About this Journal Submit a Manuscript Table of Contents
Journal of Industrial Engineering
Volume 2013 (2013), Article ID 841584, 22 pages
http://dx.doi.org/10.1155/2013/841584
Research Article

A Study of Multicriteria Decision Making for Supplier Selection in Automotive Industry

Faculty of Engineering and Technology, Multimedia University, Jalan Ayer Keroh Lama, 75450 Bukit Beruang, Melaka, Malaysia

Received 8 November 2012; Accepted 29 January 2013

Academic Editor: C. K. Kwong

Copyright © 2013 Nadia Jamil et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

This paper is designed to present the effectiveness of group multicriteria decision making in automotive manufacturing company focusing on the selection of suppliers in Malaysia. The process of selecting suppliers is one of the most critical and challenging endeavor in any supply chain management. There are five decision making tools being analyzed in this study, namely, analytical hierarchy process (AHP), fuzzy analytical hierarchy process (FAHP), technique for order performance by similarity to ideal solution (TOPSIS), fuzzy technique for order performance by similarity to ideal solution (FTOPSIS), and fuzzy analytical hierarchy process integrated with fuzzy technique for order performance by similarity to ideal solution (FAHPiFTOPSIS). The scores of ranking among the suppliers in each MCDM tools (AHP, FAHP, TOPSIS, FTOPSIS, and FAHPiFTOPSIS) show significantly comparable variation. Scores of the best supplier is then compared to the lowest supplier for all MCDM tools whereby this reflects that the highest percentage goes to TOPSIS with scoring of 79.37%. On the contrary, FAHPiFTOPSIS demonstrated the lowest score variation of 22.42% which indicates that FAHPiFTOPSIS is able to eliminate biasness in supplier selection process.

1. Introduction

A supply chain is a system which connects several departments from procurement of raw materials, to manufacturing, warehousing, and distribution of the products to the customers. Part of the contribution to supply chain complexity is the geographical outsourcing for cheaper supply and new market penetration. The complexity of supply chain is aggravated further when industry rely too much on multirange products and frequent introduction of new products as a strategy to meet different segmented market demands.

In automotive industry, such situation is rampant. The frequent introduction of new models and shorter product lifecycles compounded by fast order-delivery require high level of agility and flexibility of the suppliers, thereby, exacerbating the supply chain complexity. Hence, the right selection of supplier becomes more complicated. With the mounting complexity of supply chain, the selection of the suppliers becomes very challenging. The recent incident in Fukushima, Japan, devastated by massive earthquake and nuclear disaster, and major floods in Thailand, had affected severely many Malaysian industries as well as industries in other parts of the world [1]. Those Malaysian companies which have their suppliers associated with these companies have suffered critical production problems, in particular, automotive industry. Thus, purchasing function for each organization is increasingly seen as the most important role in supplier selection [2].

In the previous years the automotive industry has witnessed an unprecedented turmoil. Such crises had affected the European and Asian automotive industry and had gravely stricken the American automobile industry. The first half of 2009 had indeed been a very faltering year due to economy recession. Based on press report released by MAA [3] dated 19 January 2011 (Figure 1), sales of passenger cars in Malaysia had dropped by 7.5% from year-to-date September 2008 to September 2009 (see Table 1) affected by the bad economic slowdown.

tab1
Table 1: Vehicles Sales in 2008 and 2009.
841584.fig.001
Figure 1: Vehicles sales in 2008 and 2009.

The economy is currently in transition from recession to recovery in tandem with financial markets improvement. Year 2010 shows an increase trend in automotive industry. In 2010, the total industry volume of vehicles has picked up and increased to 605,156 units and, in fact, surpassed the highest recorded volume of 552,316 units in 2005.

Table 1 shows total vehicles produced and assembled from 1980 to 2010. From Table 2, total vehicles produced and assembled in 2010 were 605,156 units, increased by 12.7% from 536,905 units produced the year before. In 1980 the production statistics shows only two types of vehicles: passenger cars and commercial vehicles. In 1990, the statistics had included production figures for 4WD vehicles. In 2007, there was a reclassification where passenger cars, 4WD/SUV, window van, and MPV models are all categorized as passenger vehicles. Commercial vehicles had been reclassified to include trucks, prime movers, pick-up, panel vans, bus, and others. The data in Table 1 is summarized graphically in Figure 2 which illustrates the bar graph for passenger and commercial vehicles produced and assembled in Malaysia from year 1980 to 2010.

tab2
Table 2: Summary of passenger and commercial vehicles produced and assembled in Malaysia from 1980 to 2010.
841584.fig.002
Figure 2: Bar graph of passenger and commercial vehicles produced and assembled in Malaysia from 1980 to 2010 [3].

The data in Table 2 is summarized in bar graph shown in Figure 3 to highlight the significant increase in the number of passengers cars produced and assembled from 1980 to 2010. The total number of vehicles produced and assembled also showed a good progress of increment from year to year. To see the production trend of commercial vehicles from 2005 to 2010, Figure 3 is presented. From Figure 3, it is noted that the production of commercial vehicles did not exceed more than 100,000 units.

841584.fig.003
Figure 3: Line graph of passenger and commercial vehicles produced and assembled in Malaysia from 2005 to 2010 [3].

In Figure 3, the line graph representing the commercial vehicles illustrates a slight increment of production from 2009 to 2010 which is an increase of 6.8% of 45,147 units. In terms of percentage contribution of commercial vehicles to the market, it gradually increases from 8.8% in 2005 to 10.2% in 2010 [4].

