The Journa of Mathematics and Computer Science Avaiabe onine at http://www.tjmcs.com The Journa of Mathematics and Computer Science Vo.2 No.2 (2011) 311-318 Performance Measurement Using Fuzzy Logic Controer Hadi Shirouyehzad 1,, Hamidreza Panjehfouadgaran 2, Reza Dabestani 3,, Mostafa Badakhshian 4, Facuty Member of IE Department Isamic Azad University, Najafabad Branch, hadi.shirouyehzad@gmai.com MSc. of Industria and system engineering, hamidfouadgaran@gmai.com Management department, Isfahan University, reza.dabestani@gmai.com R&D Department, Deputy of Transportation and Traffic, Municipaity of Isfahan, m.badakhshian@gmai.com Received: September 2010, Revised: November 2010 Onine Pubication: January 2011 Abstract The era of gobaization has begun and organizations endeavor to increase their market share in the competitive environment. To achieve the mentioned goa, organizations shoud increase their effectiveness as a major strategy in order to improve their performance. Performance measurement as a manageria key can be used for monitoring activities in organizations. s seection is one of the issues which infuence the efficiency of organizations. Therefore, performance measurement of vendors pays a vita roe in firms. Many conceptua and anaytica modes have been deveoped for addressing vendor seection probems. Hence, a suitabe approach is needed to consider a the factors in order to seect the most efficient vendor. In this paper, fuzzy ogic controer as a robust and easy understanding approach is appied to transform the quantitative variabe to inguistic terms in order to measure the vendors performance. Four criteria which can infuence vendors performance are considered. The criteria are service quaity, price, ateness deiveries and rate of rejected parts. Keywords: Seection, Performance Measurement, Fuzzy Logic Controer (FLC). 1, Facuty member 2, Corresponding author: MSc. of Industria and Systems Engineering 3, MSc. candidate, Industria Management 4, MSc. of Industria and Systems Engineering 311
1. Introduction Suppy chain management (SCM) is a process which tries to fufi the customers request. SCM consist of suppiers, manufacturers, distributors, retaiers and customers which product fow from downstream to upstream and information fows in both directions [1], [2]. Improving efficiency and effectiveness as the most vauabe factors from the customers point of view are one of the major goas of SCM [3]. To achieve the mentioned goa, organizations shoud concentrate on improving their activities in order to increase their competitive advantages in the whoe suppy chain rather than the singe firm. One of the crucia decision making probems in SCM is vendor seection which may infuence organizations performance directy. Therefore, seecting the most suitabe vendor at the right time by predetermined quota of aocation is one of the managers chaenges [4]. There must be an ongoing process to define the procurement requirement needed to support the company s business pan and its operating mode. The vaue of these requirements shoud be considered in addition to price of vendors products. The vaue of product quaity, service eves, just in time deivery and technica support are some of the factors which can be considered in the process of vendor s evauation. Once there is an understanding of the current purchasing situation and an appreciation of what organizations needs to support its business pan and operating mode, a search can be done for seecting the most suitabe vendors. This way can hep the organization to evauate its vendors and identify the most efficient ones. [1] seection is the process which review, evauate and choose the vendors to become a part of organization. It is one of the most important decision making probems, since seecting the suitabe vendors significanty reduce the purchasing costs and improves corporate competitiveness. Evauation methodoogies have ranged from simpe weighted scoring modes to advanced mathematica programming modes [5]. Appropriate vendors or suppiers seection is one of the fundamenta strategies for achieving the quaity of output in any organization, which has a direct infuence on the company. The importance of vendor seection has been stated in the iterature [6]. seection decisions are aso an important component of production and ogistics management in many firms. These decisions are typicay compicated for severa reasons. First, potentia options may require to be assessed on more than one criterion. Dickson [7] identified 23 criteria that have been recognized by purchasing managers in various vendor seection probems. Wind and Robinson [8] concuded that most vendor seection decisions invoved mutipe criteria. Muti-criteria evauation has been recognized to be particuary important in manufacturing strategies where criteria such as reiabiity of deivery and product quaity have increased importance [9], [10]. The second compication of vendor seection decisions is that individua vendors may have different performance characteristics based on the different criteria. For exampe, the vendor who can suppy an item for the east per unit price may not have the best deivery, quaity or performance in compare with other vendors. The third compication surrounding the vendor seection decision comes from interna poicy imitations and externay imposed system constraints paced