| The rapid evolution of automated machine tools in manufacturing has raised the prospect of automated control of the production process. Automated scheduling of a production facility highlights issues which have either never been addressed, or have received only passing interest in the operations management literature. This dissertation addresses one aspect of this scheduling problem, namely the scheduling issues which arise when the buffer space for holding work-in-process inventory is restricted. Understanding this problem is especially relevant to an automated shop floor control system for two important reasons. Firstly, automated decision-making requires the storage space on the shop floor to be discretized so as to make each storage location independently addressable for directing the material handling system. Secondly, building large buffers between machines can be expensive, especially if the storage system is automated. Previous studies on restricted buffers on the shop floor have mainly been concerned with their optimal size and location, and very little is known about scheduling jobs through the facility once these buffers are in place.; This dissertation considers the deterministic scheduling problem which arises when the buffer space is restricted. We first consider the jobshop scheduling problem with limited local buffers available for storage of work-in-process. The problem is modelled as an integer linear program, a lower bound is developed, the complexity of the problem is studied, some heuristics are proposed and results of computational experiments are reported. The special case of this problem is then considered, where multiple machines feed jobs into a single machine and compete for a common limited local buffer. Lower bounds and heuristics are also proposed for this problem and analysed through computational experiments. Finally, we consider the central buffer problem, i.e. the situation where there are no local buffers, but unlimited buffer storage is available at a cost of the time spent traveling to and from the central storage area. An integer linear program is developed for this problem, a lower bound is proposed, some theoretical results about the problem are presented including a complexity result, and heuristics are presented and analysed through computational experiments. |
