You are on page 1of 10

Neural Comput & Applic

DOI 10.1007/s00521-016-2533-z

ORIGINAL ARTICLE

Supplier selection using fuzzy AHP and TOPSIS: a case study


in the Indian automotive industry
Vipul Jain1 • Arun Kumar Sangaiah2 • Sumit Sakhuja3 • Nittin Thoduka3 •

Rahul Aggarwal3

Received: 21 April 2016 / Accepted: 8 August 2016


Ó The Natural Computing Applications Forum 2016

Abstract Supplier selection is one of the key activities of 1 Introduction


purchase management in supply chain. Supplier selection is
a multifaceted problem relating qualitative and quantitative Suppliers play a major role in the production, delivery and
multi-criteria. This paper deals with a supplier selection service of competitive quality products. Purchased mate-
problem in an Indian automobile company. The work rials and services represent up to 80 % of the total product
presents selection of headlamp supplier using integrated costs in most high technology industries [14]. For any
fuzzy multi-criteria decision-making approaches: analyti- manufacturing or service business, selecting the right
cal hierarchy process (AHP) and technique for order of upstream suppliers is a key success factor that will sig-
preference by similarity to ideal solution (TOPSIS). The nificantly reduce purchasing cost, increase downstream
selection process starts with identifying the criteria based customer satisfaction, and improve competitive ability
on literature review and interviewing industry experts. [22]. The increasing demands from end customers force
Weights to criteria are assigned using AHP, and suppliers companies to focus on their core competence and let the
are ranked using AHP and TOPSIS. Consistency tests are suppliers carry out a larger part of the work than before.
carried out to check the quality of expert’s inputs. Also, This makes companies much more dependent on their
sensitivity analysis is done to check the robustness of the suppliers, and this increases the need of coordination and
approach. The results address that fuzzy approaches could adoption of supply chain management (SCM). It also
be effective and more accurate than the existing approaches means that a larger part of the supply chain has to be
for supplier selection problems. managed and to do this the network horizon must be
extended. This needs to be done in order to create pro-
Keywords Supplier selection  AHP  TOPSIS  cesses that are more efficient but also to reduce the risk of
Consistency test  Sensitivity analysis disruptions [9, 37].
Supplier selection is a strategic decision that companies
take as a result of which they identify and evaluate
potential suppliers with the view of contracting with them
offering products and services of high value [5]. The rapid
advances and competition between companies needs
effective management of supply chain for gaining more
& Arun Kumar Sangaiah profit to the industry. Nowadays, choosing a supplier
arunkumarsangaiah@gmail.com
involves more than just scanning a series of price lists. The
1
Victoria Business School, Victoria University of Wellington, choice of supplier should depend on a wide number of
Wellington, New Zealand criteria such as quality of the product, value for money of
2
School of Computing Science and Engineering, the product, reliability of the product, product variety, on-
VIT University, Vellore, India time delivery, quality of the relationship, performance
3
Department of Mechanical Engineering, G.D.Goennka World history, financial status, brand name, warranty, environ-
Institute & Lancaster University, Sohna, India mental performance. All these criteria, to some extent,

