Modeling the Mean and Variance of Product Cycle Time using a Discrete-Event Simulation Model of a Manufacturing Facility
by
Bruce Ankenman
Northwestern University
Coauthors: Barry L. Nelson, Feng Yang

One of the major issues confronting a manufacturing manager is how to reduce production costs and enhance productivity. Adding production equipment increases throughput, but this will also increase the production cost. In factory level operations, a manager wants to concurrently maximize throughput while minimizing cycle time and Work In Process (WIP). Particularly, in a technology-driven industry like a semiconductor wafer fabrication facility (fab), a reduction in the cycle time variability may be the only short-term improvement possible, because of the long lead times associated with procuring equipment. We discuss the fitting of metamodels for both the mean and variance of cycle time-throughput curves developed from discrete event simulation models of semiconductor manufacturing facilities. We focus on a model family that is appropriate for the mean, the variance, and higher moments of the cycle time curve as a function of throughput of the factory. These metamodels together allow for quick evaluation of “what if” production scenarios.

Date received: August 22, 2003