, H. Chao, L. Di, Y. Chen, Lat eral direct ional fractional order (PI); cont rol
of a small fixed-w ing unmanned aerial vehicles: cont roller designs and flight
t ests, IET Cont rol Theory Appl. 5 (18) (2011) 21562167.
[250] H. de Marina, F. Pereda, J. Giron-Sierra, F. Espinosa, UAV at t it ude est imat ion
using unscent ed kalman filt er and TRIAD, IEEE Trans. Ind. Electron. 59 (11)
(2012) 44654474.

506

A. Shukla, H. Karki / Robotics and Autonomous Systems 75 (2016) 490507

[251] H. Kim, M. Kim, H. Lim, C. Park, S. Yoon, D. Lee, H. Choi, G. Oh, J. Park, Y. Kim,
Fully aut onomous vision-based net -recovery landing system for a fixed-w ing
UAV, IEEE/ASM E Trans. M echatronics 18 (4) (2013) 13201333.
[252] P. Sujit , S. Saripalli, J.B. Sousa, Unm anned aerial vehicle pat h follow ing: A
survey and analysis of algorithms for fixed-w ing unmanned aerial vehicless,
IEEE Control Syst. 34 (1) (2014) 4259.
[253] H. Oh, D. Turchi, S. Kim , A. Tsourdos, L. Pollini, B. W hit e, Coordinat ed st andoff
t racking using pat h shaping for mult iple UAVs, IEEE Trans. Aerosp. Electron.
Syst . 50 (1) (2014) 348363.
[254] F. W arsi, D. Hazry, S. Ahmed, M . Joyo, M . Tanveer, H. Kam arudin, Z. Razlan,
Yaw , Pitch and Roll cont roller design for fixed-w ing UAV under uncert ainty
and pert urbed condition, in: Signal Processing its Applicat ions, CSPA, 2014
IEEE 10t h Int ernational Colloquium on, 2014, pp. 151156.
[255] G. Saggiani, B. Teodorani, Rot ary w ing UAV pot ent ial applications: an
analytical study t hrough a matrix met hod, Aircr. Eng. Aerosp. Technol. 76 (1)
(2004) 614.
[256] L. M urat et , S. Doncieux, J.-A. Meyer, A biomimet ic react ive navigation
system using t he opt ical flow for a rot ary-w ing UAV in urban environm ent ,
in: Proceedings of t he Int ernational Session on Robot ics.
[257] S.D. Hanford, L.N. Long, J.F. Horn, A small sem i-autonomous rot ary-w ing
unmanned air vehicle (UAV), AIAA2005-7077.
[258] B. Ahm ed, H. Pota, M . Garrat t , Rot ary w ing UAV posit ion control using backstepping, in: IEEE Conference on Decision and Cont rol, 2007, pp. 19571962.
[259] B. Ahmed, H.R. Pot a, M . Garratt , Flight cont rol of a rot ary w ing UAVa
practical approach, in: 47t h IEEE Conference on Decision and Cont rol, 2008.
CDC 2008, IEEE, 2008, pp. 50425047.
[260] O. Meister, N. Friet sch, C. Ascher, G. Trom mer, Adapt ive pat h planning for
VTOL-UAVs, IEEE Aerosp. Electron. Syst. M ag. 24 (7) (2009) 3641.
[261] M .-D. Hua, T. Hamel, P. M orin, C. Samson, A control approach for thrust propelled underactuat ed vehicles and its applicat ion t o vt ol drones, IEEE
Trans. Automat. Cont rol 54 (8) (2009) 18371853.
[262] B. Ahmed, H.R. Pot a, M . Garrat t , Flight cont rol of a rot ary w ing UAV using
backst epping, Internat . J. Robust Nonlinear Control 20 (6) (2010) 639658.
[263] B. Ahm ed, H.R. Pot a, Dynamic compensat ion for control of a rot ary w ing UAV
using posit ive posit ion feedback, J. Int ell. Robot . Syst . 61 (14) (2011) 4356.
[264] M . Ilarslan, M. Bayrakceken, A. Arisoy, Avionics syst em design of a mini VTOL
UAV, IEEE Aerosp. Electron. Syst. M ag. 26 (10) (2011) 3540.
