Anda di halaman 1dari 10

An AHP/DEA Methodology for 3PL Vendor Selection in 4PL

He Zhang1, Xiu Li1, and Wenhuang Liu2


1

National Engineering Research Center for CIMS, Dept. of Automation, Tsinghua University, 100084 Beijing, P.R. China zhanghe98@mails.tsinghua.edu.cn, lixiu@cims.tsinghua.edu.cn 2 Graduate School at Shenzhen, Tsinghua University, 518055 Shenzhen, China liuwh@sz.tsinghua.edu.cn

Abstract. Vendor selection is an important and complex problem, which contains many criteria. The analytic hierarchy process (AHP) can be very useful in reaching a likely result which can satisfy the subjective opinion of the decision maker or the evaluation team. On the other hand, the Data Envelopment Analysis (DEA) can select vendors objectively with the quantitative data. In this paper, a four-step model based on both AHP and DEA is formulated and applied to a case study. The use of the proposed model can give precise evaluation combining the subjective opinion from the decision makers with the objective data of the relevant factors. In other words, this approach can add subjective factors to the evaluation without losing the objective precision of the selection.

1 Introduction
Nowadays, in order to survive in increasing competitions, suppliers try to find better system design, logistics process management, data collection, and storage. They are outsourcing their entire set of supply chain processes from a single organization fourth party logistics (4PL). Zhang [1] presented a framework of the decision support system of 4PL, as shown in Figure 1. We can find that 3PL vendor selection is an important problem because 4PL is established upon 3PL vendors. There have been some reported research efforts focusing on 3PL vendor selection. Aghazadeh [2] developed a five-step method to choose an effective 3PL provider. He also presented four relevant criteriasimilar value, information technology systems, key management, and relationship. Knemeyer and Murphy [3] evaluated the performance of 3PL in a relationship marketing perspective. Six relationship dimensions of trust, communication, opportunistic behavior, reputation, satisfactory prior interactions, and relationship-specific investments are used to form a model to evaluate 3PL vendors. Meade and Sarkis [4] established a conceptual model for selecting and evaluating third-party reverse logistics provider with AHP. Menon et al. [5] analyzed the relevant criteria of selecting 3PL logistics services. Nine factors were gathered and divided into four groups. Yan et al. [6] developed a model of decision support system based on case-based reasoning for 3PL evaluation.
W. Shen et al. (Eds.): CSCWD 2005, LNCS 3865, pp. 646 655, 2006. Springer-Verlag Berlin Heidelberg 2006

An AHP/DEA Methodology for 3PL Vendor Selection in 4PL

647

Information management 3PL vendor management Management of the communication between customers and vendors 4PL information system Evaluation and optimization Optimization and selection of the logistics schemes 3PL vendor selection Storage management Decision management Finance management

Fig. 1. A framework of 4PL decision support system

Although all these efforts developed their methods for selecting or evaluating 3PL vendors, some requirements cannot be satisfied. At first, most methods are qualitative. The usual way that they make their evaluations is to list all the criteria in a form and ask the decision makers to give their evaluations for each criterion. In this paper, a quantitative method is presented to solve this problem. Analytic hierarchy process (AHP) was designed to solve complex problems involving multiple criteria [7,8]. It allows decision makers to specify their preference using a simple scale, which can be very useful in helping a group or an individual to make a synthetic decision. Narasimhan [9], Nydick and Hill [10], and Partovi et al. [11] suggested the use of AHP for vendor selection. They suggested AHP as the proper approach because of its inherent capacity to handle qualitative and quantitative criteria used in vendor selection problems. The hierarchical structure used in formulating the AHP model can enable all the members of the evaluation team to visualize the problem systematically in terms of relevant criteria and sub-criteria. On the other hand, the AHP model may produce a result affected by subjective attitudes of the decision makers greatly rather than the quantitative data, because of the importance of the views given by the decision makers. During the past two decades, Data Envelopment Analysis (DEA) has emerged as an important tool in the field of efficiency measurement. DEA is a nonparametric approach that does not require any assumption about the functional form of production function [12]. Weber [13] and Liu et al [14] suggested evaluating vendors with the use of DEA, which can avoid the subjective factors of decision makers. From what has been presented above, we can make a conclusion that DEA is a quantitative method to evaluate vendors. There have been several previous attempts in the literature to tie between AHP and DEA. Stern et al. [15] presented a two-stage model which combined AHP and DEA. However, these approaches actually use AHP and DEA separately and have the limitations of controlling the subjective factors in AHP.

