The operating characteristics of wafer production facilities are extremely dynamic, driven by short product life cycles, rapid equipment obsolescence and recurring layout expansion. These factors also have an impact on the design of the automated material handling system (AMHS). The AMHS must be able to react and accept change as rapidly as the production process dictates. The AMHS design engineer faces a significant challenge in that modeling efforts must be proactive and anticipate the long-term requirements of a given facility. There are, of course, several methods available for addressing this issue. This paper points out the limitations to these methods when applied to AMHS modeling and proposes an alternative. Specifically, behavioral trends from historical data can be exploited when appropriate. Simulation results from legacy designs may prove to be an efficient indicator of the validity of new designs.