Available online at www.sciencedirect.

com

ScienceDirect
Procedia Engineering 196 (2017) 278 – 285

Creative Construction Conference 2017, CCC 2017, 19-22 June 2017, Primosten, Croatia

An Integrated Multi-Criteria-Decision-Making Tool for Mechanical

Designer Selection
Gul Polata, Befrin Neval Bingolb*, Ozgen Varc
a
Associate Professor, Istanbul Technical University, Istanbul 34469, Turkey
b
Research Assistant , Istanbul Technical University, Istanbul 34469, Turkey
c
Graduate Student , Istanbul Technical University, Istanbul 34469, Turkey

Abstract

In construction projects, many parties, e.g., owner, design team, general contractor, subcontractors, consultants, suppliers, etc.,
take part and collaborate. The qualifications of those parties directly affect the success of a construction project. Therefore, those
parties should be selected in an objective and systematic way to complete the project successfully. The mechanical design group
is an important participant of a design team, whose expertise area is very complicated. The design process of the mechanical
group requires multi-disciplinary effort. The technical capability, experience, and communication skills of these parties play a
critical role. However, in practice, the selection of the mechanical design team is usually made based on the lowest bid criterion,
which may result in additional costs due to working with unqualified, incapable, and insufficient companies. The main objective
of this study is to propose an integrated approach, which combines two multi-decision criteria decision-making techniques, to
select the most appropriate mechanical design team in construction projects. In the proposed approach, Analytical Hierarchy
Process (AHP) and Complex Proportional Assessment (COPRAS) methods are used together and various conflicting and
compromising criteria are considered simultaneously. In this study, the AHP technique is used to calculate the weights of the
criteria and the COPRAS method is used to rank the design team alternatives. In order to demonstrate how this approach can be
applied to a real life problem, a case study is presented.

© 2017
© 2017TheTheAuthors. Published
Authors. by Elsevier
Published Ltd. This
by Elsevier Ltd. is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the scientific committee of the Creative Construction Conference 2017.
Peer-review under responsibility of the scientific committee of the Creative Construction Conference 2017
Keywords: AHP, COPRAS, mechanical designer, multi-decision-making, integrated approach.

* Corresponding author. Tel.: +90-212-285-3737; fax: +90-212-285-6587.

E-mail address: bbingol@itu.edu.tr

1877-7058 © 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the scientific committee of the Creative Construction Conference 2017
doi:10.1016/j.proeng.2017.07.200
 Gul Polat et al. / Procedia Engineering 196 (2017) 278 – 285 279

1. Introduction

In construction projects, two key factors play a critical role in the success, which are (1) the quality of design
documents and (2) the quality of materials and workmanship. High quality design reduces constructability problems
and clashes. Therefore, general contractors should thoroughly select the design team in order to overcome potential
delays and cost overruns. However, this selection is not an easy task as it requires long evaluation process. The
mechanical design systems are the most challenging and least known phase of the construction design process. The
mechanical design process requires multi-disciplinary effort including the deployment of equipment and distribution
systems. In most of the construction projects, the coordination of the mechanical systems is very complicated. The
design process of the mechanical systems is one of the most complex aspects of construction projects [1]. The proper
design process of mechanical systems requires effective communication between all parties of the construction
project, such as architectural and structural design teams and other special trades [1]. The mechanical systems can be
very complicated depending on the complexity of the required system and the high level of usage. There are several
different usage areas of mechanical systems in construction projects and clashes among these systems should be
avoided in the design phase. Otherwise, constructability problems can be encountered during the construction stage,
which in turn may cause delays, cost overruns and reworks [1]. Generally, the bid price is a key criterion in the
selection process of a design team. However, the design process of the mechanical systems necessitates proper
technical capability, enough experience, and good communication skills in a design company. Therefore, the
technical capability, experience, and communication skills of the mechanical design company will affect the quality
of the design served by this company. The evaluation process of a mechanical design team should be made based on
multiple criteria. Decision support systems are very helpful and practical tools for selecting the most appropriate
company based on knowledge and expertise of experts with saving valuable time [2]. The main objective of this
study is to propose an integrated approach to select the most appropriate mechanical design team in construction
projects. In the integrated approach, Analytical Hierarchy Process (AHP) and Complex Proportional Assessment
(COPRAS) methods are combined to determine the best mechanical design team for a construction project. In the
proposed model, the AHP technique is used to determine the weights of the criteria and the COPRAS method is used
to rank the alternative mechanical design companies. The proposed approach is applied in a real life project.

