A model has been developed which describes the transient performance (which is more realistic than assuming steady state performance) of large-scale assembly systems. This model has computer run times which are low and affordable (CPU cost of $1.00, for example), compared with prohibitive costs for simulation runs ($700, for example). The model is based on two assumptions, which were partially verified in the preceding work. One purpose of this research is to complete the verification. If this can be done, the result should be a planning process to guide the system designer in developing and evaluating alternative designs; a quantitative procedure to assure effective material management; and a descriptive model that quantifies all relevant trade-offs in test/rework strategies. The approach involves four tasks. First is development of the mathematics required to manipulate the model and to reduce errors resulting from approximations. Second is generation of a computer program that can apply the model for both description and prescription. Third will be evaluation of the tools in actual production situations. Fourth will be dissemination of the results, including making the computer program available in usable form.
