Fuzzy Expert System for Evaluation of Leanness Achievement Degree (LAD) in Supply Chains

Transcription

1 Australian Journal of Basic and Applied Sciences, 5(6): , 2011 ISSN Fuzzy Expert System for Evaluation of Leanness Achievement Degree (LAD) in Supply Chains 1 M. Zarei, 2 H. Khademi Zare, 3 M. Jamali Paghaleh 1 Department of Industrial Engineering, Islamic Azad University, Shiraz Branch, Iran. 2 Department of Industrial Engineering, Yazd University, Yazd, Iran. 3 Young Researchers Club, Islamic Azad University, Zahedan Branch, Iran. Abstract: In today's world of global markets and stiff competition in every product, it becomes imperative for companies to explore ways to improve their productivity in terms of implementing flexible and standardized technology, and adopting proven management principles. Lean production is one of the initiatives that many major businesses all around the world have been trying to adopt to remain competitive in the increasingly global market. The adoption of lean principles leads to a positive outcome with stable and/or increasing profitability. This study presents a framework, based upon Expert System (ES), to evaluate the Leanness Achievement Degree (LAD) of companies/ organizations. The proposed ES evaluates LAD of four company branches,producing dairy products, by five criteria including fuzzy evaluation results, human resource training and involvement, supplier information sharing, continuous improvement, and customer relationship management. By benchmarking the branch that obtained the highest LAD and taking into consideration its top scores for the five mentioned criteria in the proposed ES, other company branches can improve their lean achievement. Key words: Leanness Achievement Degree (LAD), Expert System (ES), Fuzzy logic, Lean Attributes (LAs). INTRODUCTION The core thrust of lean production is the capability of working synergistically to create a high quality system that produces finished products at the pace of customer demand with little or no waste (Shah and Ward, 2003). Lean promises significant benefits in terms of increased organizational and supply chain communication and integration (Crute et al., 2003; Scherrer-Rathje et al., 2009). Thus, evaluating the Leanness Achievement Degree (LAD) of organizations is a cause for concern. In this paper, a framework is presented to evaluate LAD of companies. The methodology was used in a big company of dairy products, here called ABC Company, which has four branches expanded in four provinces of Iran; each company branch produces dairy products in a big factory. By employing the approach in ABC Company we aimed to evaluate and compare LAD of its four branches. The branch with the highest LAD would be then recognized as benchmark. The methodology is based upon Expert System (ES) which evaluates LAD by five criteria. Moreover, fuzzy logic is exploited to translate linguistic judgments required for expressing Lean Attributes (LAs)' scores, obtained by each branch, into numerical values. The remainder of the paper is organized as follows. In the second section, lean production, and ES framework are presented as are commonly discussed in literature. Moreover, ES and its application in food industry is discussed. In the third section, LAs are identified on the basis of findings from the literature, and the approach to evaluate LAD is detailed. Finally, the last section contains the concluding remarks and future research directions. 2. Review of the Literature: 2.1. Lean Production: Lean production, originating from the Toyota Production System is one of the initiatives that many major businesses all around the world have been trying to adopt in order to remain competitive in the increasingly Corresponding Author: Mahnaz Zarei, Department of Industrial Engineering, Islamic Azad University, Shiraz Branch, Iran. 1443