2. The Proposed Approach

The proposed integrated approach to select the most appropriate mechanical design team in construction projects
consists of eight steps, which can be categorized in two phases. The steps of the proposed approach are presented in
Figure 1.
Phase-I Phase-II

Forming decision matrix of

Defining the mechanical
mechanical designer selection
designer selection model
problem
1 5

Developing the weighted

Developing hierarchy of the
normalized decision matrix of
mechanical designer problem
the selection problem
2 6

Applying COPRAS method to

Constructing pairwise
find preferences of the
comparison matrices by group
mechanical designer
decision making
3 alternatives 7

Determining weights of the

Ranking the mechanical
main criteria in the hierarchy
designer alternatives
using AHP
4 8

Fig. 1. The steps of the proposed integrated approach.

280 Gul Polat et al. / Procedia Engineering 196 (2017) 278 – 285

In the first phase, the mechanical designer selection problem is defined. Then, the decision making team, who are
in charge of mechanical designer selection in the company, is formed, and this decision-making team identifies the
main and sub-criteria that may affect the mechanical designer selection and develops the hierarchy of the mechanical
designer selection model. Thirdly, the decision making team constructs pairwise comparison matrices of the
mechanical designer selection problem. In the final step of the first phase, the weights of the main and sub-criteria of
the mechanical designer selection problem are calculated using the AHP method. In the second phase, the decision
matrix, which includes the evaluations of the decision-making team members of the alternative mechanical designer
companies according to the main and sub-criteria, for the mechanical designer selection problem is firstly formed. In
the second step, the weighted normalized matrix is calculated. Then, the preference scores of the mechanical design
team alternatives are found using the COPRAS method. Finally, the mechanical designer alternatives are ranked
based on their preference scores.

2.1. Analytical Hierarchy Process (AHP)

The AHP method was first introduced by Saaty in 1980 to solve complex decision-making problems. This
method is based on evaluations of the alternatives in terms of relative importance according to the criteria set by
decision makers to solve decision-making problems [3]. The AHP method allows using combination of qualitative
and quantitative criteria at the same time. Also, its methodology is easy to imply. Depending upon these features, the
AHP is the most preferred multi-criteria decision-making method among the other alternatives. In this method, the
pairwise comparison ratios are combined in a ratio scale [4]. Basically, the AHP consists of four main steps, which
are (1) defining the decision-making problem with all related information, (2) forming the hierarchy of the decision-
making problem in a broad perspective (i.e., defining goal, setting criteria, sub-criteria and alternatives) , (3)
constructing pairwise comparison matrices for the entire hierarchy using Saaty’s Rating Scale (see Table 1), and (4)
determining the priorities of the alternatives according to the criteria and/or sub-criteria implying mathematical
computations of the method [5].

Table 1. Saaty’s Rating Scale.

Intensity of
Definition
Importance
1 Equal importance
3 Somewhat more important
5 Much more important
7 Very much important
9 Absolutely more important
2-4-6-8 Intermediate values
Explanation
Two factors contribute equally to the objective.
Experience and judgment slightly favor one over the other.
Experience and judgment strongly favor one over the other.
Experience and judgment very strongly favor one over the other.
The evidence favoring one over the other is one of the highest possible validity.
When compromise is needed.

In the AHP method, the pairwise comparison matrix is composed in accordance with the number of criteria. In
this matrix, the diagonal elements are directly set to the “1” due to the comparison of the criterion with itself. Also,
the elements in the decision matrix are reciprocal according in the direction of the diagonal. Thus, the decision
maker needs to only evaluate the upper or lower triangle of the decision matrix. The decision matrix formed based
on the given information is illustrated in Equation 1.