123
Neural Comput & Applic

influence the supplier selection. The factors are weighed 2 Theoretical foundations
based on the company’s needs, priorities, and strategy. The
selection criteria and the importance of each criterion vary The earlier studies [6, 13, 16–21, 32–34, 36, 40] have
from industry to industry. Firms should adopt a strategic incorporated various multi-criteria decision-making
approach to supplier relationship management and avoid a (MCDM) approaches such as DEMATEL, TOPSIS,
‘‘one-size-fits-all’’ strategy for supplier relationship man- ELECTRE, TODIM in the decision-making process. Only
agement [12, 15, 31, 39]. Having more than one supplier few studies [1, 2, 38] have integrated fuzzy AHP and
can help in reducing the risk if one supplier fails to deliver TOPSIS approaches in the decision-making process.
the products. Moreover, we have not found hybrid AHP-TOPSIS for the
India is probably the most competitive country in the evaluation of supply chain risk. Based on the context, we
world for the automotive industry. The Indian Automotive have integrated AHP and TOPSIS approaches for evalu-
Industry has one of the largest markets in the world, and ation of supply chain risk . In this work, MCDM
the car manufacturing industry is the sixth largest in the approaches are implemented to select best supplier for a
world. The automotive industry in India accounts for about component in an automobile industry. The initial step in
22 % of the country’s manufacturing gross domestic pro- this process is the identification of appropriate selection
duct (GDP). It does not cover 100 % of technology or criteria for a particular product type. This is done by
components required to make a car, but it is giving a good reviewing the literature on supplier selection problems and
97 %. Procurement is especially important in the automo- also interactions with industry experts. After the selection
tive industry because a large number of components are criteria are identified, the next step is to have a pair-wise
assembled and it is not possible for a company to produce comparison of criteria on Saaty scale. These inputs are
all those components. There are also a large number of then used to assign weights to the criteria using AHP
vendors for each component and intense competition methodology.
among them. The frequent introduction of new models and
shorter product lifecycles compounded by fast order-de- 2.1 Analytical hierarchy process (AHP)
livery require high level of agility and flexibility of the
suppliers, thereby exacerbating the supply chain com- The AHP helps decision makers find the one that best suits
plexity. With the mounting complexity of supply chain, the their goal and their understanding of the problem. It pro-
selection of the suppliers becomes very challenging. vides a comprehensive and rational framework for struc-
Recently, a number of studies [4, 7, 8, 10, 11, 23–27, 30] turing a decision problem, for representing and quantifying
have utilized MCDM approaches, either independently, or its elements, for relating those elements to overall goals,
hybridization of two or more MCDM approaches. Subse- and for evaluating alternative solutions.
quently, recent studies, we can find hybridization of The decision problem is first decomposed into a
MCDM approaches for the example DEMATEL-TOPSIS hierarchy of easily comprehensible subproblems. Each
[32], DEMATEL-TOPSIS-ELECTRE [33], AHP-PRO- subproblem is analyzed independently. The hierarchy
METHEE [35], GMM-TOPSIS [3]. On the other hand, the consists of the decision goal, the criteria for evaluating
hybridization of fuzzy AHP and TOPSIS in the context of the alternatives, and the alternatives for reaching the
evaluating the supply chain risk has not been available in goal. Table 1 defines the Saaty scale [28, 29] used for
the existing literatures. Thus, addressing this research gaps AHP.
inspired us to propose a hybrid approach based on fuzzy The AHP methodology is explained with the help of
AHP and TOPSIS approaches to evaluate the supply chain equations described below:
risk in the Indian automotive industry. Equation 1 shows the pair-wise comparison matrix.

Table 1 Saaty scale used in AHP (Saaty 1987)


Intensity of importance Definition Explanation

1 Equal importance Two factors contribute equally to the objective


3 Somewhat more important Experience and judgment slightly favor one over the other
5 Much more important Experience and judgment strongly favor one over the other
7 Very much more important Experience and judgment very strongly favor one over the other
9 Extremely more important The evidence favoring one over the other is of the highest possible affirmation
2, 4, 6, 8 Intermediate scores

123
Neural Comput & Applic

2 3
a11  a16 2.2 Technique for order preference by similarity
6 . .. .. 7 to ideal solution (TOPSIS)
A ¼ 4 .. . . 5 ð1Þ
a61  a66
TOPSIS is based on the concept that the chosen alternative
The matrix A is a m 9 m real matrix, where m is the should have the shortest geometric distance from the pos-
number of evaluation criteria considered. Each entry ajk of itive ideal solution and the longest geometric distance from
the matrix A represents the importance of the jth criterion the negative ideal solution. It is a method of compensatory
relative to the kth criterion. If ajk [ 1, then the jth criterion aggregation that compares a set of alternatives by classi-
is more important than the kth criterion, while if ajk [ 1, fying weights for each criterion, normalizing scores for
then the jth criterion is less important than the kth criterion. each criterion, and determining the geometric distance
If two criteria have the same importance, then the entry ajk between each alternative and the ideal alternative, which is
is 1. The entries ajk and akj satisfy the following constraint: the best score in each criterion.
ajkakj = 1. The initial step in TOPSIS, like in any other multi-cri-
Equation 2 is used to find the operator equation of teria decision-making technique, is to select the criteria and
matrix for normalizing. the alternatives. After that, the criteria are given weights by
the decision makers. Then the alternatives are given scores
X
j¼1
w¼ aij ¼ ½w1 ; . . .; w6  ð2Þ for each criterion and the decision matrix is formed.
j¼6 The first step is to form the decision matrix using Eq. 6.
2 3
After the pair-wise comparison matrices are com- x11    x1n
6 . .. .. 7
plete, the next step is to normalize the matrix using DM ¼ 4 .. . . 5 ð6Þ
operator equation determined by Eq. 2. This is done xm1    xmn
using Eq. 3.
2 3 The decision matrix is then normalized using Eq. 7.
a11    a16 xij
X
j¼1
6 . h i
A= ¼ 1 6 . .. .. 7
7: 1=    1= NDM ¼ rij ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Pm 2 ð7Þ
w a ij :wj ¼ 4 . . . 5 w1 w6 x
i¼1 ij
j¼6
a61    a66
2 a11 3 The weighted normalized decision matrix is formed
=w1    a16=w6
6 . using Eq. 8.
¼6 .. .. 7
7
4 .. . . 5 ð3Þ V ¼ vij ¼ wj :rij ð8Þ
a61= a
w1    66=w6 The positive ideal solution and negative ideal solution
The next step is to determine the normalized principle are determined for each criterion using Eqs. 9 and 10,
eigenvector using Eq. 4. respectively.
 