on the purchasing process. Interna poicy constraints exist either impicity or expicity in the purchasing process such as the number of vendors, minimum or maximum order quantities, the use of minority vendors, etc. Simiary, suppiers may impose some constraints in the buying process ike their own minimum/maximum order quantities based on their production capacity or their wiingness to be in contact with a particuar firm. These constraints utimatey infuence the number of vendors and the order quantities in purchasing function [11]. Partnership with better vendors can hep organizations to improve their performance in different aspects. Therefore, for seecting the most suitabe vendors, their performance shoud be measured. In vendor seection probem, a number of methods with different criteria have been deveoped and examined in SCM. Wind and Robinson [8] proposed a inear weighting method for choosing the best vendors. Gregory used previous works and inked them to a matrix to achieve a new approach to 312
sove the probem. During the recent years, some new approaches were proposed by schoars. Buffa and Jakson for measuring the price, quaity and deivery, used goa programming (GP) agorithm to sove the probems. Sharma et a. proposed a GP method for measuring price, quaity and ead time under budget constraint to vendor seection. Data enveopment anaysis, anaytic hierarchy process and fuzzy ogic are aso the methods and approach which have been used in vendor seection probems [11]. Fuzzy ogic controer (FLC) is usefu when the probem is too compex to be soved with quantitative approaches [4]. During the recent years, a number of researches have been done by appying fuzzy ogic controer in order to measure vendors performance. FLC is aso used for assessing suppiers and monitoring suppy chain networks [12], [13], [14]. In this paper, service quaity, average of ate deivery times, price and rate of rejected parts are considered as the main factors which may infuence vendors performance. Parasuraman mode is aso used to measure vendors service quaity. Data are gathered through a checkist based on the five dimensions of SQ mode and vendors performance is measured through fuzzy ogic approach. The output of this approach is a score from twenty which shows the organization s performance based on the mentioned criteria. 2. Service Quaity Service quaity is a concept that has agitated considerabe interest and discussed because of its difficuties in both defining and measuring it with no overa consensus. A number of different "definitions" has been stated to expain service quaity concept. One that is commony used to expain service quaity is the extent to which a service meets customers needs or expectations [15]. Hence, Service quaity can be defined as the difference between customer expectations of service and perceived service. If expectations are greater than performance, then perceived quaity is ess than satisfactory and hence customer dissatisfaction occurs [16]. Numerous service quaity iteratures have studied conceptuaization, measurement, impementation, and management of service quaity. The concept of service quaity was estabished after there had been a growing interest in the quaity of goods served. Garvin [17] was amongst the first schoars who examined the quaity concepts to cover both goods and services. Garvin expained the perceived quaity as the subjective perception of quaity through indirect measures of quaity comparison. Gronross [18] introduced perceived service quaity as a resut of comparing the rea experience with the expectation of a customer before consuming the service. Based on the perceived service quaity concept Parasuraman et a. [16] appied premises from other previous studies to form their mode of service quaity. The ideas incuded a consumer had difficuty in evauating service quaity rather than goods quaity, that a perception of service quaity was deveoped from a comparison of consumer expectation with actua service performance, aso quaity evauation invoved the evauation of both the process and outcome of service deivery. Consequenty, the conceptua mode study of Parasuraman et a. [16] presented five generic dimensions or factors which are as foows: [19] (1) Tangibes. Physica faciities, equipment and appearance of personne. (2) Reiabiity. Abiity to perform the promised service dependaby and accuratey. (3) Responsiveness. Wiingness to hep customers and provide prompt service. (4) Assurance (Incuding competence, courtesy, credibiity and security). Knowedge and courtesy of empoyees and their abiity to inspire trust and confidence. (5) Empathy (incuding access, communication, understanding the customer). Caring and individuaized attention that the firm provides to its customers 313