ª a11 a12  a1n º

« 1 a 22  a2n »
« a12 »
A « (1)
    »
« »
« 1 1  a nm »
¬ a n1 an 2 ¼

The pairwise comparison matrix is mainly used to evaluate the priority vector. The priority vector calculations
start with the normalization procedure, which is performed by dividing each element to the sum of its column. After
this process, the priority vector is obtained by taking the arithmetic mean of the each row in the normalized matrix.
 Gul Polat et al. / Procedia Engineering 196 (2017) 278 – 285 281

These priorities present the weights of the decision alternatives in the hierarchy of the problem and are used to select
the best option. In the AHP method, the consistency of the pairwise comparison matrix should also be checked. For
this evaluation, it is necessary to calculate consistency index and then consistency ratio. The consistency index (CI)
is determined with the aid of Equation 2.

(O max  n )
CI (2)
( n  1)

where λmax is the maximum Eigenvalue, n is the number of criteria/alternatives of the decision problem. The
consistency ratio (CR) is computed via Equation 3.

CI
CR (3)
RI

In the Equation 3, RI refers to the random index value, and Table 2 presents some random index values for
different number of items. The consistency ratio (CR) of the pairwise comparison matrix should be smaller than 0.1
(i.e., CR<0.1) [6]. If it is not, the pairwise comparison matrix should be reformed by the decision maker(s).

Table 2. Random Index.

Number of Items Random Index
1 0
2 0
3 0.58
4 0.90
5 1.12
Number of Items Random Index
6 1.24
7 1.32
8 1.41
9 1.45
10 1.49

2.2. Complex Proportional Assessment (COPRAS)

In multi-criteria decision-making problems, the decision maker should select the best alternative among the
several alternatives by taking into consideration the defined attributes for the problem in question. Therefore, several
multi-criteria decision-making methods have been proposed by researchers to make a sound decision in this broad
area. The Complex Proportional Assessment (COPRAS) is one of the multi-criteria decision-making methods which
was first developed in 1994 and then improved in 1996 [7]. This method uses the stepwise ranking, and alternatives
are evaluated based on their significance and utility degree [8]. COPRAS has been applied to the various kinds of
decision problems because of its simplicity [9]. The steps of COPRAS method are briefly explained as follows:
Step-1: Constructing the decision matrix X with n number of alternatives and m number of attributes, as shown in
Equation 4:

ª x11 x12  x1n º

«x x22  x2 n »
X « 21 » (4)
«     »
« »
¬ x n1 xn 2  xnm ¼

Step-2: Normalizing the elements of decision matrix by using Equation 5:

xij
xij ; i 1, n and j 1, m (5)
¦
n

xij
i 1
282 Gul Polat et al. / Procedia Engineering 196 (2017) 278 – 285

In Equation 5; j presents attribute and i represents an alternative. The normalized decision matrix is illustrated in
Equation 6:

ª x11 x12  x1n º

« »
« x21 x22  x2 n »
X (6)
«    »
« »
«¬ xn1 xn 2  xnm »¼

Step-3: In this step, the weighted normalized decision matrix is determined by using Equation 7:

xˆij xij . q j ; i 1, n and j 1, m (7)

where ‫ݍ‬௝ is the weight of jth attribute. The weighted normalized decision matrix is shown in Equation 8:

ª xˆ11 xˆ12  xˆ1n º

« xˆ xˆ 22  xˆ 2 n »
Xˆ « 21 » (8)
«    »
« »
¬ xˆ n1 xˆ n 2  xˆ nm ¼

Step-4: The sums of weighted normalized values are evaluated both for beneficial attributes (i.e., maximization is
preferable) (use Equation 9) and non-beneficial attributes (i.e., minimization is preferable) (use Equation 10).

¦ xˆ
k

Pj ij
(9)
i 1

¦ xˆ
m

Rj ij
(10)
i k 1

where k represents the total number of beneficial attributes which are going to maximized, and m represents the
total number of non-beneficial attributes which are going to minimized.
Step-5: In this step, the relative weight of each alternative which is denoted with Qj are determined via Equation
11.