2 a11 3 PIS ¼ Vjþ ¼ MAXi Vij ð9Þ
=w1    a16=w6 2 1 3
1Xa 
j¼1
6 .  
Wj ¼ ¼6 .. .. .. 77 6 .. 7 1 NIS ¼ Vj ¼ MINi Vij ð10Þ
n j¼6
1j
w j 4 . . 5:4 . 5:n
a61= a 1 The geometric distance of each alternative from the
w1    66=w6
2 j¼1 3 positive and negative ideal solutions is calculated using
1Xa  Eqs. 11 and 12, respectively.
6 1j
w1 7
6 n j¼6 7 vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
6 7 uX
6 7 u n  þ 2
¼66
.. 7
7 ð4Þ Si ¼ t Vj  Vij ð11Þ
6 . 7 j¼1
61X j¼1
 7
4 a6j 5 vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
w6 uX
n j¼6 0 u n   2
Si ¼ t Vj  Vij ð12Þ
The final weights of the alternatives are determined using j¼1

Eq. 5.
The relative closeness to the ideal solution is determined
W ¼ MMULTðArrayi ; Arrayj Þ ð5Þ using Eq. 13

123
Neural Comput & Applic

Headlamp Supplier

Product Price/cost Quality of Manufacturing Warranty On me Environmental Brand name


quality relaonship capability delivery performance

Lumax Automove Lighng Bosch

Fig. 1 AHP hierarchy for headlamp supplier selection

Table 2 Aggregated pair-wise


Criteria PQ P/C QOR MC W OTD EP BNOS
comparison of the criteria
Product quality (PQ) 1 1 2 5 3 2 4 3
Price/cost (P/C) 1 1 3 4 2 2 3 2
Quality of relationship (QOR) 1/2 1/3 1 3 1 ‘ 1 1
Manufacturing capability (MC) 1/5  1/3 1 1/3  1/2 1/2
Warranty (W) 1/3 ‘ 1 3 1 ‘ 1 1
On-time delivery (OTD) 1/2 ‘ 2 4 2 1 3 3
Environmental performance (EP) 1/4 1/3 1 2 1 1/3 1 1
Brand name of supplier (BNOS) 1/3 ‘ 1 2 1 1/3 1 1
P
Sum = each column 4 1/9 4 3/7 11 1/3 24 11 1/3 7 14 1/2 12 1/2

Table 3 Normalized criteria


PQ P/C QOR MC W OTD EP BNOS Average
matrix and weights
PQ 0.242 0.226 0.176 0.208 0.264 0.289 0.275 0.240 0.240482
P/C 0.242 0.226 0.264 0.166 0.176 0.289 0.206 0.160 0.216653
QOR 0.121 0.075 0.088 0.125 0.088 0.072 0.068 0.080 0.089957
MC 0.048 0.056 0.029 0.041 0.029 0.036 0.034 0.040 0.039538
W 0.080 0.113 0.088 0.125 0.088 0.0722 0.068 0.080 0.089613
OTD 0.121 0.113 0.176 0.166 0.176 0.144 0.206 0.240 0.168218
EP 0.060 0.075 0.088 0.083 0.088 0.048 0.068 0.080 0.074145
BNOS 0.080 0.113 0.088 0.083 0.088 0.048 0.068 0.080 0.081393

0
S from two experts, and those were aggregated by taking
RelativeCloseness ¼ Ci ¼ 0 i  ð13Þ their average values. The methodology explained in pre-
Si þ S i
vious section is adopted to assign weights and rank the
The alternative with the highest relative closeness is the alternatives. The hierarchy of the supplier selection is
best alternative. shown in Fig. 1.
Step 1 Assigning weight to criteria
3 Numerical illustration The aggregated pair-wise comparisons of the criteria on
the Saaty scale are shown in Table 2. Equation 2 is used to
The present case discusses the supplier selection of head- find the operator equation of the matrix.
lamps in leading automobile company in India. The criteria Equations 3 and 4 are used to normalize the matrix and
were identified and shortlisted after interviewing industry assign weights to the criteria. The normalized matrix and
experts. The inputs were taken in pair-wise comparison assigned weights are shown in Table 3.

