A Simulation Study of Operating Policies in a Hypothetical Dual-Constrained Job Shop

This paper describes a simulation study of the relative impact of due-date assignment, dispatching and labor assignment decision rules on the performance of a hypothetical dual-constrained job shop. Six criteria are used to measure the performance of the decision rules. They are mean flow-time, variance of flow-time, mean lateness, variance of lateness, proportion of jobs late, and total labor transfers.

Multiple regression is the principal method of analysis of the results of the experiments. This technique provides regression coefficients and analysis of variance statistics for the decision rules for the various performance measures. Since the objective of this study is to go beyond simple statements of the significance of the various decision rules, the multiple regression results are used with analysis of variance statistics to indicate the relative impact of the decision rules. The regression coefficients and the omega squared (ω2) indices [Hays, W. L. 1962. Statistics. Holt, Rinehart and Winston, New York, 406–408.] indicate that the relative importance of the due-date assignment, dispatching and labor assignment decision rules is dependent upon the measure of performance considered. In addition, the relative importance and optimality of the dispatching rules is dependent upon the due-date tightness for selected performance measures.