Fuzzy Analytical Hierarchy Process Part Routing in FMS

Routing flexibility provides the ability of FMS to efficiently encounter traffic problems caused by machine breakdown, excessive workload, etc. The advantages of imposing routing flexibility can be fully obtained by a competent part roufing rule. In this study, Fuzry Analytical Hierarchy Process (Fuzry AHP) is applied to form part routing rules from the attributes of alternate machines: workload on machine buffer, processing time, and the probability that the part being routed to the alternate machine can be processed before the machine fails. By means of Fuzzy AHP, the burdensome mathematical model can be avoided by extracting the relationship between the athibutes from human experience and knowledge instead. The relationships between the attributeS are dynamic which are changed according to the urgency of the part being routed. Three proposed fuzzy-based rules, FuzzyAHP, FuzzyNF and FuzzyWINQ are compared with W[NQ, NINQ, SPT and RAN. The measures of performance are mean flow time, mean tardiness, mean lateness, proportion of tardy jobs, and system utilization. For tardiness and system utilization, FuzzyWINQ performs significantly better than other rules. l. Introduction As competition in industry becomes more intense, a novel production concept named Flexible Manufacturing System (FMS) has been developed to replace conventional production systems. Since FMS uses highly productive and flexible computer-controlled machines as well as automated material handling systems, it is capable of producing a variety of part types, increasing productivity, and reducing production cost. One important property inherent in FMS is its flexibility, which is the capability to encounter and react efficiently to changes in the environment and process requirements. Routing flexibility is one of the various types of f l e x i b i l i t y p r o v i d e d b y F M S . U n l i k e conventional job shops where routing decisions are always made before parts enter the system, FMS always provides alternate processing routes for each part. Shmilovici and Maimon [] define routing flexibility as the ability of a system to route parts to alternate machines in case of breakdowns, or as a response to the prevailing machine loading situation. In this case, parts can be routed to altemate paths to avoid traffic problems, i.e. congestion, or bottlenecks. Part routing can also increase system utilization and decrease makespan. More benefits of routing flexibility are reported by [2-3]. Yao [4] developed routing entropy to evaluate routing flexibility relating to the availability of an individual workstation. Kumar [5] extended Yao's results to include additional technological constraints of the system. Chandra and Tombak [6] employ Linear Programming model to maximize expected contribution of the system which reflects the system routing flexibility. They also point out that routing flexibility depends on a number of alternate machines and their reliability. F lex ib le machines, tool t ransport systems, and routing strategy are the main ingredients of routing flexibility. The routing strategy is very important because inappropriate application of conventional fixed routing to FMS can impede FMS from its fullpotentiality.