The development of methods for the design and evolution of reconfigurable cellular manufacturing systems.
The concept of reconfigurable manufacturing is presently being researched due to the need for production systems that are able to economically respond to changes in markets and the rapid introduction of new products. Cellular Manufacturing Systems (CMS) are a central concept in just-in-time and lean manufacturing. Although CMS are able to provide a strategic operating advantage, machine cell clusters do not remain optimal over an extended period of time. The concept of a Dynamic CMS (DCMS) has received attention in recent years; a DCMS is a system where the layout of machines change in order to improve the responsiveness of CMS to changing production requirements. A deficiency in existing DCMS methods is that reconfiguration plans are generated without the consideration of an initial design of the factory floor space for future change. This research distinguishes Reconfigurable CMS (RCMS) from DCMS, as a system that is designed at the outset for changes to system layout and cell configurations. The concept of a Factory Configuration Template (FCT) is proposed in this research; the FCT is a design of the factory floor space to ensure the feasible implementation of reconfiguration plans generated by mathematical models. A nine step method for FCT design is presented that uses a Simultaneous Fuzzy Clustering Heuristic to develop manufacturing cells and part families. A Tabu Search algorithm was develop to generate the optimal arrangement of machine sites in cells. Three multi-period machine assignment models were developed that determine reconfiguration plans based on changing product demand and the introduction of new products. The models that were developed included two integer linear programs that determine the distribution of machine resources among cells over multiple periods. A quadratic zero-one programming model was developed that distributes machines among available sites in cells to promote unidirectional part flow. The results show that RCMS is able to provide a more economical solution than traditional CMS with the added advantage of improved part flow in the system.