3. Short Description of Fuzzy Inference System Approach The pure fuzzy ogic system is the system where the fuzzy rue base consists of a coection of fuzzy IF THEN rues, and the fuzzy inference engine uses these fuzzy IF THEN rues to determine a n mapping from fuzzy sets in the input universe of discourse U R to fuzzy sets in the output universe of discourse V R based on fuzzy ogic principes. The fuzzy IF THEN rues are of the foowing form: () R : IF x 1 is Then y is Where G F and i F 1 and... x n is F n G are fuzzy sets, T x ( x 1,..., x ) U and y V are input and output inquistic n variabes, respectivey, and =1, 2,..., M. Practice has shown that these fuzzy IF THEN rues provide a convenient framework to incorporate human experts knowedge. Each fuzzy IF THEN rue of Eq. (1) defines a fuzzy set F1 x... xfn G in the product space U V. The most commony used fuzzy ogic principe in the fuzzy inference engine is the so-caed sup-star composition, described in detai in [15]. The mini-inference rue and product-inference rue are seected as good from an axiomatic strength point of view, and aso because these inferences rues are computationay simpe [15]. In our appication, we wi use the mini-inference type of inferences rues in our adaptive fuzzy system appication. In order to use the pure fuzzy ogic system in engineering systems where inputs and outputs are reavaued variabes, the most straightforward way is to add a fuzzifier to the input and a defuzzifier to the output of the pure fuzzy ogic system. The fuzzy rue base and fuzzy inference engine are the same as those in the pure fuzzy ogic system. In the iterature, this fuzzy ogic system is often caed the fuzzy ogic controer since it has been mainy used as a controer. It was first proposed by [16], and has been successfuy appied to a variety of industria process and consumer products. A detaied description of this fuzzy ogic system can be found in [15]. The Mamdani fuzzy ogic system has many attractive features. First, it is suitabe for engineering systems because its inputs and outputs are rea-vaued variabes. Second, it provides a natura framework to incorporate fuzzy IF THEN rues from human experts. Third, there is much freedom in the choices of fuzzifier, fuzzy inference engine, and defuzzifier, so that we may obtain the most suitabe fuzzy ogic system for a particuar probem [15]. 3. Fuzzy Logic in Performance Fuzzy ogic is used in performance measurement especiay for vendor performance. Kwong et a. [12] appied scoring method with fuzzy expert to suppier assessment in the industry. Mentioned approach has some advantages such as decreasing the errors with impementing the membership functions and evauation of knowedge with utiizing the fuzzy rues. Lau et a. [13] is appied fuzzy to monitor the suppy chain network for the reiabe and accuracy reasons. Cahn and Qi [14] proposed an innovative method for performance measurement. They caimed that this method is beneficia for foowing reasons; first fuzzy heps to reduce the human errors in judgment and decision making and second, defuzzification provide benchmarking for the managers to anayze the effectiveness of the organization. Jain et a. [20] mixed genetic agorithm and FLC to optimize the 314
evauation. Ohdar and Ray [21] appied data coection and genetic agorithm for suppier assessment. Unahabhokha et a. [22] presented fuzzy expert approach to deveop forward ooking performance measurement for deivery factor in SCM. The approach is deveoped for managers to estabish a systematic way to predict future deivery performance in and identify potentia probems in the company. According to the mentioned methods and approaches, in this paper fuzzy ogic controer based on the data coection via the questionnaire deveoped to measure four criteria in vendor seection probems. The resut of the research is shown how an organization can measure the performance at the moment with these four criteria. The next section is described the methodoogy of the work. 4. Fuzzy Performance Measurement System for Seection FLC design is the major step in the methodoogy which incudes: determining membership function and designing fuzzy rues. Foowing the process of membership function design and fuzzy rues are defined. 4.1 Membership Function A fuzzy system is a system that works based on imprecise knowedge which is stated by inguistic variabes. A membership function (MF) is a part of FLC that provides information which maps the inputted numbers onto the inguistic variabes. A MF is basicay a curve that defines how each point in the input space is mapped to a membership vaue (or degree of membership) between 0 and 1. For instance, high or medium vaue is referring to MF of the fuzzy set that each point in the curve shows a number between 0 and 1. In this paper tripe membership function is defined with trianguar shape with three MFs degree. Trianguar membership function shape was chosen because it is most popuar in the performance measurement [12], [13], [21]. Range of the membership function depended on the aternatives that were designed for each criterion in the questionnaire. 4.2 Fuzzy Rues The rue base of the proposed FLC was defined in three steps; firsty, tota numbers of interactions between the input variabes of the FLC were defined. The numbers of rues were estabished based on the permutation with number of membership function and numbers of criteria. The numbers of factors are six and numbers of membership functions are four. The tota amount of rues are 3^4 equa to 81 rues. The rues iustrate a estates between criteria and their reationships. The output of the rues are determined by experts in the company. Finay, rues and membership functions are entered in the MATLAB software and Mamdani type is chosen to evauate the output of the company. 315