648

H. Zhang, X. Li, and W. Liu

This paper proposes a model for 3PL vendor selection in 4PL, containing several relevant criteria and sub-criteria. This approach works within the framework of DEA; basically, it advanced the DEA analysis with AHP. We first present the model AHP/DEA, which combines DEA and AHP. Afterwards, the AHP/DEA model will be applied to a case study to show the improvement of this approach. After analyzing the results, the relevant advantages and disadvantages of this model are given to show the merits and limitations of this model.

2 The AHP/DEA Model


2.1 Relevant Theory The AHP Theory. The AHP [7] is designed for subjective evaluations based on multiple criteria, which are organized in a hierarchical structure. The goal, criteria and sub-criteria are placed in higher levels and each alternative in the lower levels are evaluated by each criterion. After establishing the AHP model, the opinion of the relevant decision makers is gathered to create several pairwise comparison matrices in which their subjective judgment of each pair of items is accessed. With these pairwise comparison matrices, we should determine the normalized weighs by using some proper approaches, such as sum-approach [11]. Then these normalized priority weights can be combined to synthesize the solution of the vendor selection problem. The DEA Theory. As a non-parametric approach, DEA assumes that there are n decision-making units (DMU), among which each one consumes various amount of m different inputs to produce s different outputs. Based on these two sets of multiple criteria, DEA deals with classifying the DMUs into two categories, efficient and inefficient. This efficient frontier is determined by the most efficient DMUs under study, based on the notion of Pareto optimality. This concept states that a specific DMU is efficient if and only if the performance of other DMUs does not show that some of its inputs or outputs can be improved without worsening some of its other inputs or outputs. Conversely, a DMU is said to be Pareto inefficient if the performance of other DMUs is able to show that some of its inputs or outputs can be improved without worsening some of its other inputs or outputs. Algebraic model like what has been discussed can be framed with the given inputs and outputs [7]. In this paper, an improved mode, model [15], is used to calculate the efficiency. The C 2 R model is shown as the following:
max T y0 = VP T T s.t. x y j 0 ( P) T j x0 = 1 0 , 0 j = 1,L n
xi = ( x1i ,L , xmi )T is the input

, are the weight vector

yi = ( y1i ,L , ysi )T is the output

2.2 AHP/DEA Methodology The methodology solves the problem in a procedure, as illustrated in Figure 2.

An AHP/DEA Methodology for 3PL Vendor Selection in 4PL


Establishment of system of criteria
AHP for qualitative data evaluation DEA for pairwise comparison judgment matrices (PCJM) in AHP AHP for final evaluation

649

Fig. 2. Procedure of the AHP/DEA methodology

2.3 Establishment of System of Criteria From a survey which is conducted in order to identify the relevant evaluation criteria, we identified the criteria shown below: General Company Considerations - Price (PR) - Financial stability (FS) - Experience in the same industry or with similar companies (EX) - Location (LO) - International scope (IS) Capabilities Information systems and technology capabilities (ISTC) Customer service (CS) Capacity to accommodate and grow the clients business (CB) Flexibility to handle unique requirements (FHR) Responsiveness to unexpected problems (RUP) The ability to meet or exceed promises (AMP) Quality Service quality and performance, e.g. Six Sigma, ISO 9000 (SQP) Commitment to continuous improvement (CCI) Quality of the providers management team (QPM) Client Relationship Availability (AV) Subjective feel between the partners (SBP) Service (SER) Labor relations Human resource policies (HRP) Availability of qualified talent (AQT) The above identified factors are considered as the relevant elements that are used to formulate an appropriate system of criteria for 3PL vendor selection. All the correspondent evaluations are shown in Table 1.