[265] A. Roberts, A. Tayebi, Adapt ive posit ion t racking of VTOL UAVs, IEEE Trans.
Robot. 27 (1) (2011) 129142.
[266] B. Herisse, T. Hamel, R. M ahony, F.X. Russott o, Landing a VTOL unmanned
aerial vehicle on a moving platform using optical flow , IEEE Trans. Robot . 28
(1) (2012) 7789.
[267] P. Brow n, Feasibilit y st udy of a ship-borne rotary UAV recovery syst em,
School of Aerospace, Civil, and Mechanical Engineering, Universit y College,
ADFA, UNSW .
[268] A. Brow n, The development of a secondary lock dow n syst em for a ship
borne rot ary UAV landing plat form, Init ial thesis report , UNSW @ADFA.
[269] M .K. Samal, S. Anavat ti, M . Garrat t, Real-time neural net w ork based
ident ificat ion of a rot ary-w ing UAV dynamics for aut onomous flight , in: IEEE
Internat ional Conference on Industrial Technology, 2009. ICIT 2009, IEEE,
2009, pp. 16.
[270] K. M ansfield, T. Eveleigh, T. Holzer, S. Sarkani, Unmanned aerial vehicle
smart device ground cont rol st at ion cyber security t hreat m odel, in: Tech.
for Homeland Security (HST), IEEE, 2013, pp. 722728.
[271] Y. Kang, M . Yuan, Soft w are design for mini-t ype ground control station of
UAV, in: nt ernational Conference on Electronic M easurement Inst ruments,
2009, pp. 4-7374-740.
[272] B. Kayayurt, I. Yayla, A. Yapici, C. Kucukoguz, Ground cont rol stat ion avionics
soft w are development in anka UAV, in: Digital Avionics System s Conference,
DASC, 2011 IEEE/AIAA 30th, 2011, pp. 5B6-15B6-7.
[273] R.H. Klenke, J. M cBride, H. Klenke, A reconfigurable, linux-based, flight
cont rol system for small UAVs, in: AIAA Infot ech Conference and Exhibit on
Aerospace, 2007, pp. 110.
[274] E.M . At kins, R.D. Eubank, A.T. Atkins, A reconfigurable, linux-based, flight
cont rol syst em for small UAVs, in: AIAA Infot ech Conf. and Exhibit on
Aerospace, 2009, pp. 110.
[275] A. Nemra, N. Aouf, Robust INS/GPS sensor fusion for UAV localizat ion using
sdre nonlinear filt ering, IEEE Sens. J. 10 (4) (2010) 789798.
[276] J. Wendel, O. M eist er, C. Schlaile, G.F. Tromm er, An int egrat ed gps/mems-im u
navigat ion system for an aut onomous helicopter, Aerosp. Sci. Technol. 10 (6)
(2006) 527533.
[277] S. Saripalli, G.S. Sukhatm e, J.F. Montgomery, An experiment al study of the
autonomous helicopt er landing problem, in: Experimental Robot ics VIII,
Springer, 2003, pp. 466475.
[278] J. Zhang, H. Yuan, Analysis of unmanned aerial vehicle navigat ion and height
cont rol syst em based on gps, J. Syst . Eng. Electron. 21 (4) (2010) 643649.
[279] S.D.L. Parra, J. Angel, Low cost navigation syst em for UAV, Aerosp. Sci.
Technol. 9 (2005) 504516.
[280] R. Johnson, J. Sasiadek, J. Zalew ski, Kalman filter enhancement for UAV
navigat ion, Simul. Ser. 35 (1) (2003) 267272.
[281] N. Sunderhauf, S. Lange, P. Protzel, Using t he unscent ed kalman filt er in
mono-SLAM w it h inverse depth paramet rizat ion for autonomous airship
cont rol, in: IEEE Int ernat ional Workshop on Safet y, Securit y and Rescue
Robotics, 2007. SSRR 2007, IEEE, 2007, pp. 16.
[282] H.G. de Marina, F.J. Pereda, J.M. Giron-Sierra, F. Espinosa, UAV at tit ude
estimation using unscented kalman filter and triad, IEEE Trans. Ind. Electron.
59 (11) (2012) 44654474.

