O𝕡timal Run Quantities for an Assembly System with Random Yields

Abstract An assembly system consists of m production lines, each of which produces a certain type of component. The components are then fed into an assembly stage, where each set of m components, one unit from each different type, are assembled. Due to uncertainties involved in the production lines, the component yields are random. The optimization problem is to find the run quantities of the components, in order to minimize the total setup cost while guarantee at a certain level the probability of achieving a given target yield from the assembly stage. Solution procedures are developed that efficiently solve the problem under convex setup costs and increasing failure rate (IFR) component yield distributions. Also studied are the computation of the expected yield and the effect of variability.