In May 2011, based on a press release by Malaysian Automotive Association (MAA), automotive industry has shown a healthy increase of sales compared to the same period the year before (see Table 3). However, in comparison to sales in May 2010 and May 2011, it showed that there was a reduction of 11.5%. This impact is due to the earthquake and tsunami which happened in Japan. More than 80% of the vehicles produced were dominated by passenger vehicles.

tab3
Table 3: Summary of passenger and commercial vehicles produced and assembled in Malaysia from May 2010 to May 2011.

Currently, automotive industry is slowly and gradually shifting towards Asian countries, mainly due to high cost and saturation of automobile industry in the west and the increase in demand in Asia. The principal driving markets for Asian automobile industry are China, India, and ASEAN nations. The future of automotive industry in the Asian countries in particular Thailand, Philippines, Malaysia, and Indonesia is bright and promising because of the (ASEAN Free Trade Area) AFTA with tariffs currently at 0 to 5% [5].

Malaysia is a country that has a long history of making cars. There are many local and international cars assemblers and manufacturers in the country. The employees of the industry are widely regarded as skilled, well educated, and trainable. Located strategically in the ASEAN region which has a population of more than 500 million people, Malaysia offers vast opportunities for global automotive and component manufacturers and suppliers to set up their manufacturing and distribution operations in the country. Mercedes-Benz assembly plant has proven its great prospect where its plant located in Pekan, Pahang, initially assembled only 4 units per day for one model and today assembles the S-Class, E-Class, and C-Class Mercedes with annual volume reaching 5,000 units [6].

Supplier selection decision process considers qualitative and quantitative criteria [2, 7]. It is a complex process whereby it involves many criteria, not necessarily quality, cost, and delivery. Such decision making process that involves multiple criteria is classified as a multicriteria decision making (MCDM) process.

One of the MCDM is the analytical hierarchy process (AHP) whereby it is a theory of mathematical for decision making and measurement introduced by Saaty [8]. It assists decision makers to make effective decision based on its goals, criteria, and alternatives. AHP can be applied in making decisions which are unstructured and complex and consists of multiple criteria [9]. This method has been used in various areas including performance evaluation, supplier selection, credit scoring, project management, resource allocation, distribution channel management, inventory management, promotion and recruitment decisions, portfolio management, energy resources planning, technology management, financial planning, budgeting decisions, socioeconomic planning, common vote prediction, and conflict resolution [10]. The Technique for Order Preference by Similarity to the Ideal Solution (TOPSIS) is another mathematical model for MCDM. TOPSIS advocates two artificial alternatives, that is, the ideal alternative, the one which has the best level for all attributes considered, and the negative ideal alternative, the one which has the worst attribute values. TOPSIS selects the alternative that is the closest to the ideal solution and farthest from negative ideal alternative [11]. For the purpose of this research and without any prejudice to the others, two MCDM mathematical models have been selected, that is, AHP and TOPSIS.

In making decision for supplier selection, it is agreed by [2, 12] that the decisions made are very often involved by several decision makers. The purpose of having more people in decision making is to avoid any weaknesses or prejudice in the selection process. In order to make a decision which reflects human thinking, a system needs to be realistic. In making decision, decision makers prefer to evaluate a criterion with a certain level of tolerance rather than a fixed value judgment [13]. Due to this, one system which implies a human-like thinking model is introduced which is known as fuzzy logic [14].

This system describes a matter with a certain degree of characteristic which is also known as membership function. Membership function is a graphical representation which associates with the magnitude of input and ultimately determines an output response. There are different membership functions associated with each input and output response. Details about membership function are explained in Section 2.

The remainder of this paper is organized as follows: in Section 2, fuzzy concepts and the integration between fuzzy AHP and TOPSIS are explained; Section 3 shows the methodology used to conduct this study; Section 4 explained the results and discussion, finally in Section 5, the conclusions are presented.

2. Background of Fuzzy Concepts

Fuzzy sets can be simply defined as a set with fuzzy boundaries whereby the values of boundaries is multivalued unlike the two-valued Boolean logic. In the fuzzy theory, fuzzy set of universe is defined by function which is called the membership functions of set . This notation can be expressed as follows [15]: where

The above set explains the membership (characteristic) functions of which has the value ranges from 0 to 1. It allows a wide range of possible values. The value from 0 to 1 in this set represents the degree of membership of element in set . The membership function is commonly illustrated in terms of membership curve.

2.1. Triangular Fuzzy Number (TFN)

There are 2 types of TFN which are symmetry and unsymmetry. Symmetry TFN is used in this paper since it enables users to easily calculate, understand, and capture the vagueness in people’s verbal assessments [16].

TFN can be defined in three numbers, () which represented the smallest possible value, the most promising value, and the largest value which describes the fuzzy event. This representation can be observed from Figure 4 which illustrates symmetry TFN [17].

841584.fig.004
Figure 4: Symmetry TFN.