650

H. Zhang, X. Li, and W. Liu Table 1. Correspondent evaluations of all the criteria Criteria PR FS EX LO IS ISTC CS CB FHR RUP AMP SQP CCI QPM AV SBP SER HRP AQT Evaluation Actual prices Qualitative evaluation Qualitative evaluation Qualitative evaluation Qualitative evaluation Qualitative evaluation Percentage of customers satisfaction Qualitative evaluation Qualitative evaluation Qualitative evaluation Percentage of untimely jobs Qualitative evaluation Qualitative evaluation Qualitative evaluation Qualitative evaluation Qualitative evaluation Qualitative evaluation Qualitative evaluation Qualitative evaluation

2.4 AHP for Qualitative Data Evaluation Obviously, there are many criteria, such as FS and EX that cannot be evaluated with quantitative data directly and their evaluations usually are qualitative. In this step, all the qualitative evaluations should be transferred to quantitative values. The purpose of the AHP is to provide a vector of weights expressing the relative importance of those 3PL alternatives for each qualitative criterion. AHP requires four steps: (1) structuring the hierarchy of criteria and alternatives for evaluation; (2) assessing the decision-makers evaluations by pairwise comparisons; (3) using the eigenvector method to yield priorities for criteria and for alternatives by criteria; and (4) synthesizing the priorities of the alternatives by criteria into composite measures to arrive at a set of ratings for the alternatives. The scale of importance is defined in Table 2 according to Satty 1-9 scale [7] for pairwise comparisons.
Table 2. Correspondent evaluations of all the criteria Intensity of important 1 3 5 7 9 2, 4, 6, 8 Definition Equal importance Moderate importance Strong importance Very strong Extreme importance For compromise between the above values

An AHP/DEA Methodology for 3PL Vendor Selection in 4PL

651

2.5 DEA for Pairwise Comparison Judgment Matrices (PCJM) in AHP When evaluating alternatives with AHP, the number of the relevant sub-criteria for one goal should be less than 10, because of the burden of computation. In this step, all the quantitative data should be quantified with DEA for the establishment of PCJM which will be used in the final step. Let j be the correspondent 3PL vendor

j ( j = 1, 2,L , n) and all the criteria can be divided into two groups:

xij (i = 1, 2,L , m) much smaller, much better. Criteria which present the input factors in DEA.

yrj (r = 1, 2,L , s ) much larger, much better. Criteria which present the output
factors in DEA. Thus all the above criteria are divided into two groups as shown in Table 3.
Table 3. Inputs and outputs of DEA Group Input Output Relevant Criteria PR, AMP FS, EX, LO, IS, ISTC, CS, CB, FHR, RUP, SQP, CCI, AV, SBP, SER, HRP, QPM, AQT

Zhu et al. [16] proposed and developed a DEA method with preference. In all the criteria presented above, the decision-makers often have an order of preference. For any two 3PL vendors i and j , the relevant efficiency Eij and E ji are calculated as the following models:
max Eij = T yi max E ji = T y j T T T T s.t. y j x j 0 s.t. yi xi 0 T T T T xi = 1, yi 1 x j = 1, y j 1 Pij Pji k , k = 1,L , s q k , k = 1,L , s q k +1 k 0, k = s q + 1,L , s 1 k +1 k 0, k = s q + 1,L , s 1 s s

( xi : Input;

yi : Output; , : Weight vector; : Positive infinitesimal) In the above equations, there is a priority order in the last q output factors:
us q +1 us q + 2 L u s

Then the ratio of the relevant efficiency can be calculated as aij = Eij E ji ,

a ji = E ji Eij , aii = 1 . All the relevant pairwise comparisons can be calculated with
this DEA method.

