Design and Operational Analysis of Virtual Cellular Manufacturing Systems

Abstract

The research focus in the VCM domain is primarily attributed to the benefits that these systems afford within the present volatile manufacturing environment. VCM systems (VCMSs) offer the potential to harness the benefits of both the cellular and functional manufacturing systems. However, this often comes at a cost: the design and operation of VCMSs can become a complex issue. The present work pointedly addresses this complexity which therefore becomes the central theme of this research effort. newlineWe first address the design aspect within VCMSs. Specifically, our focus is on flow-shop based VCMSs wherein arriving parts are first clustered into families based on processing similarities. Then, the families are sent for processing to functional departments where they compete for available workstations and switch-overs between families entails a pre-defined setup time. In this context, we propose a novel family-based procedure. We compare the performance of the proposed approach with an existing procedure available in the literature. newlineHaving dealt with the design aspect within flow-shop type VCMSs, we next intentionally choose a job-shop type VCM for studying the operational aspects. For a hypothetical VCMS, we construct a full-factorial simulation experiment assuming the presence of the following experimental factors: process plans, product-mixes, shop loads, setup-to-process time ratios. The relative performance of different due-date based scheduling rules were compared with the other conventional heuristics within the assumed VCM. Performance comparisons were based on the following measures: average tardiness, average flowtime, makespan, and average workstation work-in-process. newlineGiven that the above simulation study was based on the assumption of a single (unaltered) process plan per simulation run, in the final part of the thesis, we ventured to explore the possible differences in VCM performance in the presence of changing part-mixes necessitating alterations in process plans during a single simulation run. Again, a full-factorial experiment was conducted in the presence of the sequential process plans on account of the dynamically changing part-mixes. Results of this Periodic VCMS vis-à-vis due-date scheduling heuristics afford an interesting comparison with the earlier results. newline