Throughout this study, the commonly used algebraic operations for fuzzy numbers are addition and multiplication. The fuzzy operators shown below were adapted from [18]. Let and be two triangular fuzzy numbers with their parameters shown as follows: Fuzzy numbers addition is defined by On the other hand, fuzzy numbers multiplication is calculated as shown below:

2.2. AHP and Fuzzy AHP

The analytical hierarchy process (AHP) was first developed by Saaty, in mid of 1970 [19]. It is a decision analysis method that considers both qualitative and quantitative attributes in decision making. It decomposes complex problem into tangible and intangible entities, transforms them into systematic hierarchies of criteria and alternatives, and ranks them in order from most to least importance or desirable. As such the complex problem can easily be comprehended and each level can be analyzed independently. Each criterion from the AHP structure will be compared against a given alternatives, thus, allowing it to judge the intensity of the importance of one criterion over the other based on the given alternatives [19].

AHP is simple, systematic, and a very useful approach which integrates the matrix theory. In defining the weights of criteria and comparing the alternatives, a set of pair-wise comparison has been developed by Saaty [8] as part of the steps in AHP. Since the concept of AHP considers only real exact number, the set of pair-wise comparison by Saaty has been modified to suit human’s judgement. This is the main reason, and the scale of weight and criteria preferences are defined in terms of linguistic variables in fuzzy numbers as presented in Tables 4 and 5 [2]. Linguistic variables are used to evaluate and assess the ratings of suppliers and weight of criteria. It is defined as variables with values expressed by words in a natural language [30].

tab4
Table 4: Linguistic variables for importance of the criteria.
tab5
Table 5: Linguistic variables for performance of the alternatives.

There are three principles used in solving problems with AHP [9] in supplier selection which are as follows.(i)AHP establishes the priorities based on sets of pair-wise comparisons. (ii)AHP score is built on human attributes and judgements where the intensity of each attribute or judgment is set according to its hierarchy over the other.(iii)AHP synthesizes these judgments by using the hierarchy framework to obtain the overall priority of the elements or factors.

The concept of AHP and FAHP is mainly the same, and the difference of FAHP is that it analyses the numbers in terms of fuzzy numbers but as for AHP the numbers analyzed are crisp numbers. FAHP was firstly introduced by Van Laarhoven and Pedrycz in 1983 [31] whereby the fuzzy judgment is represented by triangular fuzzy number.

There are 6 steps in the process of decision making using FAHP. These steps are as follows.(1)Firstly, form a decision matrix of the importance of each criterion with respect to each other.(2)The membership function of triangular fuzzy number is defined by three real numbers which is mathematically described as . The comparison matrix is derived from (6) as (3)Calculate the   (objective) which is defined by Chang in 1992 [32] in his study of fuzzy extent analysis. The computations of are explained in (7) as (4)The next step is the calculation of (total objective) whereby the equation used is displayed in (8) as (5)Upon getting the result for , weight of each criterion with respect to the objective is calculated. The equations used are shown in (9) as (6)The final stage is defuzzification whereby triangular fuzzy numbers are transformed into real numbers which is defined as weights () and can be computed in (10) as

The steps of 1 to 6 will be repeated for each alternative in terms of the criterion which will be called as the weight of factors in terms of alternatives.

2.3. TOPSIS and Fuzzy TOPSIS

TOPSIS has been used in various fields and a number of applications such as outsourcing of logistics service, weapon selection, supplier selection analyzing business competition, and many other applications [24, 25, 33]. This method considers three types of criteria which are cost, qualitative, and quantitative. Generally, there are five steps in solving a problem using TOPSIS [34]. These steps are as follows:(1)Obtain normalized decision matrix for alternatives over criteria. Let score of option with respect to criterion . Data can be normalized as below: for and .(2)Construct weighted normalized decision matrix, , whereby is a set of weights for each criteria and is the normalized decision matrix.(3)Identify the ideal alternative (extreme performance on each criterion, ) and negative ideal alternative (reverse extreme performance on each criterion, ) as whereby if if and whereby if if .(4)Develop a distance measure over criterion to both ideal () and negative ideal alternative () as (5)For each alternative, determine the relative closeness to the ideal solution, .   is equal to the distance of the negative ideal solution divided by the sum of the distance of the negative ideal and ideal solution given by

TOPSIS only considers crisp values, whereas human judgments are usually uncertain and could not be evaluated using fix numbers. In spite of this, fuzzy numbers are used to replace all the crisp values in TOPSIS. In decision making, it is difficult to give a certain judgement, hence by integrating fuzzy logic and TOPSIS it will eliminate the uncertainty of the decision made [35]. Fuzzy TOPSIS has been successfully applied in several MCDM problems [3639].

In performing FTOPSIS, Chen in 2000 [36] underlined 9 steps to be followed. The steps of the algorithm in this multicriteria decision making tool are as follows.

Step 1. A committee of decision makers were formed to evaluate the alternatives based on the goal defined using the linguistic variables in Table 6.

tab6
Table 6: Linguistic variables for the importance weight of each criterion.

Step 2. Appropriate linguistic variables for the ratings of the alternative with respect to criteria were chosen and these linguistic ratings used are taken from Table 7 [36].

tab7
Table 7: Linguistic variables for the ratings.

Step 3. The importance weights of criteria and the ratings of three candidates are converted into fuzzy values. In order to obtain the fuzzy decision matrix and fuzzy weights of three alternatives, average value of each criteria for 3 decision makers was taken.

Step 4. The next step of the FTOPSIS analysis is the normalization of fuzzy decision matrix. This step is performed by taking each respective value and divide it with the maximum number of the particular criteria for all the 3 candidates.

Step 5. Construction of fuzzy weighted normalized decision matrix was then generated. In order to generate these values, values in Step 4 were divided by the weight of the respective column of fuzzy weighted normalized decision matrix.

