Suchergebnisse

ABSTRACTWhile a fairly large body of literature dealing with just‐in‐time (JIT) manufacturing exists, most of it is descriptive. JIT research consists primarily of case studies, surveys, a few analytic models (primarily dealing with Kanban), and some simulations. This research builds on the existing literature to develop and test an exploratory model of JIT purchasing with an empirical field study of an actual JIT implementation. While the results basically follow hypotheses developed from the literature, some anomalies were found. The research provides both a methodology for further JIT study and an indication of the possible relative importance of JIT purchasing characteristics as they exist in a North American installation. Especially insightful is the fact that the relative importance changes depending on whether customer inventory or supplier inventory is used as the dependent variable. With customer inventory, the variables measuring delivered lot size, the size of release changes, and supplier manufacturing lead time were significant. With supplier inventory, only supplier manufacturing lot size was significant.

This research compares Material Requirements Planning (MRP), Kanban, and Period Batch Control (PBC) as alternative approaches to the planning and control of multi‐cell manufacturing involving flow cells and assembly. Since previous research on performance of these systems in cellular manufacturing has been primarily conceptual, the experiments reported here provide new insights into their comparative performance.The results show that the production environment is a major factor in system choice. Three operating factors—Master Production Schedule (MPS) volume variation, MPS mix variation, and setup time/lot size—clearly affect system choice. All systems performed well under Justin‐Time (JIT) conditions; there was no advantage to Kanban. Under the mixed conditions of high MPS variation, but small setup time/lot sizes, PBC produced superior performance compared to Kanban and MRP. Under non‐JIT conditions, MRP was seen as clearly more effective. Finally, the results indicate that when conditions permit very small lot sizes relative to requirements, Kanban may perform best, even when MPS variation is high.