123
Neural Comput & Applic

Table 4 Aggregated pair-wise comparison

Product Quality Lumax Automotive Lighting Bosch Price/Cost Lumax Automotive Lighting Bosch
Lumax 1 1/5 1/4 Lumax 1 6 6
Automotive Lighting 5 1 2 Automotive Lighting 1/6 1 1
Bosch 4 ½ 1 Bosch 1/6 1 1
Sum=∑each column 10.00 1.70 3.25 Sum=∑each column 1.33 8.00 8.00
a) Aggregated pair-wise comparison of the supplier for the b) Aggregated pair-wise comparison of the supplier for the price/cost criterion
product quality criterion
Manufacturing Capability Lumax Automotive Lighting Bosch
Quality of Relationship Lumax Automotive Lighting Bosch
Lumax 1 ¼ 1/7
Lumax 1 2 7
Automotive Lighting ½ 1 4 Automotive Lighting 4 1 1/3
Bosch 1/7 ¼ 1 Bosch 7 3 1
Sum=∑each column 1.64 3.25 12.00 Sum=∑each column 12.00 4.25 1.48
c) Aggregated pair-wise comparison of the supplier for quality of d) Aggregated pair-wise comparison of the supplier for the manufacturing
relationship criterion capability criterion

Warranty Lumax Automotive Lighting Bosch On Time Delivery Lumax Automotive Lighting Bosch
Lumax 1 1/2 1/5 Lumax 1 2 1/2
Automotive Lighting 2 1 ½ Automotive Lighting 1/2 1 1/4
Bosch 5 2 1 Bosch 2 4 1
Sum=∑each column 8.00 3.50 1.70 Sum=∑each column 3.50 7.00 1.75
e) Aggregated pair-wise comparison of the supplier for the warranty f) Aggregated pair-wise comparison of the supplier for the on time delivery
criterion criterion

Environmental Lumax Automotive Lighting Bosch Brand Name of the Supplier Lumax Automotive Lighting Bosch
Performance
Lumax 1 1/3 1/7
Lumax 1 1/6 3
Automotive Lighting 6 1 9 Automotive Lighting 3 1 1/2
Bosch 1/3 1/9 1 Bosch 7 2 1
Sum=∑each column 7.33 1.28 13.00 Sum=∑each column 11.00 3.33 1.64
g) Aggregated pair-wise comparison of the supplier for the h) Aggregated pair-wise comparison of the supplier for the brand name of the
environmental performance criterion supplier criterion

Table 5 Normalized matrix and weights

Product Quality Lumax Automotive Lighting Bosch Average Price/Cost Lumax Automotive Lighting Bosch Average
Lumax 0.1 0.117647059 0.07692308 0.09819 Lumax 0.75 0.75 0.75 0.75
Automotive Lighting 0.5 0.588235294 0.61538462 0.567873 Automotive Lighting 0.125 0.125 0.125 0.125
Bosch 0.4 0.294117647 0.30769231 0.333937 Bosch 0.125 0.125 0.125 0.125
5b: Normalized matrix and weights for the price/cost criterion
5a: Normalized matrix and weights for the product quality criterion

Quality of Lumax Automotive Bosch Average Manufacturing Lumax Automotive Bosch Average
Relationship Lighting Capability Lighting
Lumax 0.608696 0.615384615 0.58333333 0.602471 Lumax 0.083333 0.058823529 0.09677419 0.079644
Automotive Lighting 0.304348 0.307692308 0.33333333 0.315124 Automotive 0.333333 0.235294118 0.22580645 0.264811
Lighting
Bosch 0.086957 0.076923077 0.08333333 0.082404 Bosch 0.583333 0.705882353 0.67741935 0.655545
5c: Normalized matrix and weights for the quality of relationship criterion 5d: Normalized matrix and weights for the manufacturing capability criterion