Step 6. Fuzzy positive ideal solution (FPIS) and fuzzy negative ideal solution (FNIS) were then defined based on the values of normalized positive and negative triangular fuzzy numbers.

Step 7. Distance of each alternative from FPIS and FNIS was then calculated. The distance of two fuzzy numbers is calculated in (17) as

Step 8. After complete computing the positive and negative distance, these values were then summed up to obtain the total distance measurement for positive values, , and negative values, .

Step 9. Finally, obtain the final ranking of the alternative closeness coefficient, CC. In order to achieve the CC, (18) is used as
As per quoted by Krohling and Campanharo in 2011 [35], the triangular fuzzy number used is very effective for solving decision making problem which involved uncertainty and imprecise judgement.

2.4. Fuzzy AHP Integrated with Fuzzy TOPSIS (FAHPiFTOPSIS)

Integration between fuzzy AHP and fuzzy TOPSIS is believed to be able to make decision making more practical and reliable for decision makers due to its human-like thinking capability. For FAHPiFTOPSIS, the process of decision making comprises of 9 steps [2, 30].(1)Construct pair-wise comparison matrices among all the elements or criteria in the dimensions of the hierarchy system and convert it into fuzzy.(2)Define the fuzzy geometric mean as (3)Define the fuzzy weights as (4)Aggregating the ranking of criteria and suppliers’ performance criteria ranking as where (5)Suppliers’ performance as where (6)Normalizing the fuzzy decision matrix as where (7)Weighted normalization of the fuzzy decision matrix as where (8)Distance to positive and negative ideal solution using vertex method [36]. Let and between two triangular fuzzy numbers is as The best level of solution denoted by and the worst level of solution denoted by are defined in (30) as where The distances ( and ) between suppliers from and can be achieved by the area compensation method which is shown in the following equations: (9)Closeness coefficient as

2.5. Summary of Advantages and Disadvantages of MCDM Models

All the 5 models mentioned above are basically focusing on the theoretical part of the decision making. In this subsection the models are compared with respect to the differences. Based on Table 8, every model has its advantages. However, there is one common advantage which caters for all 5 models which is the effectiveness to handle qualitative and quantitative data.

tab8
Table 8: Summary of advantages of MCDM models.

Table 8 summarized the advantages of 5 models of focus in this study. For FAHP, AHP, and FAHPiFTOPSIS, these models are able to decompose complex problems into hierarchy which simplify the process of decision making and reduce the biasness by checking the inconsistency of rating. This is different for TOPSIS and FTOPSIS whereby the weights defined by decision makers are directly inserted into the models without any pairwise comparison. Models incorporated with fuzzy logic such as FAHP, FTOPSIS, and FAHPiFTOPSIS can adopt linguistic variables which convert expression of words to fuzzy numbers. FAHPiFTOPSIS is a versatile MCDM which adopted the decomposed hierarchy system of FAHP and the best method of ranking issue by FTOPSIS.

Referring to Table 9, it analysed the disadvantages of every model. These disadvantages can be compensated by combining the model with another MCDM model, for instance FAHPiFTOPSIS whereby FAHP is integrated with FTOPSIS.

tab9
Table 9: Summary of disadvantages of MCDM models.

Since FAHPiFTOPSIS integrated with FAHP and FTOPSIS, there are more steps that need to be performed. When comparing FAHP to FTOPSIS, FAHP requires more complex computation which includes pairwise comparison. When the number of alternatives and criteria increased, pairwise comparison becomes cumbersome. For AHP and FAHP besides the advantage of decomposing the problems, this might also lead to a large and lengthy process. AHP and TOPSIS are not integrated; hence, the models will not be able to reflect human thinking style.

3. Research Methodology

Companies involved in this study are automotive supply companies, Malaysian automotive manufacturers, and foreign automotive manufacturing plants operating in Malaysia together with the cooperation of four major tyre manufacturers in Malaysia. The process of the study started with the determination of the important criteria and their ranking which was made through surveys conducted on supply companies for automotive manufacturers. The data and information collected were analysed in depth and compared against traditional criteria.

Five methods of MCDM which are AHP, FAHP, TOPSIS, FTOPSIS, and an integrated FAHPiFTOPSIS model will be investigated. The effectiveness of all the five MCDM models will be compared. This study will be conducted through questionnaires distributed to Malaysian automotive manufacturers and foreign companies which supply preassembled parts to automotive manufacturers. The respondents of this study comprise 4 major tyre manufacturers and 12 automotive manufacturing plants.

3.1. Phase 1: Data Collection

This phase starts with the design of the data collection protocol, which concerns data to be collected and how to collect the data. The data used in this paper are collected and obtained from several sources such as questionnaires, face-to-face interviews, and phone interviews. The questionnaire aimed at identifying the priorities of various criteria highlighted in selecting suppliers as attached in Appendix C. In this phase, the interviews have been conducted by visiting the selected automotive industry. Phone interviews were also conducted whenever necessary. The purpose of this interview is to have a clearer picture on how supplier selection process is performed and at the same time to clarify any doubts regarding the questionnaires answered earlier. Nine main criteria were identified by the author for supplier selection process in automotive industry based on initial interview conducted with one of the major local car manufacturers. These criteria are shown in Appendix A.

