Safie, Fayssal M. (1988, ASQC) George C. Marshall Space Flight Center, Huntsville, AL
A methodology is presented for assisting designers in the development of inspection requirements and plans for automated multistage production systems. In opposition to the conventional practice of developing the inspection program as a final step in process planning, in this study we address the problem of specifying the inspection requirements for a given product early enough during the design phase to explore tradeoffs and changes among alternative designs, processes, and inspection procedures. A general inspection system cost model is formulated for a serial multistage production system. The system consists of n operations each followed by an inspection. Following this inspection, items either pass on to a subsequent operation, divert to a repair station, or are scrapped. The model involves the formulation of a nonlinear program whose solution determines the maximum attainable reliability of an accepted item under certain conditions of operation, cost, and quality levels among stages. Specifically, the nonlinear formulation is designed to determine the stage inspection requirements that maximize the product reliability level for a given cost. The generation of locally optimal solutions coupled with the ability to interactively examine tradeoffs between cost, product reliability, and inspection requirements, allow engineers to easily generate designs that are pragmatic and at least near optimal in economic terms. A numerical example is included to illustrate the method.