A Directly-Coupled Two-Stage Unpaced Line

Abstract

For a directly-coupled two-stage production system whose stochastic processing times may follow any distribution form, we present a simple procedure for estimating its average cycle time. We then demonstrate systematically the effect of the skewness and kurtosis of processing times on this average cycle time. Finally, we consider a version of the “unbalancing” problem that is more realistic than the classical Hillier-Boling version; we found that for a two-stage directly-coupled system, one should perfectly balance the mean and the skewness of the stages' processing times, but the variance and the fourth moment need not be balanced. Also, if the stages have unequal mean processing times, the station with larger mean should also have larger skewness.