3.2. Phase 2: Analyzing and Comparing Supplier Selection Models Using AHP, FAHP, TOPSIS, FTOPSIS, and FAHPiFTOPSIS

In this phase, real quantitative data are used. After identifying the criteria which are critical and important in selecting suppliers, the next stage of the project will be structuring the existing AHP, FAHP, TOPSIS, FTOPSIS, and FAHPiFTOPSIS models for supplier selection using data obtained from the questionnaire. Results obtained from the survey which consists of goal, criteria, subcriteria, and alternatives are used in structuring the models.

The models used for AHP and TOPSIS are from the founders themselves which are Saaty [8] and Hwang and Yoon 1981 [11], whereas for FAHP, the calculation method applied in this research is based on Chang’s approach [32]. Chang introduced a new approach for handling the fuzzy AHP—the use of TFNs for a pairwise comparison scale of the fuzzy AHP and the use of the extent analysis method for the synthetic extent values of the pairwise comparisons. Chang’s approach is one of the most popular approaches in the fuzzy AHP field. In this work, Chang’s extent analysis method is preferred whereby the steps of this approach are relatively easier than the other fuzzy AHP approaches and are similar to the conventional AHP. Considering Buckley [40], the model developed utilizes trapezoidal calculation and is tremendous. For FTOPSIS, a general extension of TOPSIS in fuzzy environment has been developed by Chen 2000 [36] which is the model utilized in this study, whereas for FAHPiFTOPSIS the model used is by Sun 2010 [30].

Once these five models are structured, the results produced by the models are then analysed. The purpose of this analysis is to determine the qualities, shortcomings, and biasness of the models. For AHP, FAHP, and FAHPiFTOPSIS, bias in decision making can be eliminated due to its flexibility to check the inconsistency and ability to decompose problems into hierarchies [21]. To determine which model has the less variation which means less bias in selection process, computation of variation percentage between models is performed in the analysis section [41, 42].

3.3. Phase 3: Testing and Validation of the Models

The final step of this paper is to validate the models using existing data collected from the questionnaire. This is to prove the consistency of the results for the five models with the results of the supplier selection of in-house procedures of the companies under study. The results of the models are then analyzed with the current practice of the selected companies. The main purpose of this step is to test if the results are consistent with the models and also to observe the best method relevant to the companies in obtaining the best supplier for automotive industry.

The system architecture example of the models is illustrated in Appendix B which involves one decision maker, decision maker 1 (DM1), and three potential alternatives, supplier 1 (S1), supplier 2 (S2), and supplier 3 (S3) with 9 criteria adopting the triangular values of fuzzy. From the figure, it can be seen that the system has three major components which areA”: selection of criteria ranking and supplier’s performance,B”: interpretation of data,C”: results.

In the first component (A), decision maker is required to pairwise compare the suppliers’ performance with respect to the criteria. Further ranking of the criteria based on its importance towards the goal is then performed. Accordingly, performance of the suppliers based on the criteria is then selected. In the second component (B), the interpretation of data takes place where all the data collected from the survey are inserted as input for all the models. All the calculation of analysis is done using Microsoft Excel 2010. The pair-wise comparison and ranking utilized the linguistic variable method which are variables with values expressed by words in a natural language [30]. All the data inputs are selected based on the linguistic variable table provided. Component “C” is the final major component of the example in obtaining the best supplier according to the desired criteria.

3.4. Application in Automotive Industries

In Tables 10 and 11, the data tabulated are the number of respondents and level of importance for criteria 1 to 9 responded by 12 major automotive manufacturers and 4 major tyre suppliers in Malaysia.

tab10
Table 10: Weight and level of importance of criteria 1 to 9 of 12. Major automotive manufacturers.
tab11
Table 11: Weight and level of importance of nine (9) criteria of 4 major tyre suppliers.

4. Results and Discussion

Upon analyzing the various decision making tools described in Section 3, Figure 5 is presented. The figure illustrates the overall ranking for 5 different decision making tools used for supplier selection process in automotive industries using the results tabulated in Table 12.

tab12
Table 12: Actual scores of Supplier 1, 2, 3, and 4 according to 5 decision making tools.
841584.fig.005
Figure 5: Summary of various decision making tools based on actual score.

Referring to Table 12, these are the values of actual scores for Supplier 1, 2, 3, and 4 according to the decision making tools applied. From this table, the values are then normalized as in Table 13. The equation used for normalization is presented in (34) below. The normalised scores in Table 13 were then used to plot the bar graph of Figure 6. Normalized values are used to make values comparable with each other. It is a process of reducing the values into a standard scale [43].

tab13
Table 13: Normalised scores of Supplier 1, 2, 3, and 4 according to 5 decision making tools.
841584.fig.006
Figure 6: Summary of various decision making tools based on normalised priority.

Generally, it can be observed from Figure 6 that the ranking of supplier selection process for each different decision making tool followed the order of S3 > S4 > S2 > S1 where Supplier 3 exhibits the highest priority level while Supplier 1 exhibits the lowest priority level. The normalized score for Supplier 3 hits above average of 0.3, whereas for Supplier 1 and 2 the difference in normalized scoring is very small which indicates that both suppliers are competitively comparable.