¦R
n

Qj Pj  j 1
(11)
¦
n
1
Rj
j 1 Rj

Step-6: The optimality criterion K is evaluated in this step (Equation 12).

K max Q j ; j 1, n (12)
j

Step-7: In the final step, the utility degree of each alternative is calculated via Equation 13, where Qj is the
significance of the alternative obtained from Equation 11, and Qmax is the maximum relative significance of value.
The ranking of the alternatives can be ordered based on the utility degree calculated from Equation 13.
 Gul Polat et al. / Procedia Engineering 196 (2017) 278 – 285 283

Qj
Nj u 100% (13)
Qmax

3. Case Study: Mechanical Designer Selection Problem

In this study, the integrated mechanical designer selection model is tested with the real case, which is a housing
project in İstanbul, Turkey. This project consists of 28 blocks and total construction area is 260,000 m2. The
anticipated duration of the project is 22 months. The selection process of the mechanical designer team mainly
depends on the characteristics of the undertaken task. Therefore, the decision makers, who are responsible for the
mechanical designer selection process, should participate in developing the mechanical designer selection model.
For that reason, first the decision making team was formed and the model was developed based on their opinions
and evaluations.

3.1. Hierarchy of the Mechanical Designer Selection Problem

In the general contractor company, four experienced engineers were responsible for the selection process of
mechanical designer team. Thus, the face to face interviews were made with these authorized staff. These engineers
identified seven main criteria, which are: cost (C1), experience (C2), reference (C3), the largest project completed by
the company (C4), quality of service (C5), diversity of working area (C6), and quality of design (C7), 3 sub-criteria
under the main criterion C5, which are: quick response (SC51), rate of attendance to meetings (SC52), and qualified
workmanship (SC53), and 2 sub-criteria under the main criterion C7, which include: no clashes with other projects
(SC71) and clarity of the details in design documents (SC72). In the studied project, general contractor invited six
mechanical designer companies – which passed the pre-qualification phase- to bid for the project. These companies
are named A1 to A6 as alternatives of mechanical designer companies.

3.2. Determining the Weights of the Main and Sub Criteria of the Selection Problem

After constructing the hierarchy of the mechanical designer selection problem, the AHP method is used to
determine the weights of the identified main criteria and sub-criteria. For that reason, four decision makers were
asked individually to construct pairwise comparison matrices for main criteria and sub-criteria of the selection
problem. The four pairwise comparison matrices were then aggregated by taking the geometric means of each
preference in order to reach a group decision. Finally, the mathematical computations of the AHP method were
applied to find weights of the main and sub-criteria of the selection problem. Table 3 shows the weights of main
criteria of the mechanical designer selection problem.

Table 3. Aggregated pairwise matrix of main criteria for the mechanical designer selection problem.
Criteria C1 C2 C3 C4 C5 C6 C7 Weights
C1 1.000 1.861 1.861 3.568 2.913 5.244 2.632 0.300
C2 0.595 1.000 1.000 1.565 2.000 2.913 1.414 0.160
C3 0.595 1.000 1.000 1.861 2.060 3.201 1.316 0.170
C4 0.293 0.639 0.537 1.000 0.841 2.213 0.485 0.090
C5 0.369 0.500 0.485 1.189 1.000 2.515 0.639 0.100
C6 0.211 0.343 0.312 0.452 0.398 1.000 0.435 0.050
C7 0.407 0.707 0.760 2.060 1.565 2.300 1.000 0.130 C.R.= 0.017

According to the findings, the “C1-Cost” has the highest weight on the selection of a mechanical designer
problem. It is followed by the “C3-Reference” and the “C2- Experience” with second and third higher weights; hence
these factors have also significant importance in the selection process. On the other hand, the “C6-Diversity of
working Area” has the least importance in the selection process due to its lowest weight. The consistency ratio of the
aggregated pairwise comparison matrix is also checked, and it is below the 0.1.
284 Gul Polat et al. / Procedia Engineering 196 (2017) 278 – 285

The aggregated pairwise comparison matrix of three sub-criteria (i.e., SC51, SC52, and SC53) that identified under
the “C5-Quality of Service” was given in Table 4.