[283] C.G. Prevost , A. Desbiens, E. Gagnon, Ext ended kalman filt er for st ate
est im at ion and traject ory predict ion of a moving object det ect ed by an
unmanned aerial vehicle, in: Am erican Cont rol Conference, 2007. ACC07,
IEEE, 2007, pp. 18051810.
[284] S. Ronnback, Developm ent of an INS/GPS navigat ion loop for an UAV
(M ast ers thesis), Lulea Tekniska Universit et , Lulea, Sw eden, 2000.
[285] A.G. Kallapur, S.G. Anavat ti, UAV linear and nonlinear estimation using
ext ended kalman filt er, in: CIM CA/IAWTIC, 2006, p. 250.
[286] G. Chow dhary, R. Jat egaonkar, Aerodynamic paramet er estimat ion from
flight dat a applying ext ended and unscent ed kalman filt er, Aerosp. Sci.
Technol. 14 (2) (2010) 106117.
[287] E. Frazzoli, M .A. Dahleh, E. Feron, Trajectory tracking control design for
aut onomous helicopt ers using a backst epping algorit hm, in: American
Control Conference, 2000. Proceedings of t he 2000, vol. 6, IEEE, 2000,
pp. 41024107.
[288] T. Lee, Y. Kim, Nonlinear adapt ive flight cont rol using backst epping and
neural netw orks cont roller, J. Guid. Cont rol Dyn. 24 (4) (2001) 675682.
[289] T. Espinoza, A. Dzul, R. Lozano, P. Parada, Backst epping-sliding m ode cont rollers applied to a fixed-w ing UAV, in: Unmanned Aircraft System s, ICUAS,
2013, pp. 95104.
[290] B. Ahmed, H.R. Pot a, Flight control of a rot ary w ing UAV using adapt ive
backst epping, in: IEEE Int ernat ional Conference on Cont rol and Aut om at ion,
2009. ICCA 2009, IEEE, 2009, pp. 17801785.
[291] H.-S. Ju, C.-C. Tsai, S.-W. Liu, Design of longitudinal axis full envelope cont rol
law by adapt ive backstepping, in: Netw orking, Sensing and Cont rol, 2004
IEEE Internat ional Conference on, vol. 2, IEEE, 2004, pp. 755760.
[292] J.S. Jang, D. Liccardo, Sm all UAV aut omat ion using M EMS, IEEE Aerosp.
Electron. Syst. M ag. 22 (5) (2007) 3034.
[293] X. Fu, Z. Zhou, W . Xiong, Q. Guo, Mems-based low -cost flight cont rol syst em
for sm all UAVs, Tsinghua Sci. Technol. 13 (5) (2008) 614618.
[294] G. Conte, P. Dohert y, An int egrat ed UAV navigat ion syst em based on aerial
image matching, in: Aerospace Conference, 2008 IEEE, IEEE, 2008, pp. 110.
[295] E.E.K. Filt er, Vision-aided navigation for small UAVs in GPS-challenged
environment s.
[296] J. Hermansson, A. Gising, M. Skoglund, T. Schn, Autonomous landing of an
unmanned aerial vehicle.
[297] C. M iao, J. Li, Autonomous landing of small unmanned aerial rot orcraft based
on m onocular vision in GPS-denied area, IEEE/CAA J. Automat. Sinica 2 (1)
(2015) 109114.
[298] Y. Jin, Y. Liao, A. M inai, M. Polycarpou, Balancing search and target response
in cooperative unmanned aerial vehicle (UAV) t eam s, IEEE Trans. Syst . Man
Cybern. 36 (3) (2005) 571587.
[299] A. Sinha, A. Tsourdos, B. Whit e, Mult i UAV coordinat ion for t racking t he