Figure 1. Membership Function in Seection 5. Resuts In this paper, tweve P.V.C vendors of one of the Iranian companies in hose production industry are considered and anaysed. The Data were gathered through some checkists and some interview with the managers of the company. The Data are presented in the Tabe 1. The resuts of anaysing P.V.C vendors are shown in the Tabe 2. Based on the method of this paper, high numbers of output iustrate the poor performance and ow numbers show high performance. Hence, the owest number shows the best performance. According to the Tabe 3, vendor number 6 has the highest eve of performance. In contrast, vendors number 9 and 12 have the owest eve in comparison with others. Tabe1. Data for Seection Data Lateness Service Quaity Price Deiveries Rate of Reject Parts 1 25 29 14 3 2 22 24 6 5 3 82 30 8 6 4 20 25.5 10 3 5 57 29.5 20 8 6 10 25 6 3 7 28 24.5 14 4 316
8 47 28.5 12 4 9 45 27 12 6 10 39 27 24 4 11 12 28.5 4 5 12 35 27.5 10 8 Output Performanc e Tabe2. Output Performance for tweve P.V.C s Output Output Performanc Performanc e e Output Performanc e 1 5 4 5.54 7 5 10 5 2 5.96 5 5 8 5 11 5.89 3 5 6 2.5 9 8.5 12 7.78 5. Concusions One of the success keys for managers is seecting the most appropriate vendors based on their performances. It has been proved that vendor seection may infuence organization s performance, product quaity and services directy. Hence, this question is remained that, how indicators can be monitored in the organization? Which vendor is better in comparison with others? To answer the mentioned questions fuzzy ogic controer is appied as a robust method in vendor seection probem. In this paper, four effective factors incuding service quaity, price, ateness deiveries and rate f reject parts were considered in the performance measurement. Then, FLC approach, appropriate membership functions and fuzzy rues were used and defined in the MATLAB software. Finay, a framework for measuring performance was obtained that change the inguistic terms to quantitative variabes. References [1] Hugos, M. Essentias of Suppy Chain Management, New Jersey: John Wiey, 2003. [2] Chopra, S. and Meinde, P. Suppy chain management: strategy, panning and operations, Prentice Ha, 2001. [3] Mentzer, J.K. Fundamentas of suppy chain management: tweve drivers of competitive advantage, Sage Pubication, 2004. [4] Kumar, M., Vrat, P., and Shankar, R. A fuzzy goa programming approach for vendor seection probem in a suppy chain, Computers and Industria Engineering, Vo. 46, pp. 69-85, 2004. [5] Tauri, S., Narasimhan, R., and Nair, A., performance with suppy risk: a chance-constrained DEA approach, Internationa Journa of Production Economics, Vo. 100, No. 2, pp. 212-222, 2006. [6] Weber, C.A., Current, J.R. and Benton, W.C., seection criteria and methods, European Journa of Operationa Research, Vo. 50 No. 1, pp. 2-18, 1991. [7] Dickson, G.W., An anaysis of vendor seection systems and decisions, Journa of Purchasing, Vo. 2, No. 1, pp. 5-17, 1966. [8] Wind, Y. and Robinson, P.J., The determinants of vendor seection: the evauation function approach, Journa of Purchasing and Materias Management, August, pp. 29-41, 1968. [9] Chapman, S.N., Just-in-time suppier inventory: An empirica impementation mode, Internationa Journa of Production Research, Vo. 27, No. 12, pp. 1993-2007, 1989. [10] Chapman, S.N., and Carter, P.L. Suppier/customer inventory reationships under just-in-time, Decision Sciences, Vo. 21, pp. 35-51, 1990. [11] Weber, C.A., Current, J. & Desai, A., An optimization approach to determining the number of vendors to empoy, Suppy Chain Management: An Internationa Journa, Vo. 5, No. 2, pp. 90-98, 2000. 317
[12] Kwong, C.k., Lp, W.H., and Chan, J.W.K. Combining scoring method and fuzzy expert, Integrated Manufacturing Systems, Vo. 13, No. 7, pp. 512-519, 2002. [13] Lau, H.C.W., Pang, W.K., and Wong, C.W.Y. Methodoogy for monitoring suppy chain performance: a fuzzy ogic approach Logistics Information Management, Vo. 15, No.4, pp.271-280, 2002. [14] Chan, F.T.S., and Qi, H.J. An innovative performance measurement method for suppy chain management, Suppy Chain Management: An Internationa Journa, Vo.8, No.3, pp.209-223, 2003. [15] Guimaraes, A.C.F., and Lapa. C.M.F. Fuzzy Inference to risk Assessment on Nucear Engineering Systems, Appied Soft Computing, Vo. 7, pp.17-28, 2007. [16] Parasuraman, A., Zeitham, V. A., and Berry, L. L. A conceptua mode of service quaity and its impications for future research, Journa of Marketing, Vo. 49, pp.41-50, 1985. [17] Garvin, D. A., Managing Quaity: The Strategic and Competitive Edge. 1st Edition, New York: The Free Press, 1988. [18] Gronroos, C., Quaity comes to service. In E. E. Scheuing and W. F. Christopher (Eds.). The Service Quaity Handbook (pp. 17-24). New York, USA: AMACOM, 1993. [19] Kang G. D., James J., Aexandris K., Measuring of interna service quaity, Managing service quaity, Vo 12 No. 5, pp. 278-291, 2002. [20] Jain, V., Tiwari, M.K., and Chan, F.T.S. Evauation of the suppier performance using an evoutionary fuzzy-based approach Journa of Manufacturing Technoogy Management, Vo.15, No.8, pp.735-744, 2004. [21] Ohdar, R., and Ray, P.K. Performance measurement and evauation of suppiers in suppy chain: an evoutionary approach, Journa of Manufacturing Technoogy Management, Vo.15, No.8, pp.723-734, 2004. [22] Unahabhokha, C., Patts, K., and Hua Tan, K. Predictive performance measurement system: A fuzzy expert system approach, Benchmarking: An Internationa Journa, Vo.14, No.1, pp.77-89, 2007. 318