Scheduling virtual cellular manufacturing systems

Jocelyn Rene Drolet, Purdue University

Abstract

The efficiency of FMS comes from the fact that routing decisions are made at the last moment based on conditions prevailing at the time. The combinatorial complexity which evolves when dealing with multiple routings, large variety of parts, unexpected events, etc, makes it difficult to develop any decision-making scheme based on look-ahead information. In this context, synchronizing the movement of parts and tools so that the right combination ends up at the right workstation at the right time is not possible; frequent delays and deadlocks are likely to occur. To help resolve these problems, a new concept in small batch manufacturing has been developed. In this thesis we begin by defining a new manufacturing concept termed Virtual Cellular Manufacturing System (VCMS). The VCMS is an expansion of the virtual cell concept proposed by McLean and al. (MCL86) in the early 1980's. Second, a time-phased linear programming algorithm has been developed for scheduling these systems. The scheduling algorithm permits the creation of virtual cells and schedules them under workstations and tools availability constraints. Equally important, the algorithm is characterized by its polynomial time complexity. The size of the model grows quickly but linearly and remains reasonable for a typical size factory. It is efficient enough to be executed in real time for scheduling any virtual cell based system within hundreds of seconds with current computer technology. Third, an object oriented simulation of a VCMS has been performed and various variables of interest have been examined. The results showed improvement of nearly 20% in both, machine utilization and throughput over the group technology cell system. In addition, the average makespan and the number of parts in the system showed a significant improvement.

Degree

Ph.D.

Advisors

Moodie, Purdue University.

Subject Area

Industrial engineering

COinS