dispersion of a contaminant cloud in an urban region, Eur. J. Control 15 (34)
(2009) 441448.
[300] M . Tortonesi, C. Stefanelli, E. Benvegnu, K. Ford, N. Suri, M . Linderman,
M ult iple-UAV coordinat ion and communications in t actical edge netw orks,
IEEE Commun. Mag. 50 (10) (2012) 4855.
[301] W . Xinzeng, C. Linlin, L. Junshan, Y. Ning, Based on pso algorithm m ultiple task
assignment s for cooperating UAVs, in: Educat ional and Information Technology, ICEIT, 2010 Internat ional Conference on, vol. 2, 2010, pp. V2-25V2-28.
[302] A. Gil, K. Passino, S. Ganapat hy, A. Sparks, Cooperat ive task scheduling for
net w orked uninhabited air vehicles, IEEE Trans. Aerosp. Elect ron. Syst. 44 (2)
(2008) 561581.
[303] S.K. Hong, Fuzzy logic based closed-loop st rapdow n att it ude syst em for
unmanned aerial vehicle (UAV), Sensors Act uators A 107 (2) (2003) 109118.
[304] M . Hardesty, S. Kennedy, S. Dixon, T. Berka, J. Graham, D. Caldw ell, Development of navigation and aut omat ed flight cont rol syst em solut ions for
m aritime VTOL UAS operations, Association for Unmanned Vehicle System s
Int ernationals Unmanned Syst ems North America.
[305] R.T. W ainner, M.B. Frish, B.D. Green, M.C. Laderer, M .G. Allen, J.R. M orency,
High alt it ude aerial natural gas leak det ect ion syst em, Final Report Prepared for US Depart ment of Energy under Grant No. DE-FC26-04NT42268,
PSI-1454/TR-2211.
[306] J.-H. Kim, S. Sukkarieh, Flight t est result s of gps/ins navigat ion loop for an
aut onomous unm anned aerial vehicle (UAV), in: Proceedings of t he 15t h
Int ernational Technical M eeting of t he Sat ellit e Division of The Inst it ut e of
Navigat ion, ION GPS 2002, 2001, pp. 510517.
[307] C.-S. Yoo, I.-K. Ahn, Low cost GPS/INS sensor fusion syst em for UAV
navigation, in: Digit al Avionics Syst ems Conference, 2003. DASC03. The
22nd, vol. 2, IEEE, 2003, 8A.
[308] J.-H. Kim, S. Sukkarieh, S. W ishart, Real-t im e navigation, guidance, and
control of a UAV using low -cost sensors, in: Field and Service Robot ics,
Springer, 2006, pp. 299309.
[309] P. Doherty, G. Granlund, K. Kuchcinski, E. Sandew all, K. Nordberg, E. Skarm an, J. W iklund, The w itas unmanned aerial vehicle project , in: ECAI, 2000,
pp. 747755.
[310] A. Ryan, M . Zennaro, A. How ell, R. Sengupt a, J.K. Hedrick, An overview of
emerging result s in cooperative UAV cont rol, in: Decision and Cont rol, 2004.
CDC. 43rd IEEE Conference on, vol. 1, IEEE, 2004, pp. 602607.
[311] E. Past or, J. Lopez, P. Royo, UAV payload and mission control hardw are/soft w are archit ect ure, IEEE Aerosp. Electron. Syst . Mag. 22 (6) (2007)
38.
[312] T.X. Brow n, B. Argrow , C. Dixon, S. Doshi, R.-G. Thekkekunnel, D. Henkel,
Ad hoc UAV ground net w ork (augnet), in: AIAA 3rd Unmanned Unlimited
Technical Conference, 2004, pp. 111.