Warranty Lumax Automotive Bosch Average On Time Lumax Automotive Bosch Average
Lighting Delivery Lighting
Lumax 0.125 0.142857143 0.11764706 0.128501 Lumax 0.285714 0.285714286 0.28571429 0.285714
Automotive Lighting 0.25 0.285714286 0.29411765 0.276611 Automotive 0.142857 0.142857143 0.14285714 0.142857
Lighting
Bosch 0.625 0.571428571 0.58823529 0.594888 Bosch 0.571429 0.571428571 0.57142857 0.571429
5e: Normalized matrix and weights for the warranty criterion 5f: Normalized matrix and weights for the on time delivery criterion

Environmental Lumax Automotive Bosch Average Brand Name of Lumax Automotive Bosch Average
Performance Lighting the Supplier Lighting
Lumax 0.136364 0.130434783 0.23076923 0.165856 Lumax 0.090909 0.1 0.08695652 0.092622
Automotive Lighting 0.818182 0.782608696 0.69230769 0.764366 Automotive 0.272727 0.3 0.30434783 0.292358
Lighting
Bosch 0.045455 0.086956522 0.07692308 0.069778 Bosch 0.636364 0.6 0.60869565 0.61502
5g: Normalized matrix and weights for the environmental performance criterion 5h: Normalized matrix and weights for the brand name of the supplier criterion

123
Neural Comput & Applic

Table 6 Weighted standardized decision matrix


Criteria/alternative Lumax Automotive lighting Bosch Max Min

Product quality 0.023613 0.13656349 0.080306 0.13656 0.02361


Price/cost 0.16249 0.027081662 0.027082 0.16249 0.02708
Quality of relationship 0.054196 0.028347593 0.007413 0.05419 0.00741
Manufacturing capability 0.003149 0.010470119 0.025919 0.02591 0.00314
Warranty 0.011515 0.024787926 0.05331 0.05331 0.01151
On-time delivery 0.048062 0.02403121 0.096125 0.09612 0.02403
Environmental performance 0.012297 0.056674176 0.005174 0.05667 0.00517
Brand name of supplier 0.007539 0.02379583 0.050058 0.05005 0.00753
P
AHP result = each column 0.322862 0.331752006 0.345386 (best alternative)

Table 7 Geometric distance


Criteria/alternative Lumax Automotive lighting Bosch
from (a) positive ideal solution
and (b) negative ideal solution (a)
Product quality 0.012758 0 0.003165
Price/cost 0 0.018335411 0.018335
Quality of relationship 0 0.00066816 0.002189
Manufacturing capability 0.000518 0.000238667 0
Warranty 0.001747 0.000813493 0
On-time delivery 0.00231 0.005197491 0
Environmental performance 0.001969 0 0.002652
Brand name of supplier 0.001808 0.000689707 0
P
Si* = H (solution - positive ideal solution)2 0.145294 0.161068089 0.1623
(b)
Product quality 0 0.01275782 0.003214
Price/cost 0.018335 0 0
Quality of relationship 0.002189 0.000438264 0
Manufacturing capability 0 5.35994E-05 0.000518
Warranty 0 0.00017616 0.001747
On-time delivery 0.000577 0 0.005197
Environmental performance 5.07E-05 0.002652297 0
Brand name of supplier 0 0.000264293 0.001808
P
Si’ = H (solution - negative ideal solution)2 0.145439 0.127837529 0.111735

Table 8 Relative closeness


Lumax Automotive lighting Bosch

Si* ? Si’ 0.290732 0.288905619 0.274035


Si’/(Si* ? Si’) 0.500249 (best alternative) 0.442488899 0.40774

It can be seen from Table 3 that product quality (PQ) is of alternatives for each criterion. The operator equation for
the most important selection criteria with a weight of each matrix is formed using Eq. 2. Equations 3 and 4 are
0.240482 followed by price/cost and on-time delivery. The used to normalize the comparison matrices of the suppliers
least important criterion is manufacturing capability. for each criterion. Table 5a–h shows the normalized
matrices and assigned weights.
Step 2 Assigning weights to suppliers with respect to
criteria Step 3 Rank suppliers using AHP.
For each criterion, the suppliers are compared on the The weighted normalized decision matrix is constructed
Saaty scale. Table 4a–h shows the aggregated comparison using Eq. 8. The ranking of the suppliers based on AHP is