652

H. Zhang, X. Li, and W. Liu

2.6 AHP for Final Evaluation In this step, all the pairwise comparisons calculated in Section 2.5 will be used to form a PCJM in order to perform the final evaluation with AHP. With all the pairwise comparisons, the PCJM for AHP evaluation is formed as following:

a12 1 1 a 1 A = 12 M M 1 a1n 1 a2 n

L a1n L a2 n O M L 1

Then PCJM A should be used to calculate the maximum eigenvalue

max

and the

eigenvector w . So the ranking of all the alternatives can give the decision-makers useful reference information.

3 Case Study
First we consider a problem of selecting a proper vendor in four 3PL vendors whose names are A, B, C and D ( n = 4 ). This problem will be solved by the approach presented above. The data of all the criteria of the four 3PL vendors are shown in Table 4. All the qualitative evaluations have been transferred to quantitative data by AHP during the process presented in Section 2.4.
Table 4. Data of the four 3PL vendors Criteria PR FS EX LO IS ISTC CS CB FHR RUP AMP SQP CCI QPM AV SBP SER HRP AQT Vendor A 500 0.2736 0.4954 0.1469 0.5086 0.3954 95% 0.3472 0.1563 0.1879 5% 0.0729 0.4605 0.125 0.0813 0.4715 0.2242 0.1815 0.1055 Vendor B 550 0.1120 0.2125 0.0854 0.1209 0.1225 75% 0.1423 0.4078 0.2035 10% 0.1350 0.1340 0.125 0.1544 0.1083 0.2242 0.0556 0.1860 Vendor C 1200 0.4915 0.1001 0.3851 0.2668 0.3676 98% 0.3829 0.0781 0.0606 3% 0.2521 0.1074 0.5 0.4758 0.1653 0.0698 0.4589 0.4720 Vendor D 450 0.1229 0.1920 0.3827 0.1036 0.1145 80% 0.1276 0.3577 0.5480 9% 0.5400 0.2981 0.25 0.2884 0.2550 0.4818 0.3041 0.2365

An AHP/DEA Methodology for 3PL Vendor Selection in 4PL

653

Considering there is a priority order as presented above:

uCS u EX u SER u FS others ,


With the equations P and Pji , the relevant PCJM A can be calculated as: ij
1.0000 0.8541 A= 1.0000 1.0000 1.1708 1.0000 1.0000 1.0000 1.0000 0.8732 1.0000 1.0000 1.0000 1.1452 1.0000 1.0000

The eigenvector of matrix A can be calculated as the vector w : T w = [ 0.2598,0.2321,0.2498,0.2582] . Thus we can conclude that 3PL vendor A is better than 3PL vendor D, 3PL vendor D is better than 3PL vendor C and 3PL vendor C is better than 3PL vendor B.

4 Remarks
In Section 2.3, all the correspondent criteria are made by various logistics surveys and some practical projects. When there is a specific 3PL vendor selection problem, all these criteria can be selected according to the practical requirement of the decisionmakers. In Section 2.4, when using AHP to quantify all the qualitative evaluations, the opinions of the experts should be examined by the consistency ratio (CR) of each PCJM, which should be compared with the rule-of-thumb value of C.R (RCR). If the calculated CR is well below the corresponding RCR, it clearly implies that the decision maker is consistent in assigning pairwise comparison judgments. Otherwise, the PCJMs are invalid and should be reassigned by the decision maker [7]. When there are many selected 3PL vendors, using triangle questionnaire, improved nine-point scale and reformative pairwise comparison judgment matrices can improve the precision of this approach based on the AHP model [17]. In Section 2.5, the priority order is designed according to the opinion of the decision-makers. It should be pointed out that, if there are no priority orders, the final evaluation may not be performed normally. That is to say that, if there is no difference among the priorities of all these criteria, a vendor, who is the best in an unimportant criterion, may be evaluated in the same way as others. So decision-makers should choose their focused factors in order to evaluate effectively.

5 Conclusion
As explained in Section 1, 3PL vendor selection is an important problem. We first identify the relevant criteria for selecting a 3PL vendor. Then a procedure is established and this methodology is generally effective to a 3PL vendor selection problem. After ascertaining all the criteria, AHP is used to quantify all the qualitative evaluations and all these quantitative data are used to establish the final PCJM by DEA.

