A beam heuristic method for mixed-model scheduling in setups

Mixed-model scheduling involves determining a production sequence for multiple products along a single assembly line. While the goal of using this approach is often determining a schedule which keeps the usage of parts as constant as possible, there is an implicit but often unrealistic assumption that setup times are negligible. When setup times are assumed to be significant, the sequencing decision becomes a multi-objective problem which needs to minimize both the usage rate and the number of setups. However, the objectives of a low usage rate and a minimum number of setups are frequently in opposition with one another. To address this situation, this research presents an efficient frontier approach to support sequencing decisions that are made as a result of the compromise necessitated by these two conflicting goals. A beam-search heuristic is used to effectively generate the efficient frontiers needed to solve the problem.