Algorithms for mixed-model sequencing with due date restrictions

In a Mixed-Model Sequencing (MMS) environment, customer demands with respect to timeliness of delivery conflict with the production-centered goal of smooth component utilization. This research is designed to be an initial step in the process of dealing with both balance and lateness objectives. The results show that a slight decrease in one-time production leads to substantial gains in smoothing component utilization. The fundamental task in addressing the multiple-objective environment is to distinguish those problem models in a sequence, whose due dates cause the schedule to be unbalanced with respect to part usage. This paper also presents some insights into the behavior of the multi-objective MMS problem. Inherent characteristics of the MMS are revealed and explained, primarily with respect to demand patterns and lateness restrictions. Because of the problem complexity it is necessary to employ a heuristic to produce quality solutions for large problems. This study evaluates a set of heuristics, focusing on the tradeoffs between lateness measures, smooth component utilization, and computation time. This research involves the first application of local search heuristics to this problem. One heuristic, the Border Swap, proves to be an extremely effective mechanism for recognizing problem models, and solving the multi-objective problem.