A. Shukla, H. Karki / Robotics and Autonomous Systems 75 (2016) 490507

[313] J. Allred, A.B. Hasan, S. Panichsakul, W . Pisano, P. Gray, J. Huang, R. Han, D.
Law rence, K. Mohseni, Sensorflock: an airborne w ireless sensor netw ork of
micro-air vehicles, in: Proceedings of the 5t h Int ernational Conference on
Embedded Netw orked Sensor Systems, ACM , 2007, pp. 117129.
[314] E. Yanm az, R. Kuschnig, C. Bet tst et ter, Channel measurements over 802.11 abased UAV-t o-ground links, in: GLOBECOM W orkshops (GC W kshps), 2011
IEEE, IEEE, 2011, pp. 12801284.
[315] M. Bart holmai, P. Neumann, Micro-drone for gas measurement in hazardous
scenarios via remote sensing, in: Proceedings of, 2010.
[316] P. Rudol, P. Dohert y, Human body det ection and geolocalizat ion for UAV
search and rescue missions using color and therm al imagery, in: Aerospace
Conference, 2008 IEEE, IEEE, 2008, pp. 18.
[317] J. Berni, P.J. Zarco-Tejada, L. Surez, E. Fereres, Thermal and narrow band
multispect ral remot e sensing for veget ation monitoring from an unmanned
aerial vehicle, IEEE Trans. Geosci. Remot e Sens. 47 (3) (2009) 722738.
[318] J. Everaerts, N. Lew yckyj, D. Fransaer, Pegasus: design of a st rat ospheric long
endurance UAV syst em for remot e sensing, in: The Internat ional Archives of
the Photogrammetry, Remote Sensing and Spat ial Inform at ion Sciences 35,
Part B.
[319] A.S. Lom ax, W . Corso, J.F. Et ro, Employing unmanned aerial vehicles (UAVs)
as an element of t he integrated ocean observing syst em, in: OCEANS, 2005.
Proceedings of MTS/IEEE, IEEE, 2005, pp. 184190.
[320] E. Donnay, Use of unm anned aerial vehicle (UAV) for the det ect ion and
surveillance of marine oil spills in t he belgian part of the nort h sea.
[321] F. M utt in, Umbilical deployment modeling for t et hered UAV det ect ing oil
pollut ion from ship, Appl. Ocean Res. 33 (4) (2011) 332343.
[322] W.J. Lehr, The potential use of small uas in spill response, in: Internat ional
Oil Spill Conference, Vol. 2008, American Pet roleum Inst it ut e, 2008,
pp. 431433.
[323] W. Zang, J. Lin, Y. W ang, H. Tao, Investigating small-scale w ater pollution w ith
UAV remote sensing t echnology, in: World Aut om at ion Congress (WAC),
2012, IEEE, 2012, pp. 14.
[324] N. M oham ed, I. Jaw har, J. Al-Jaroodi, L. Zhang, M onit oring underw at er
pipelines using sensor netw orks, in: High Performance Comput ing and
Com mun., HPCC, 2010, pp. 346353.
[325] M.S. BenSaleh, S.M . Qasim , A.M. Obeid, A. Garcia-Ortiz, A review on w ireless
sensor net w ork for w ater pipeline monit oring applicat ions, in: Collaboration
Technologies and Syst ems, CTS, 2013 Int ernational Conference on, 2013,
pp. 128131.
[326] N. M ohamed, I. Jaw har, A fault tolerant w ired/w ireless sensor netw ork
architecture for monitoring pipeline infrastructures, in: Sensor Technologies
and App., 2008, pp. 179184.
[327] S. Petersen, P. Doyle, S. Vatland, C.S. Aasland, T.M . Andersen, D. Sjong,
Requirement s, drivers and analysis of w ireless sensor net w ork solut ions
for t he oil gas industry, in: Emerging Technologies and Fact ory Automation,
2007. ETFA. IEEE Conference on, 2007, pp. 219226.
[328] M. Akhondi, A. Talevski, S. Carlsen, S. Pet ersen, Applicat ions of w ireless sensor
net w orks in the oil, gas and resources indust ries, in: Advanced Information
Netw orking and Applicat ions (AINA), IEEE, 2010, pp. 941948.
[329] A.O. Adejo, S.J. Onumanyi, J.M. Anyanya, S.O. Oyew obi, Oil and gas process
monit oring t hrough w ireless sensors net w orks: A survey, Ozean J. Appl. Sci.
6 (2) (2013) 10771079.
[330] H. Yu, M . Guo, An efficient oil and gas pipeline m onit oring syst ems based on
w ireless sensor net w orks, in: Information Securit y and Intelligence Control,
ISIC, 2012 Internat ional Conference on, 2012, pp. 178181.
[331] K. Anupam a, N. Kamdar, S. Kam alam pet, D. Vyas, S. Sahu, S. Shah, A w ireless sensor netw ork based pipeline monitoring syst em, in: Internat ional
Conference on Signal Processing and Integrated Netw orks, SPIN, 2014,
pp. 412419.
[332] J. Sheldon, R. Kube, H. Zhang, Oil sand screen modelling using part ial least
squares regression, in: IEEE Conf. on Aut om at ion and Logist ics, 2008, pp.
29362940.
[333] A. Abdelgaw ad, M. Bayoumi, Remote measuring for sand in pipelines
using w ireless sensor netw ork, IEEE Trans. Instrum . M eas. 60 (4) (2011)
14431452.
[334] H. Zhang, F. Zhang, Y. W ang, G. Zhang, W ireless sensor netw ork based antitheft system of m onit oring on pet roleum pipeline, in: M echanic Automation
and Control Engineering, M ACE, 2011, pp. 46894692.
[335] I. Jaw har, N. M ohamed, K. Shuaib, A framew ork for pipeline infrast ruct ure
monit oring using w ireless sensor netw orks, in: W ireless Telecommunications Symposium, 2007, pp. 17.

