9713750 Fowler This grant provides funding for the development of methods for optimizing the design and operation of semiconductor wafer fabrication facilities. Descriptive issues concerning performance evaluation, as well as prescriptive issues concerning scheduling and planning will be studied. The goal of the research on performance evaluation is to develop methodologies for determining the impact of process time variabilities, release policies, scheduling policies, hot lots, equipment failures, product mix, and cluster tools on the cycle-time performance of the plant. In the area of scheduling, the goal is to design and analyze policies for making real-time decisions concerning when to release new wafer lots into the plant, and how to schedule the large numbers of stations in re-entrant fabrication lines in the presence of equipment failures, set-up time constraints, batching considerations, and operator availability constraints. In the area of planning, the goal is to study the issues of yield learning, equipment utilization over time, effects of adding equipment over time, ramp-up, and non-stationary behavior. Also, problems of managing capital outlays over time in the face of demand uncertainty will be studied. If successful, the results of this research will lead to improvements in the design and operation of semiconductor fabrication facilities. They will provide methods for the prediction of the performance of large fabrication facilities before they are built. They will also provide methods for efficiently scheduling the complex operations of expensive fabrication plants so as to improve their cycle-time performance. The research work will thus contribute to the understanding of how to design and operate large wafer fabrication facilities.