The scores for the ranking among the four suppliers in each MCDM tool (AHP, FAHP, TOPSIS, and FTOPSIS) show significantly comparable variation. The percentage of variation scores for 4 suppliers is −36.9%, −40.33%, 26.8%, and 32.2% when comparing AHP, FAHP, TOPSIS, and FTOPSIS with FAHPiFTOPSIS, respectively. These variations are obtained from the average of the original scoring and then compared with respect to FAHPiFTOPSIS. Table 14 showed the variations of models with respect to each other. The computation can be performed as in (34) whereby the calculation for AHP percentage of variation is taken here, for example. The reliability of the model is based on the less variation of a model. Based on Table 14, when the variations are summed up for each MCDM model, FAHPiFTOPSIS achieved less variation, whereas the high variation model is FAHP for 3 computations with respect to FAHPiFTOPSIS, FTOPSIS, and TOPSIS

tab14
Table 14: Variations of models with respect to each other.

The following evaluation to determine the variation is based on the difference of scores achieved by the best supplier with the less preferred supplier, and Table 15 is generated. The data for Table 15 can be obtained from Table 13. Referring to Table 15 and Figure 7, the highest percentage goes to TOPSIS with scoring of 79.73% followed by FTOPSIS with percentage of 63.53 and FAHP with 29.49%. On the contrary, FAHPiFTOPSIS decision making tool demonstrated the lowest score variation of 22.42% followed by 26.84% of AHP. This indicates that FAHPiFTOPSIS is able to eliminate the “biasness” and ambiguity in supplier selection process.

tab15
Table 15: Difference of normalized scores for the best supplier (Supplier 3) and the less preferred supplier (Supplier 2).
841584.fig.007
Figure 7: Difference in percentage comparison between the best supplier and the less preferred supplier.

5. Conclusions

This research has successfully demonstrated the applicability of the MCDM tools (AHP, FAHP, TOPSIS, FTOPSIS, and FAHPiFTOPSIS) by automotive industry in the selection of suppliers. As a result, few points can be concluded as follows.(1)Six additional parameters were observed to have equal important criteria in automotive industry besides three classic criteria, namely, price, delivery, and quality which are support service, technology, factory capacity and capability, supplier background, flexibility, and other management system requirements.(2)AHP, FAHP, TOPSIS, FTOPSIS, and FAHPiFTOPSIS are all applicable and accurate for supplier selection in automotive industry. (a)For AHP and FAHP, when the number of suppliers and alternatives becomes big, more criteria will need to be considered and decision maker will face a challenge to perform pairwise comparison in a big matrix.(b)For TOPSIS and FTOPSIS, it is a simpler method whereby weights are needed as the input of this decision making tool irrespective with the number of alternatives.(c)The results from AHP, FAHP, TOPSIS, and FTOPSIS show great variation in the final ranking scores. However, FAHPiFTOPSIS decision making tool demonstrated the lowest score variation of 22.42%. This indicates that FAHPiFTOPSIS application is able to effectively eradicate any ambiguity or fuzziness in the process.

Appendices

A. Supplier Selection Criteria

See Table 16.

tab16
Table 16: Supplier selection criteria.

B. Generic System Architecture of FAHP, FTOPSIS, and FAHPiFTOPSIS for Decision Makers in Selecting the Best Supplier

See Figure 8.

841584.fig.008
Figure 8

C. Questionnaire Developed for the Study

C.1. Developed Questionnaire

The developed questionnaire is called:Supplier selection process for automotive industry.

C.1.1. Supplier Selection Criteria

Supplier selection evaluations are normally being conducted by companies in order to evaluate their suppliers’ performance and reliability. Traditionally, supplier selection evaluations are based on price/cost, quality, and delivery. However, in today’s management, most companies do not only consider the three primary criteria, but they also take into account technology, support service, flexibility, supplier background, management tool systems, and factory capacity and capability.

As part of my Master research in manufacturing engineering, I would like to seek you a great assistance to provide me your critical input on supplier selection. Below you will find the criteria listed. What you have to do is to fill in your inputs in each question by writing the numbers according to the ranking in the right box.

  Delivery/Lead Time. Delivery of supply: whichever is applied to your company, please indicate the degree of importance on the following delivery criteria as expected from your suppliers. See Table 17.

tab17
Table 17

  Support Service. See Table 18.

tab18
Table 18

  Quality Factor. Supplier selection is not only limited to cost, but also to the performance of other competitive dimensions of quality. Meeting customer requirements is one example, but there are other important factors involved. Please indicate the degree of importance on the following quality criteria as expected from your suppliers. See Table 19.

tab19
Table 19

  Technology. Suppliers have to cope with technology to be more competitive. Suppliers which are responsive to new technology are more favorable to any company. Please indicate the degree of importance on the following criteria. See Table 20.

tab20
Table 20

  Price/Cost. Traditionally, any company will go for the lowest cost from their suppliers. Consequently, the cost control system administered by any company may have some effect on the suppliers. The criteria below show some of the important issues related to cost and price. Please indicate their level of importance as applicable to your company. See Table 21.

tab21
Table 21

  Factory Capacity and Capability. See Table 22.

tab22
Table 22

  Supplier Background. The additional requirements criteria involve supplier’s background. Below are some of the important factors for suppliers. Please indicate their level of importance. See Table 23.

tab23
Table 23

  Flexibility. See Table 24.

tab24
Table 24

  Management System. See Table 25.

tab25
Table 25
C.1.2. Rating of Suppliers under Various Criteria

Please state your level of importance of the supplier with respect to the criteria on the right boxes regarding the numbers below. See Table 26.