123
Neural Comput & Applic

Table 9 Summary of the consistency test calculated using Eq. 5 or by finding the sum of columns of
Pair-wise comparison matrix Consistency ratio (%)
the weighted standardized decision matrix. Table 6 shows
the weighted standardized decision matrix. The positive
Criteria 1.61091206 and negative ideal solutions are obtained using Eqs. 9 and
Product quality 2.37091849 10, respectively.
Price/cost 0.00000000
Step 4 Rank suppliers using TOPSIS
Quality of relationship 0.19057978
Manufacturing capability 3.12829819 The geometric distances from the positive and negative
Warranty 0.53256737 ideal solutions are calculated using Eqs. 11 and 12,
On-time delivery 0.00000000 respectively. Table 7a, b shows the geometric distance of
Environmental performance 5.23066173 the solution from the positive and the negative ideal solu-
Brand name of supplier 0.25402362 tions, respectively.

Fig. 2 Sensitivity analysis (a) Sensitivity Analysis of the Product Quality Factor (b) Sensitivity Analysis of the Price/Cost Factor
Bosch Bosch
Automove Lighng Automove Lighng
Lumax Lumax

0 0.1 0.2 0.3 0.4 0.5 0.6 0 0.2 0.4 0.6 0.8
Alternave Score Alternave Score

(c) Sensitivity Analysis of the Quality of


(d) Sensitivity Analysis of the Manufacturing
Relationship Factor
Capability Factor
Bosch
Bosch
Automove Lighng
Automove Lighng
Lumax
Lumax
0 0.2 0.4 0.6 0.8
0 0.2 0.4 0.6 0.8
Alternave Score
Alternave Score

(e) Sensitivity Analysis of the Warranty Factor (f) Sensitivity Analysis of the On Time
Delivery Factor
Bosch Bosch
Automove Lighng
Automove Lighng
Lumax
Lumax
0 0.1 0.2 0.3 0.4 0.5 0.6
0 0.2 0.4 0.6
Alternave Score
Alternave Score
(g) Sensitivity Analysis of the Environmental (h) Sensitivity Analysis of the Brand Name of the
Performance Factor Supplier Factor
Bosch Bosch
Automove Lighng Automove Lighng
Lumax Lumax

0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8


Alternave Score Alternave Score
(i) Sensitivity Analysis with Equal Weights

Bosch
Automove Lighng
Lumax

0 0.1 0.2 0.3 0.4 0.5


Alternave Score

123
Neural Comput & Applic

Fig. 3 Sensitivity analysis graph of the AHP results

Table 10 Ranking of the suppliers based on the AHP results comparison matrix and alternative matrices. A summary of
Supplier Weight Rank the consistency test is shown in Table 9.

Bosch 0.345386 1 3.2 Sensitivity analysis of the AHP results


Automotive Lighting 0.331752006 2
Lumax 0.322862 3 The sensitivity analysis is performed in order to test the
stability of the priority ranking. In sensitivity analysis, one
criterion is assigned a weight of 90 % and the 10 % is
Table 11 Ranking of the suppliers based on the TOPSIS results distributed among the other criteria in the ratio of the
weights assigned in the beginning; this is done for all cri-
Supplier Weight Rank
teria. The final AHP result is calculated in each case. The
Lumax 0.500249 1 criteria are also assigned equal weights. The nine cases of
Automotive Lighting 0.442488899 2 the sensitivity analysis are shown in Fig. 2a–i.
Bosch 0.40774 3 The overall sensitivity graph of the AHP results is
shown in Fig. 3. The scores of the suppliers are on the
y-axis, and the criteria are on the x-axis. The average/
The relative closeness is determined using Eq. 13. overall results are also shown in the graph.
Table 8 shows the relative closeness of different
alternatives.
Lumax is found to be the best alternative among the 4 Results and discussion
three alternatives, because its relative closeness is maxi-
mum with a value of 0.500249. The second best is Auto- The AHP results indicate that Bosch is the best supplier
motive Lighting with value of relative closeness equal to followed by Automotive Lighting and Lumax as shown in
0.442488899. Table 10.
The TOPSIS results, however, indicate that Lumax is
3.1 Consistency test the best supplier followed by Automotive Lighting and
Bosch as shown in Table 11.
The consistency test is performed in order to ensure that all The consistency of all the comparison matrices is gen-
the expert inputs are consistent. It is done for the criteria erally high. The consistency ratio is always less than 10 %.

