Almost all equipment replacement models assume that replacement technology remains constant over time and that replacement decisions are driven primarily by deterioration of the current equipment. However, in areas such as computer hardware and software, robotics, automatic control, and other electronic technologies, which are undergoing rapid technological change, replacements are made to upgrade the current system, as well as replace deteriorating equipment. Hence, it is critical to consider the state of the currently available replacement technology and the potential for technological improvements in the near future to make economic replacement decisions. Current replacement models are not well-suited to this problem. The objective of this study is to develop a useful method for analyzing the equipment replacement problem under technological change. Because technological forecasting is costly and uncertain, the basic approach will be to characterize conditions for which it is optimal to replace. These conditions will help the decision-maker determine whether forecasting is necessary and will indicate how sensitive the optimal "keep" or "replace" decision is to the accuracy of the forecast. Means for computing these conditions will be based on recent results in planning horizons and sensitivity analysis in dynamic programming. 8714696 Rode Gallium Arsenide and related compound semiconductors have progressed from the R&D to the production stage where manufacturing efforts are being planned and carried out. The processing of these materials is very different from the processing of Silicon for integrated circuits. This proposal covers a request for support of the Second U.S. Conference on GaAs Manufacturing Technology. A substantial group of leaders in the field, representing U.S. industry, universities and government, will meet at this conference. This conference will take place October 12-14 at Portland, OR. This conference represents an attempt to integrate university based manufacturing studies with industrial manufacturing engineering efforts. NSF support of this interactive role is extremely important.