507

[336] D. W u, D. Chatzigeorgiou, K. Youcef-Toumi, S. M ekid, R. Ben-Mansour,

Channel-aw are relay node placement in w ireless sensor net w orks for
pipeline inspect ion, IEEE Trans. Wireless Comm un. PP (99) (2014) 11.
[337] I. Jaw har, N. M ohamed, J. Al-Jaroodi, S. Zhang, An efficient framew ork for
aut onomous underw at er vehicle ext ended sensor net w orks for pipeline
m onit oring, in: Robotic and Sensors Environments (ROSE), IEEE, 2013,
pp. 124129.
[338] M . Rahman, H. Hasbullah, Early detect ion met hod of corrosion on buried
st eel gas pipeline using w ireless sensor net w ork, in: Int. Conf. on Com puter
and Aut omation Engineering, ICCAE, vol. 3, 2010, pp. 553556.
[339] Z. W ang, X. Zhao, X. Qian, The application and issuse of linear w ireless
sensor net w orks, in: System Science, Engineering Design and M anufacturing
Inform at ization, ICSEM, 2011 Int ernational Conference on, vol. 2, 2011,
pp. 912.
[340] R. Liu, I. Wassell, K. Soga, Relay node placement for w ireless sensor netw orks
deployed in t unnels, in: IEEE 6t h Internat ional Conference on W ireless
and M obile Computing, Net w orking and Com municat ions, W iM ob, 2010,
pp. 144150.
[341] I. Jaw har, N. M ohamed, J. Al-Jaroodi, S. Zhang, Dat a communication in
linear w ireless sensor netw orks using unmanned aerial vehicles, in: Unm anned Aircraft Syst ems, ICUAS, 2013 Int ernational Conference on, 2013,
pp. 492499.
[342] D. Misiunas, J. Vit kovsky, G. Olsson, A. Sim pson, M. Lam bert , Pipeline
break det ection using pressure t ransient monitoring, J. W ater Resour. Plann.
M anag. 131 (4) (2005) 316325.
[343] H. M ustafa, P. Chou, Embedded damage detection in w at er pipelines using
w ireless sensor net w orks, in: High Performance Computing and Communication Int. Conf. on Em bedded Softw are and Sys., HPCC-ICESS, 2012, pp.
15781586.
[344] Y. Gao, M. Brennan, P. Joseph, J. Mugglet on, O. Hunaidi, On the selection of
acoustic/vibration sensors for leak det ection in plastic w at er pipes, J. Sound
Vib. 283 (35) (2005) 927941.
[345] I. St oianov, L. Nachm an, S. Madden, T. Tokm ouline, M. Csail, Pipenet: A
w ireless sensor net w ork for pipeline monit oring, in: Informat ion Processing
in Sensor Net w orks, 2007. IPSN 2007. 6t h Int ernational Symposium on,
2007, pp. 264273.
[346] M . M eribout, A w ireless sensor net w ork-based infrastructure for real-time
and online pipeline inspect ion, IEEE Sens. J. 11 (11) (2011) 29662972.
[347] Y. Khulief, A. Khalifa, R. Mansour, M. Habib, Acoust ic det ect ion of leaks in
w at er pipelines using measurements inside pipe, J. Pipeline Syst . Eng. Pract.
3 (2) (2012) 4754.
[348] A.M. Sadeghioon, N. M etje, D.N. Chapm an, C.J. Ant hony, Sm art Pipes: smart
w ireless sensor netw orks for leak detection in w at er pipelines, J. Sensor
Act uator Net w . 3 (1) (2014) 6478.

Dr . Am i t Shuk la, born in India in 1983, has done his

Ph.D. from Im perial College London, Unit ed Kingdom , w it h
control syst ems group in 2012. At Im perial College London,
his Ph.D. w ork w as primarily focused on the Modeling,
Simulation and Cont rol of Hybrid elect ric Vehicles. He did
his B.Tech and M.tech from Indian Inst it ut e of Technology
(IIT) Kanpur, India, w ith specializat ion in t he robot ics and
control systems. His research experience expands across
various dom ains of engineering w it h some of t he key areas
of int erest s are Robotics, M echat ronics, Cont rol System s,
Pow er Electronics, Electrical Drives, Renew able Energy
Syst em and Hybrid Elect ric Vehicles. Current ly he is w orking The Pet roleum
Institute, Abu Dhabi, UAE, for development of advance robotic t echnologies t o be
used in oil and gas indust ry.
Dr . Ham ad Kar ak i , has done his Ph.D. from I Tokyo
Universit y of Technology, Japan w ith Robotics Group.
At Tokyo Universit y of Technology, his Ph.D. w ork w as
primarily focused on field robot s w hich are basically
t racked vehicles t hat have various off-road running
features and applicable in rescue missions. His research
experience expands across various domains of engineering
w it h some of the key areas of interests are Robot ics,
M echatronics and Control System s.