123
Neural Comput & Applic

The results of the sensitivity analysis of AHP mirror the 9. Christopher M, Mena C, Khan O, Yurt O (2011) Approaches to
AHP results. However, the results of the sensitivity anal- managing global sourcing risk. Supply Chain Manag Int J
16(2):67–81
ysis of TOPSIS are the opposite of the TOPSIS results. 10. Dalalah D, Al-Oqla F, Hayajneh M (2010) Application of the
Therefore, it is concluded that in this case the AHP results analytic hierarchy process (AHP) in multi-criteria analysis of the
are more robust and are used for the final ranking of the selection of cranes. Jordan J Mech Ind Eng 4(5):567–578
suppliers. Bosch is the best supplier followed by Auto- 11. Deng X, Hu Y, Deng Y, Mahadevan S (2014) Supplier selection
using AHP methodology extended by D numbers. Expert Syst
motive Lighting and Lumax. Appl 41(1):156–167
12. Dyer JH, Cho DS, Chu W (1998) Strategic supplier segmentation:
the next ‘‘best practice’’ in supply chain management. Calif
5 Conclusion Manag Rev 40(2):57–77
13. Gopal J, Sangaiah AK, Basu A, Gao XZ (2015) Integration of
fuzzy DEMATEL and FMCDM approach for evaluating knowl-
This paper has illustrated case study for the supplier edge transfer effectiveness with reference to GSD project outcome.
selection of headlamps in leading automobile company in Int J Mach Learn Cybernet. doi:10.1007/s13042-015-0370-5
India. The eight criteria were investigated in the context of 14. Gülen KG (2007) Supplier selection and outsourcing strategies in
supply chain management. J Aeronaut Space Technol 3(2):1–6
supplier selection problem. 15. Gürler AGİ (2007) Supplier selection criteria of Turkish auto-
Moreover, this study highlights the two valuable con- motive industry. J Yaşar Univ 2(6):555–569
tributions: (a) comprehensive structure of this paper is to 16. Handfield R, Walton SV, Sroufe R, Melnyk SA (2002) Applying
address the supplier selection problem, in this context a environmental criteria to supplier assessment: a study in the
application of the analytical hierarchy process. Eur J Oper Res
hybrid AHP-TOPSIS methodology has been utilized to 141(1):70–87
determine the relative importance of the supplier selection 17. Ho W, Xu X, Dey PK (2010) Multi-criteria decision making
factors via Saaty scales. (b) Accordingly, there are limited approaches for supplier evaluation and selection: a literature
studies on hybrid MCDM (AHP-TOPSIS) methods for review. Eur J Oper Res 202(1):16–24
18. Jamil N, Besar R, Sim HK (2013) A study of multicriteria
investigating supplier selection problem. Subsequently, the decision making for supplier selection in automotive industry.
future work in this area can be to develop new criteria’s by J Ind Eng 2013:1–22, Article ID 841584. doi:10.1155/2013/
adopting the new framework and hybrid computational 841584
intelligence techniques such as neuro-fuzzy, fuzzy-genetic 19. Koprulu A, Albayrakoglu MM (2007) Supply chain management in
the textile industry: a supplier selection model with the analytical
and other optimization methods (particle swarm opti- hierarchy process. In: Proceeding of the international symposium
mization, ant colony optimization and so on. The results on the analytic hierarchy process, Viña Del Mar, Chile
obtained in this paper have been compared with proposed 20. Kornyshova E, Salinesi C (2007) MCDM techniques selection
method, which is the new avenue for future research. approaches: state of the art. In: IEEE symposium on computa-
tional intelligence in multicriteria decision making, pp. 22–29
21. Kumar S, Parashar N, Haleem A (2009) Analytical hierarchy
process applied to vendor selection problem: small scale, medium
References scale and large scale industries. Bus Intell J 2(2):355–362
22. Liao CN, Kao HP (2010) Supplier selection model using Taguchi
1. Al Maliki A, Owen G, Bruce D (2012) Combining AHP and loss function, analytical hierarchy process and multi-choice goal
TOPSIS approaches to support site selection for a lead pollution programming. Comput Ind Eng 58(4):571–577
study. Doctoral dissertation, IACSIT Press 23. Marufuzzaman M, Ahsan KB, Xing K (2009) Supplier selection
2. Alsuwehri YN (2011) Supplier evaluation and selection by using and evaluation method using Analytical Hierarchy Process
the analytic hierarchy process approach. Doctoral dissertation, (AHP): a case study on an apparel manufacturing organisation.
The University of Kansas Int J Value Chain Manag 3(2):224–240
3. Aly MF, Attia HA, Mohammed AM (2013) Integrated fuzzy 24. Özkan B, Başlıgil H, Şahin N (2011) Supplier selection using
(GMM)-TOPSIS model for best design concept and material analytic hierarchy process: an application from Turkey. In: Pro-
selection process. Int J Innov Res Sci Eng Technol 2(11): ceedings of the world congress on engineering, vol 2, pp. 6–8
6464–6486 25. Pitchipoo P, Venkumar P, Rajakarunakaran S (2013) Fuzzy
4. Behzadian M, Otaghsara SK, Yazdani M, Ignatius J (2012) A hybrid decision model for supplier evaluation and selection. Int J
state-of the-art survey of TOPSIS applications. Expert Syst Appl Prod Res 51(13):3903–3919
39(17):13051–13069 26. Rezaei J, Ortt R (2013) Multi-criteria supplier segmentation using
5. Beşkese A, Şakra A (2010) A model proposal for supplier a fuzzy preference relations based AHP. Eur J Oper Res
selection in automotive industry. In: 14th international 225(1):75–84
research/expert conference TMT 27. Rouyendegh BD, Saputro TE (2014) Supplier selection using
6. Büyüközkan G (2012) An integrated fuzzy multi-criteria group integrated fuzzy TOPSIS and MCGP: a case study. Procedia Soc
decision-making approach for green supplier evaluation. Int J Behav Sci 116:3957–3970
Prod Res 50(11):2892–2909 28. Saaty RW (1987) The analytic hierarchy process—what it is and
7. Cabala P (2010) Using the analytic hierarchy process in evalu- how it is used. Math Model 9(3):161–176
ating decision alternatives. Oper Res Decis 1:1–23 29. Saaty TL (1990) How to make a decision: the analytic hierarchy
8. Chamodrakas I, Batis D, Martakos D (2010) Supplier selection in process. Eur J Oper Res 48(1):9–26
electronic marketplaces using satisficing and fuzzy AHP. Expert 30. Sadeghzadeh K, Salehi MB (2011) Mathematical analysis of fuel
Syst Appl 37(1):490–498 cell strategic technologies development solutions in the