2 global market (Pe rez and Sa nchez, 2000; Schonbergerm, 200; Womack, et al., 1990). The term 'lean production' was first used by Krafcik (1988); subsequently, Womack et al. (1990) used this term to contrast Toyota with the western 'mass production' system in his book 'The Machine that Changed the World'. The focus of this multi-dimensional approach is on cost reduction by eliminating non-value added activities, and using tools such as JIT, cellular manufacturing, total productive maintenance, production smoothing, setup reduction and others to eliminate the waste (Abdulmalek and Rajgopal, 200; Monden, 1998; Nahmias, 2001; Scherrer-Rathje, et al., 2009). The core thrust of lean production is that the mentioned tools can work synergistically to create a streamlined, high quality system that produces finished products at the pace of customer demand with little or no waste (Shah and Ward, 2003). The adoption of lean principles, which place more emphasis on levelling the production schedule (Naylor et al., 1999), leads to a positive outcome with stable and/or increasing profitability (Cox and Chicksand, 2005) Expert System in Food Industry: Expert system is a branch of applied artificial intelligence (AI), and was developed by the AI community in the mid-1960s (Liao, 2005). ES is needed to supplement, complement, or replace certain human functions (Badiru, 1988). Novel tools of artificial intelligence such as fuzzy and case-based reasoning, ESs and neural networks offer great possibilities for the food industry. Fuzzy ESs have been shown to be a valuable tool in dealing with vague and incomplete information, and in incorporating human expert knowledge into process models and control (Linko, 1998). In her paper, Linko (1998) discussed the state-of-the-art of expert systems and their potential for food-related applications. Periss (2002) presented an overview of the most relevant contributions in the field of ES applications in chemical analysis of foods, along with a critical discussion of future. Short and middle term predictions concerning the potential of ES in food analysis were made (Periss, 2002). In their paper, Lababidi and Bakers (2003) described an integrated web-based fuzzy expert system for food dryer selection. A modular approach to designing the system was employed. The system was fully implemented to run on the web and provides an excellent example of how a number of heterogeneous tools and applications can be integrated on the web (Lababidi and Bakers, 2003). 2. The Approach to Evaluate LAD: Our proposed framework to evaluate LAD by ES includes four main phases. It has a stepwise description as presented below: 3.1. Knowledge Acquisition to Identify LAs: Lean Attributes (LAs), hereafter defined as the elements which constitute the underlying structure of a lean organization, are originally conceived as core concepts of lean manufacturing. Based on a review of the normative literature (de Treville and Antonakis, 2006; Hopp and Spearman, 2004; McLachlin, 199; Narasimhan et al., 2006; Prince and Kay, 2003; Shah and Ward, 2003), LAs were defined, as shown in Table Determine Total Fuzzy Score of Las for the I Th Company Branch: Fuzzy logic provides an effective means of dealing with problems involving imprecise and vague phenomena (Dohnal et al., 1993). Thus, the major contribution of the fuzzy set theory is its ability to represent vague data (Zadeh, 1965). If a,b and c, respectively, denote the smallest possible value, the most promising value and the largest possible value that describe a fuzzy event, then the triangular fuzzy number (TFN) as shown in Fig. 1, can be denoted as a triplet (a,b,c) where, a < b < c. When a = b = c, it is a non-fuzzy number by convention. The membership function can be defined as (Chamodrakas et al., 2009; Zimmermann, 1991): If M = (a 1,b 1,c 1 ) and N = (a 2,b 2,c 2 ) represent two TFNs, then the required fuzzy calculations are performed as below (Chien and Tsai, 2000): Fuzzy Addition: M r N = (a 1 + a 2, b 1 + b 2, c 1 + c 2 ). (2) 1444

3 Fuzzy Multiplication: M r N = (a 1 a 2, b 1 b 2, c 1 c 2 ). (3) M r 1 / N = (a 1 /c 2, b 1 /b 2, c 1 /a 2 ) (4) Fuzzy and Natural Number Multiplication: r q M = (r.a, r.b, r.c). (5) In our integrated approach, fuzzy logic was exploited to translate linguistic judgments required for expressing LAs' scores obtained by each company branch, into numerical values. Experts in technical matters scored LAs in each branch by linguistic expression. The Fuzzy Score of the j th LA obtained by the i th company branch, i.e. B i, was represented as FS ij ; as an example, for LA4, i.e. volume flexibility, we selected a wide spectrum from "Very Low" to "Very High" flexibility. Linguistic judgments and corresponding fuzzy numbers are indicated in Table 2. Then the total fuzzy scores were calculated as below: TFS Bi j1 W j FS ij i = 1,...,4 (6) where TFSBi is the Total Fuzzy Score of the i th company branch, and W j is importance weight of the jth LA. In order to obtain the Lean Score of each branch, TFSB i should be de-fuzzified. Suppose M (a,b,c) is a TFN, then the defuzzified and crisp value of TFSB i is computed as: (a + 4b + c) / 6 () 4 TFSB i i1 w ( a, b, c) i = 1,...,4 (8) 4 where w is a triangular fuzzy number that indicates the mean value of TFS of all company branches being evaluated by the algorithm. Then the mean value of a and b is calculated and named a 1 ; in the same way the mean value of b and c is named b 1. If the Crisp Value of TFSB i is less than a 1, then Fuzzy Evaluation Result is Poor. If the Crisp Value of TFSB i is between a 1 and b, then Fuzzy Evaluation Result is Fair. If the Crisp Value of TFSB i is between b and b 1, then Fuzzy Evaluation Result is Good. If the Crisp Value of TFSB i is more than b, then Fuzzy Evaluation Result is Excellent. The resulting crisp values indicated as "Technical Evaluation Result" are represented as the first criterion for rules of evaluation in the proposed expert system Design the Expert System: In our expert system, five main criteria are defined to evaluate and compare LAD of company branches of the same field, as demonstrated in Fig. 2. The criteria and their relevant alternatives are presented in Table 3. The whole number of possible situations would be: =3125 In A rules series, the situations that evaluation results are "0" without considering other criteria are listed below: 1445