tab26
Table 26
C.1.3. Pairwise Comparison of 9 Criteria

According to Saaty’s table, compare the 9 criteria accordingly and fill in the table. Criteria on the left side of the table need to be compared to the criteria at the horizontal columns. See Tables 27 and 28.

tab27
Table 27: Saaty’s pair-wise comparison table.
tab28
Table 28
C.1.4. Pairwise Comparison of Suppliers with Respect to 9 Criteria

According to Saaty’s table, compare the suppliers (S1, S2, S3, and S4) with respect to the criteria. Suppliers on the left side columns are to be compared with the suppliers on the right columns. Kindly mark (/) at the preferred value when pairwise comparison is performed. See Tables 29, 30, 31, 32, 33, 34, 35, 36, 37, and 38.

tab29
Table 29: Saaty’s pair-wise comparison table.
tab30
Table 30: Criterion 1: delivery/lead time.
tab31
Table 31: Criterion 2: support service.
tab32
Table 32: Criterion 3: quality.
tab33
Table 33: Criterion 4: technology.
tab34
Table 34: Criterion 5: price/cost.
tab35
Table 35: Criterion 6: factory capacity and capability.
tab36
Table 36: Criterion 7: supplier background.
tab37
Table 37: Criterion 8: flexibility.
tab38
Table 38: Criterion 9: management tool system.

Acknowledgments

The authors would like to thank Associate Professor Dr. Mohd. Khaled Omar and all the survey respondents for their kind support and help contributed towards the implementation of this paper.