654

H. Zhang, X. Li, and W. Liu

Then AHP is used again to find the global priority weights, which represent the final evaluations of all the alternatives. The proposed AHP/DEA model is applied to a 3PL vendor selection problem. In this case, we got the quantitative evaluations of all the 3PL vendors by using this vendor selection process. With the AHP/DEA model, the relevant subjective evaluations can be transferred to quantitative values and this assessing process is more objective than the evaluation only with AHP. On the other hand, this method can also combine the subjective opinions of the decision-makers with quantitative data. Thus both the subjective opinions and quantitative data can be considered in evaluation at the same time. However, it should be noted that the computational burden would be increased with the increase in the number of criteria, as well as the number of vendors considered in the selection. This is one of the reasons that we suggested short-listing the number of vendors first and then applying the AHP/DEA model.

Acknowledgement
The work presented in this paper is supported by the National Science Foundation of China (Grant No. 70202008).

References
1. Zhang, H., Cheng, J.-C.K., Liu, W., Li, X.: Appraisement of Transporters in Fourth Party Logistics. Industrial Engineering Journal, 7(3) (2004) 3639 2. Aghazadeh, S.-M.: How to Choose an Effective Third Party Logistics Provider. Management Research News, 26(7) (2003) 5058 3. Knemeyer, A.M., Murphy, P.R.: Evaluating the Performance of Third-Party Logistics Arrangements: A Relationship Marketing Perspective. Journal of Supply Chain Management, 40(1) (2004) 3551 4. Meade, L., A.M., Sarkis, J.: A conceptual model for selecting and evaluating third-party reverse logistics. Journal of Supply Chain Management, 7(5) (2002) 283295 5. Menon, M.K, McGinnis, M.A., Ackerman, K.B.: Selection criteria for providers of thirdparty logistics services: An exploratory study. Journal of Business Logistics, 19(1) (1998) 121137 6. Yan, J., Chaudhry, P.E., Chaudhry, S.S.: A model of a decision support system based on case-based reasoning for third-party logistics evaluation. Expert Systems, 20(4) (2003) 196208 7. Satty, T.L.: The analytic hierarchy process. McGraw-Hill, New York (1980) 8. Satty, T.L., Vargas, L.G.: Decision making in economic, political, social, and technological environments with the analytic hierarchy process. RWS Publication, Pittsburgh (1994) 9. Narasimahn, R.: An analytical approach to supplier selection. Journal of Purchasing and Materials Management, 19(4) (1983) 2732 10. Nydick, R.L., Hill, R.P.: Using the analytic hierarchy process to structure the supplier selection procedure. Journal of Purchasing and Materials, 25(2) (1992) 3136 11. Partovi, F.Y., Burton, J., Banerjee, A.: Application of analytic hierarchy process in operations management. International Journal of Operations and Production Management, 10(3) (1989) 519

An AHP/DEA Methodology for 3PL Vendor Selection in 4PL

655

12. Shen, Z, Zhu, Q., Wu, G.: Theory, methodology and application of DEA. Science Publication, Beijing (1996) 13. Weber, Charles, A.: A Data Envelopment Analysis Approach to Measuring Vendor Performance. Supply Chain Management, 1(1) (1996) 2830 14. Liu, J., Ding, F., Lall, V.: Using data envelopment analysis to compare suppliers for supplier selection and performance improvement. Supply Chain Management, 5(3) (2000) 143150 15. Stern, Z.S., Mehrez, A., Hadad, Y.: An AHP/DEA methodology for ranking decision making units. International Transactions In Operational Research, 7(2) (2000) 109124 16. Zhu, Q., Shen, Z., Xu, N.: DEA method with preference. Journal of Industrial Engineering Management, 9(2) (1995) 112116 17. Zhang, Q., Nishimura, T.: Some Methods of Raising Exactness of Evaluation in AHP. Systems Engineering-Theory & Practice, 17(11) (1997) 2936

Anda mungkin juga menyukai