123
Neural Comput & Applic

automotive industry by the TOPSIS multi-criteria decision 36. Verma DS, Pateriya A (2013) Supplier selection through ana-
making method. Int J Hydrogen Energy 36(20):13272–13280 lytical hierarchy process: a case study in small scale manufac-
31. Sagar MK, Singh D (2012) Supplier selection criteria: study of turing organization. Int J Eng Trends Technol 4(5):1428–1433
automobile sector in India. Int J Eng Res Dev 4(4):34–39 37. Viswanadham N, Vedula S, Kulkarni R (2013) Orchestrating the
32. Sangaiah AK, Subramaniam PR, Zheng X (2015) A combined print supply chain in emerging markets. In: 2013 IEEE interna-
fuzzy DEMATEL and fuzzy TOPSIS approach for evaluating tional conference on automation science and engineering
GSD project outcome factors. Neural Comput Appl 26(5): (CASE), pp. 213–218
1025–1040 38. Xu L, Kumar DT, Shankar KM, Kannan D, Chen G (2013)
33. Sangaiah AK, Gopal J, Basu A, Subramaniam PR (2015) An Analyzing criteria and sub-criteria for the corporate social
integrated fuzzy DEMATEL, TOPSIS, and ELECTRE approach responsibility-based supplier selection process using AHP. Int J
for evaluating knowledge transfer effectiveness with reference to Adv Manuf Technol 68(1–4):907–916
GSD project outcome. Neural Comput Appl. doi:10.1007/ 39. Yilmaz O, Gülsün B, Güneri AF, Özgürler Ş, Endüstri BY,
s00521-015-2040-7 Besiktas MB (2011) Supplier selection of a textile company with
34. Sangaiah AK, Gao XZ, Ramachandran M, Zheng X (2015) A ANP. In: 15th international research/expert conference, ‘‘Trends
fuzzy DEMATEL approach based on intuitionistic fuzzy infor- in the development of machinery and associated technology’’,
mation for evaluating knowledge transfer effectiveness in GSD TMT 2011, Prague, Czech Republic
projects. Int J Innovative Comput Appl 6(3–4):203–215 40. Zarbini-Sydani A, Karbasi A, Atef-Yekta E (2011) Evaluating
35. Shakey BK (2006) Supplier selection using AHP and promethee- and selecting supplier in textile industry using hierarchical fuzzy
2. Int J Sci Res 6:156–160 TOPSIS. Indian J Sci Technol 4(10):1322–1334

123

You might also like