Petri net versus queuing theory for evaluation of FMS

Abstract

PurposePetri net (PN) and queuing theory are used in performance analysis of a flexible manufacturing system (FMS). They are used to determine the FMS measures of performance. These measures of performance include optimal work‐in‐process (WIP), lead time, production rate, machine utilization, and number of servers at each station. The purpose of this paper is to present a comparison between PN and queuing network tools to determine the optimum values for FMS measures of performance.Design/methodology/approachPN model with dual kanban and closed queuing network are used for analysis and performance evaluation of FMS. Integrated network analyzer and Lingo softwares are used for performance evaluation of FMS by PN tool. CANQ software is used for performance evaluation of FMS by queuing network. In both the approaches, the throughput is bounded by the utilization of the bottleneck machines.FindingsThe comparison shows that for the given number of servers, PN gives better values of performance measures for FMS. The PN optimization gives minimum WIP corresponding to the maximum production rate. Minimum WIP leads to minimum lead time.Research limitations/implicationsThe results are concluded based on one case study. In future research, the results may be achieved by doing more case studies with different numbers of system parameters and/or parameter settings.Practical implicationsUsing the PN model, the production manager may design, analyze, evaluate, and even optimize the layout of the production system for minimum WIP, maximum throughput, and reduced lead time. The determination of the total WIP, total number of stations in the production system, and the number of servers at each station may be helpful in shop floor management. It may result in more production efficiency along with ease of supervision.Originality/valueThis paper presents a first novel comparison of its kind between PN and queuing network for evaluation of FMS.