4 A: Rule 10: IF [Fuzzy Evaluation Result] = "Very Poor" THEN LAD is {0} Rule 20: IF [Fuzzy Evaluation Result] = "Poor" THEN LAD is {0} Rule 20: IF [Continuous Improvement] = "Very Poor" THEN LAD is {0} Rule 30: IF [Customer Relationship Management] = "Very Poor" THEN LAD is {0} Rule 40: IF [Supplier Information Sharing] = "Very poor" THEN LAD is {0} Thus, the remaining situations would be: = 960 Some examples of different situations and their final ranking are presented in Table 4. If w k and A k are the criteria weight of the proposed ES, and lean achievment based on the k th criteria, respectively, then: 5 wk Ak k 1 FLADp 5 p = 1,..., (9) where FLAD is the Fuzzy LAD of each situation. LAD p p1 FLAD p (a, b, c ) p = 1,..., (10) LADp is LAD of the p th situation and can be 0,1,2 or 3 that indicates Poor, Fair, Good or Excellent degree of lean achievement by Bi, respectively. Again: If the Crisp Value of FLAD p is less than a1, then LAD is Poor. If the Crisp Value of FLAD p is between a1 and b, then LAD is Fair. If the Crisp Value of FLAD p is between b and b1, then LAD is Good. If the Crisp Value of FLAD p is more than b, then LAD is Excellent. Fig. 1: Triangular fuzzy number (TFN). Fig. 2: Expert system for evaluation of LAD. 1446

5 Table 1: Lean Attributes selected from the normative literature Lean Attribute (LA) LA1 LA2 LA3 LA4 LA5 LA6 LA Stable unit cost Degree of customization Delivery responsiveness Volume flexibility Employee satisfaction Delivery reliability Cost efficiency Table 2: Linguistic judgments and corresponding fuzzy numbers. SLinguistic Expressions TFN Very High (VH) Excellent (E) (0.;1;1) High (H) Good (G) (0.5;0.;1) Medium (M) Fair (F) (0.3;0.5;0.) Low (L) Poor (P) (0;0.3;0.5) Very Low (VL) Very Poor (VP) (0;0;0.3) Table 3: Criteria and alternatives for rules of evaluation in our expert system. Row Criteria No. of Alternatives Alternative 1 Technical Evaluation Result ( LAs' evaluation) 5 Excellent (E) Good (G) 2 Human resource training & involvement 5 Fair (F) 3 Supplier Information Sharing 5 Poor (P) 4 Continuous Improvement 5 Very Poor (VP) 5 Customer Relationship Management 5 Table 4: Some examples of different situations and their final ranking. Final Technical Human resource Supplier Continuous Customer Ranking (p) Evaluation Result training & involvement Information Sharing Improvement Relationship Management E G F E G F P VP E G F P E G F P E G F P 0 (Poor) * * * * * 1 (Fair) * * * * * 1 (Fair) * * * * * 2 (Good) * * * * * 2 (Good) * * * * * 3 (Excellent) * * * * * 3 (Excellent) * * * * * Conclusions and Suggestions: The adoption of lean principles, which puts more emphasis on levelizing the production schedule (Naylor et al., 1999), leads to a positive outcome with stable and/or increasing profitability (Cox and Chicksand, 2005). Lean production promises significant benefits. However, when summarizing the lean evolution, Hines et al. (2004) presented comments on approaches that have sought to address some of the gaps in lean thinking. The objective was to provide a framework for understanding not only the evolution of lean production as a concept but also its implementation within an organization (Hines et al., 2004). In this paper an algorithm was presented in order to evaluate the degree of lean achievement in companies or between a company branches. The company branch with the highest LAD can be remarked as benchmark; so by taking into consideration the fuzzy scores (FSij) of LAs as well as other criteria for rules of evaluation in the proposed expert system with top scores, other company branches can improve their lean achievement. In the presented ES, five main criteria were defined to evaluate and compare LAD of companies/ organizations. As the four mentioned company branches have the same products and work in the same field, so their Wj are the same. In case of comparing companies or organizations of different product/ service types, for example comparing Meat, Dairy Products, and Canning foods (as some different SCs), then LAs' weights (W j ) of each one would be different. The focus of attention of future researches can be on comparing companies of different product/ servise types. ACKNOWLEDGEMENT The authors wish to express their sincere appreciation to the head managers of ABC dairy products Co. for their special cooperation. 144