Table 4. Aggregated pairwise matrix of sub-criteria for the “C5-quality of service”.

Sub-Criteria SC51 SC52 SC53 Weights
SC51 1.000 1.861 0.639 0.334
SC52 0.537 1.000 0.485 0.202
SC53 1.565 2.060 1.000 0.464 C.R.= 0.0134

Based on the findings, the “SC53- Qualified Workmanship” has the highest weight among the sub-criteria. The
aggregated matrix is also consistent (C.R. = 0.0134<0.1).
The aggregated pairwise comparison matrix of sub-criteria under the “C7-Quality of Design” (i.e., SC71 and SC72)
was given in Table 5.

Table 5. Aggregated pairwise matrix of sub-criteria for the “C7-Quality of Design”.

Sub-Criteria SC71 SC72 Weights
SC71 1.000 0.604 0.377
SC72 1.655 1.000 0.623 C.R.= 0.000

The findings indicate that “SC72- Clarity of the Details in Design Documents” has the highest weight in this
group. The consistency ratio is also below the 0.1, so the aggregated matrix is consistent.

3.3. Finding the Preferences of the Mechanical Designer Alternatives with COPRAS Method

After determining the weights of the main criteria and sub-criteria of the mechanical designer selection problem,
COPRAS method is employed to determine the ranking of the mechanical designer alternatives. In the mechanical
designer selection problem, the preferences of the four decision makers were collected to form the decision matrix.
The data of C1, C2, C3, C4 and C6 are quantitative, whereas SC51, SC52, SC53, SC71 and SC73 are qualitative and these
values are obtained by using 1 to 9 point scale (i.e., 1: Very Bad; 9: Very Good. The collected data of the sub-
criteria under C5 and C7 were combined by taking their geometric means to construct one aggregated decision matrix
for mechanical designer selection problem (see Table 6). In this selection problem, only the C1 is the non-beneficial
criterion where the smaller value is always preferred. The rest (i.e., C2, C3, C4, C6 , SC51, SC52, SC53, SC71 and SC73)
are the beneficial criteria where larger values are desirable. Therefore, C1 is minimized and the rest are maximized.

Table 6. Aggregated decision matrix of evaluation criteria for the six mechanical designer alternatives.
Mechanical designer Alternatives Opt.
Abbr. Criteria Unit Weight
A1 A2 A3 A4 A5 A6 Dir.
C1 Cost x10 TL 3
345 275 585 365 485 415 0.300 ↓
C2 Experience Year 15 6 16 9 35 28 0.160 ↑
C3 Reference Number 85 109 143 63 347 119 0.170 ↑
C4 The largest project finished by x103 m2 1700 240 369.103 390 440 460 0.090 ↑
SC51 Quick Response S. S.* 6.481 3.936 7.445 6.192 4.729 3.873 0.033 ↑
SC52 Rate of attendance to the meetings S. S.* 5.635 3.722 7.667 5.733 4.729 4.401 0.020 ↑
SC53 Qualified workmanship S. S.* 6.236 5.422 7.416 5.180 4.229 3.936 0.047 ↑
C6 Diversity of working area S. S.* 3 3 2 3 4 4 0.050 ↑
SC71 No clash with other projects S. S.* 6.236 5.692 7.483 5.692 4.472 4.000 0.049 ↑
SC72 Clarification of the details in the projects S. S.* 6.481 5.958 7.238 5.477 4.472 4.229 0.081 ↑
* S.S.: Subjective Score

After forming the decision matrix, the steps of COPRAS method described in Section 2.2 were applied. First, the
normalized decision matrix was computed via Equation 5. Then, the weighted normalized decision matrix was
determined by using Equation 7. After that, the sums of weighted normalized values for the beneficial criteria (Pj)
and for the non-beneficial criteria (Rj) were computed with aid of Equation 9 and Equation 10, respectively (see
 Gul Polat et al. / Procedia Engineering 196 (2017) 278 – 285 285

Table 7). Finally, the relative significance (Qi) and the quantitative utility (Nj) for each alternative were computed by
using Equation 11 and Equation 13 as shown in Table 7.