References

  1. V. Vaidya, Impact of Thailand Floods on Automotive Industry and Supply Chain, Frost and Sullivan, 2011.
  2. C. T. Chen, C. T. Lin, and S. F. Huang, “A fuzzy approach for supplier evaluation and selection in supply chain management,” International Journal of Production Economics, vol. 102, no. 2, pp. 289–301, 2006. View at Publisher · View at Google Scholar · View at Scopus
  3. Malaysia Automotive Association, MAA, Summary of Sales & Production Data, 2011.
  4. A. Ahmad, Market Review of 2010 and Outlook for 2011, Malaysia Automotive Association, 2011.
  5. Secretariat, ASEAN Free Trade Area (AFTA) : An update, Association of Southeast Nation, 1999.
  6. Bernama, “DRB-Hicom's Pekan Plant to See Higher Production,” Bernama, 2009.
  7. C. Kahraman, U. Cebeci, and Z. Ulukan, “Multi-criteria supplier selection using fuzzy AHP,” Logistics Information Management, vol. 16, pp. 382–394, 2003.
  8. T. L. Saaty, The Analytical Hierarchy Process, McGrawHill, New York, NY, USA, 1980.
  9. T. L. Saaty, “Axiomatic foundation of the analytical hierarchy process,” Management Science, vol. 32, no. 7, pp. 841–855, 1986.
  10. O. Cakir and M. S. Canbolat, “A web-based decision support system for multi-criteria inventory classification using fuzzy AHP methodology,” Expert Systems with Applications, vol. 35, no. 3, pp. 1367–1378, 2008. View at Publisher · View at Google Scholar · View at Scopus
  11. C. L. Hwang and K. Yoon, Multiple Attribute Decision Making-Methods and Applications, Springer, Heidelberg, Germany, 1981.
  12. L. De Boer, L. Van Der Wegen, and J. Telgen, “Outranking methods in support of supplier selection,” European Journal of Purchasing and Supply Management, vol. 4, no. 2-3, pp. 109–118, 1998. View at Scopus
  13. J. Wang, K. Fan, and W. Wang, “Integration of fuzzy AHP and FPP with TOPSIS methodology for aeroengine health assessment,” Expert Systems with Applications, vol. 37, no. 12, pp. 8516–8526, 2010. View at Publisher · View at Google Scholar · View at Scopus
  14. L. A. Zadeh, “Fuzzy sets,” Information and Control, vol. 8, no. 3, pp. 338–353, 1965. View at Scopus
  15. M. Negnevitsky, Artificial Intelligence: A Guide To Intelligent Systems, Addison-Wesley: Pearson Education Limited, Essex, UK, 2002.
  16. M. Bevilacquaa, F. E Ciarapicab, and G. Giacchettab, “A fuzzy-QFD approach to supplier selection,” Journal of Purchasing & Supply Management, vol. 12, pp. 14–27, 2006.
  17. H. Deng, “Multicriteria analysis with fuzzy pair-wise comparison,” International Journal of Approximate Reasoning, vol. 21, pp. 215–231, 1999.
  18. S. M. Ordoobadi, “Development of a supplier selection model using fuzzy logic,” Supply Chain Management, vol. 14, no. 4, pp. 314–327, 2009. View at Publisher · View at Google Scholar · View at Scopus
  19. T. L. Saaty, “A scaling method for priorities in hierarchical structures,” Journal of Mathematical Psychology, vol. 15, no. 3, pp. 234–281, 1977. View at Scopus
  20. C. C. Yang and B. S. Chen, “Supplier selection using combined analytical hierarchy process and grey relational analysis,” Journal of Manufacturing Technology Management, vol. 17, no. 7, pp. 926–941, 2006. View at Publisher · View at Google Scholar · View at Scopus
  21. R. L. Nydick and R. P. Hill, “Using the analytic hierarchy process to structure the supplier selection procedure,” International Journal of Purchasing and Materials Management, pp. 31–36, 1992.
  22. I. Ertuğrul and N. Karakaşoğlu, “Comparison of fuzzy AHP and fuzzy TOPSIS methods for facility location selection,” International Journal of Advance Manufacturing Technology, vol. 39, pp. 783–795, 2008.
  23. F. Tüysüz and C. Kahraman, “Project risk evaluation using a fuzzy analytic hierarchy process: an application to information technology projects,” International Journal of intelligent Systems, vol. 21, pp. 559–584, 2006.
  24. A. Kelemenis, K. Ergazakis, and D. Askounis, “Support managers' selection using an extension of fuzzy TOPSIS,” Expert Systems with Applications, vol. 38, no. 3, pp. 2774–2782, 2011. View at Publisher · View at Google Scholar · View at Scopus
  25. E. Bottani and A. Rizzi, “A fuzzy TOPSIS methodology to support outsourcing of logistics services,” Supply Chain Management, vol. 11, no. 4, pp. 294–308, 2006. View at Publisher · View at Google Scholar · View at Scopus
  26. H. J. Shyur, “COTS evaluation using modified TOPSIS and ANP,” Applied Mathematics and Computation, vol. 177, no. 1, pp. 251–259, 2006. View at Publisher · View at Google Scholar · View at Scopus
  27. M. F. Shipley, A. de Korvin, and R. Obid, “A decision making model for multi-attribute problems incorporating uncertainty and bias measures,” Computers and Operations Research, vol. 18, no. 4, pp. 335–342, 1991. View at Scopus
  28. O. Durán and J. Aguilo, “Computer-aided machine-tool selection based on a Fuzzy-AHP approach,” Expert Systems with Applications, vol. 34, no. 3, pp. 1787–1794, 2008. View at Publisher · View at Google Scholar · View at Scopus
  29. T. L. Saaty, Multi-Criteria Decision Making: the Analytic Hierarchy Process-Planning, Priority Setting, Resource Allocation, R W S Publications, Pittsburgh, Pa, USA, 1996.
  30. C. C. Sun, “A performance evaluation model by integrating fuzzy AHP and fuzzy TOPSIS methods,” Expert Systems with Applications, vol. 37, no. 12, pp. 7745–7754, 2010. View at Publisher · View at Google Scholar · View at Scopus
  31. P. J. M. van Laarhoven and W. Pedrycz, “A fuzzy extension of Saaty's priority theory,” Fuzzy Sets and Systems, vol. 11, no. 3, pp. 229–241, 1983. View at Scopus
  32. D. Y. Chang, Extent Analysis and Synthetic Decision, Optimization Techniques and Applications, vol. 1, World Scientific, Singapore, 1992.
  33. A. T. Gumus, “Evaluation of hazardous waste transportation firms by using a two step fuzzy-AHP and TOPSIS methodology,” Expert Systems with Applications, vol. 36, no. 2, pp. 4067–4074, 2009. View at Publisher · View at Google Scholar · View at Scopus
  34. D. L. Olson, “Comparison of weights in TOPSIS models,” Mathematical and Computer Modelling, vol. 40, no. 7-8, pp. 721–727, 2004. View at Publisher · View at Google Scholar · View at Scopus
  35. R. A. Krohling and V. C. Campanharo, “Fuzzy TOPSIS for group decision making: a case study for accidents with oil spill in the sea,” Expert Systems with Applications, vol. 38, no. 4, pp. 4190–4197, 2011. View at Publisher · View at Google Scholar · View at Scopus
  36. C. T. Chen, “Extensions of the TOPSIS for group decision-making under fuzzy environment,” Fuzzy Sets and Systems, vol. 114, no. 1, pp. 1–9, 2000. View at Scopus
  37. T. Y. Chen and C. Y. Tsao, “The interval-valued fuzzy TOPSIS method and experimental analysis,” Fuzzy Sets and Systems, vol. 159, no. 11, pp. 1410–1428, 2008. View at Publisher · View at Google Scholar · View at Scopus
  38. Y. M. Wang and T. M. S. Elhag, “Fuzzy TOPSIS method based on alpha level sets with an application to bridge risk assessment,” Expert Systems with Applications, vol. 31, no. 2, pp. 309–319, 2006. View at Publisher · View at Google Scholar · View at Scopus
  39. D. Yong, “Plant location selection based on fuzzy TOPSIS,” International Journal of Advanced Manufacturing Technology, vol. 28, no. 7-8, pp. 839–844, 2006. View at Publisher · View at Google Scholar · View at Scopus
  40. J.-J. Buckley, “Fuzzy hierarchical analysis,” Fuzzy Sets and Systems, vol. 17, no. 1, pp. 233–247, 1985.
  41. U. Nirmal, B. F. Yousif, D. Rilling, P. V. Brevern, and N. Jamil, “The potential of using treated betelnut fibres as reinforcement for tribo-bio polymeric composites subjected to dry/wet contact conditions,” in Proceedings of the International Conference on Natural Polymers (ICNP '10), Kottayam, Kerala, India, September 2010.
  42. U. Nirmal, N. Singh, J. Hashim, S. T. W. Lau, and N. Jamil, “On the effect of different polymer matrix and fibre treatment on single fibre pullout test using betelnut fibres,” Materials and Design, vol. 32, no. 5, pp. 2717–2726, 2011. View at Publisher · View at Google Scholar · View at Scopus
  43. S. Borgatti, Research Method, BA 762, May 2012.