6 REFERENCES Abdulmalek, F.A., J. Rajgopal, 200. Analyzing the benefits of lean manufacturing and value stream mapping via simulation: A process sector case study. International Journal of Production Economics, 10: Badiru, A.B., Expert systems and industrial engineers: Apractical guide to a successful partnership. Computers and industrial Engineering 14(1): Chamodrakas, I., N. Alexopoulou, D. Martakos, Customer evaluation for order acceptance using a novel class of fuzzy methods based on TOPSIS, Expert Systems with Applications, 36: Chien, C.J., H.H. Tsai, Using fuzzy numbers to evaluate perceived service and quality. Fuzzy Sets and Systems, 116: Cox, A., D. Chicksand, The limits of lean management thinking. European Management Journal 23(6): Crute, V., Y. Ward, S. Brown, A. Graves, Implementing Lean in aerospace-challenging the assumptions and understanding the challenges. Technovation, 23: Davidson, V.J., Expert systems in process control. Food Research International, 2(2): Treville, de S., J. Antonakis, Could lean production job design be intrinsically motivating Contextual, configurational, and levels-of-analysis issues. Journal of Operations Management, 24(2): Dohnal, M., J. Vystrcil, J. Dohnalova, K. Marecek, M. Kvapilik, P. Bures, Fuzzy food engineering. Journal of Food Engineering, 19(2): Hines, P., M. Holweg, N. Rich, Learning to evolve: A review of contemporary lean thinking. International Journal of Operations & Production Management, 24(10): Hopp, W.J., M.L. Spearman, To pull or not to pull: what is the question? Manufacturing and Service Operations Management, 6(2): Krafcik, J.F., Triumph of the lean production system. Sloan Management Review, Fall. Lababidi, H., C.G.J. Baker, Web-based expert system for food dryer selection. Computers and Chemical Engineering, 2: Liao, S.H., Expert system methodologies and applications-a decade review from 1995 to Expert Systems with Applications, 28: Linko, S., Expert systems-what can they do for the food industry? Trends in Food Science and Technology, 9(1): McLachlin, R., 199. Management initiatives and just-in-time manufacturing. Journal of Operations Management, 15: Monden, Y., Toyota Production System-An integrated approach to just-in-time, third ed. Engineering and Management Press, Norcross, Georgia. Nahmias, S., Production and Operations Analysis, fourth ed. McGraw Hill, New York. Narasimhan, R., M. Swink, S.W. Kim, Disentangling leanness and agility: an empirical investigation. Journal of Operations Management, 24: Naylor, J.B., M.M. Naim, D. Berry, Leagility: integrating the lean and agile manufacturing paradigm in the total supply chain. International Journal of Production Economics, 62: Pe rez, M.P., A.M. Sa nchez, Lean production and supplier relations: a survey of practices in the Aragonese automotive industry. Technovation, 20: Prince, J., J.M. Kay, Combining lean and agile characteristics: creation of virtual groups by enhanced production flow analysis. International Journal of Production Economics, 85: Schellekens, M., T. Martens, T.A. Roberts, B.M. Mackey, B.M. Nicolai, Van J.F. Impe, De J. Baerdemaecker, Computer aided microbial safety design of food processes. International Journal of Food Microbiology 24(1-2): 1-9. Scherrer-Rathje, M., T.A. Boyle, P. Deflorin, Lean, take two! Reflections from the second attempt at lean implementation.business Horizons, 52: Schonbergerm, R.J., 200. Japanese production management: An evolution with mixed success. Journal of Operations Management, 25: Shah, R., P.T. Ward, Lean manufacturing: context, practice bundles, and performance. Journal of Operations Management, 21: Womack, J.P., D.T. Jones, D. Ross, The machine that changed the world. Macmillan Publishing Company, Canada. Zadeh, L.A., Fuzzy Sets. Information and Control, 8:

7 Zimmermann, H.J., Fuzzy set theory and its Applications. (Second ed.). Boston: Kluwer Academic Publishers. 1449