Table 7. Result Matrix of COPRAS method.

Alternatives ܲ௝ ܴ௝ ͳΤܴ௝ ܳ௝ ܰ௝ Ranking
A1 0.133 0.042 23.865 0.189 89.136 2
A2 0.081 0.033 29.939 0.151 71.373 4
A3 0.117 0.071 14.074 0.151 70.967 5
A4 0.082 0.044 22.557 0.135 63.721 6
A5 0.172 0.059 16.976 0.212 100.000 1
A6 0.115 0.050 19.839 0.162 76.172 3

According to the ranking findings displayed in Table 7, the order of complete ranking is obtained as A5 ‫ظ‬A1 ‫ظ‬A6
‫ظ‬A2 ‫ظ‬A3 ‫ظ‬A4. A5 is the best mechanical designer team option with 100% utility degree for the studied project.
However, general contractor preferred A1 to work with as the mechanical designer team in real life. Therefore, the
outcome of proposed integrated model was discussed with the decision makers of the general contractor company.
They indicated that their primary intention was the offered bid price; hence the selection of a mechanical design
company mostly relied on a single criterion, which usually resulted in working with unqualified, incapable, and
insufficient companies. They concluded that the proposed integrated model can be used in the future project to make
sound decisions.

4. Conclusions

The selection of the appropriate mechanical designer company for a project enables to sustain construction
process more easily and without any problems. The general contractor should prefer a company, which serves well
designed mechanical project to overcome both complexity and defects of the mechanical system design. This study
proposed an integrated mechanical designer selection model to select the most appropriate mechanical designer
company in a more rational way. Two multi-criteria decision-making methods, namely AHP and COPRAS, were
combined to find the best mechanical designer team among the alternatives. In the integrated model, the AHP
method was used to determine the weights of the identified criteria. The COPRAS method was used to determine
the ranking of the mechanical designer alternatives. In order to demonstrate how the proposed integrated approach
can be applied in a real life project, a case study was carried out. The finding of the proposed integrated model was
discussed with the general contractor company. They indicated that they could use the proposed integrated model in
future projects to select the most appropriate mechanical designer company to complete the undertaken project more
successfully.

References

[1] D. R. Riley, P. Varadan, J. S. James, H. R. Thomas, Benefit-cost metrics for design coordination of mechanical, electrical, and plumbing
systems in multistory buildings, Journal of Construction Engineering and Management, 131(8), 887-889, (2005).
[2] E. Palaneeswaran, M. M. Kumaraswamy, Web-based client advisory decision support system for design–builder prequalification, Journal of
Computing in civil engineering, 19(1), 69-82, (2003).
[3] G. Onder, E. Onder, Cok Kriterli Karar Verme, Istanbul: Dora Yayincilik, 2014.
[4] J. Aguaron, J. M. Moreno-Jiménez, The geometric consistency index: Approximated thresholds, European journal of operational research,
147(1), 137-145, (2003).
[5] T. L. Saaty, Decision making with the analytic hierarchy process, International journal of services sciences, 1(1), 83-98, (2008).
[6] T. L. Saaty, How to make a decision: the analytic hierarchy process, Interfaces, 24(6), 19-43, (1994).
[7] E. K. Zavadskas, Z. Turskis, J. Tamosaitiene, V. Marina, Selection of construction project managers by applying COPRAS-G method,
Computer Modelling and New Technologies, 12(3), 22-28, (2008).
[8] E. K. Zavadskas, A. Kaklauskas, Z. Turskis, J. Tamosaitiene, Contractor selection multi-attribute model applynig COPRAS method with grey
interval numbers, In 20th International Conference/Euro Mini Conference on Continuous Optimization and Knowledge-Based Technologies
(EurOPT 2008) (pp. 241-247), (2008).
[9] N. Kundakci, A. Isık, Integration of MACBETH and COPRAS methods to select air compressor for a textile company, Decision Science
Letters, 5(3), 381-394, (2016).