Evaluation of dispatching rules in manufacturing using simulation and data envelopment analysis

In manufacturing environments, many decisions are involved to improve the production systems. An important decision that can impact the performance of a production system is the selection of appropriate dispatching rules for parts and/or products. Selecting the dispatching rules involves multiple conflicting objectives such as maximizing machine utilization and profit and minimizing work-in-process (WIP). In this research, a hybrid push-pull production system is considered which combines push and pull production control strategies. A simulation model is developed for the system to account for the uncertainty in processing times and order arrivals. eight dispatching rules are considered: earliest due date (EDD), fist-in-fist-out (FIFO), last-in-first-out (LIFO), shortest processing time (SPT), longest processing time (LPT), longest imminent processing time (LIPT), shortest imminent processing time (LIPT), least work next queue (LWNQ). Data envelopment analysis (DEA) is then used to evaluate the dispatching rules based on lead time, average WIP, total profit and tardiness. For the manufacturing system considered in this study, it was found that the best dispatching rule is the LPT and the least efficient is the LIFO. The results can be useful for engineers, production managers, supervisors and directors of production operations.