Capacitated lot sizing with parallel machines, sequence-dependent setups, and a common setup operator

In this paper, we consider a special variant of a capacitated dynamic lot sizing problem which, despite its practical relevance, has not been treated sufficiently in the literature. In addition to the significant complexity introduced by the sequence dependency of setup times, the problem is further complicated by the fact that there is only one single setup operator which is responsible for all setups on all machines. Hence, the multi-machine problem cannot be decomposed into multiple single machine problems, as the setup operations must be coordinated among the different machines. In addition, lots are produced in batches whereby the processing time of a lot is a step-wise function of the number of batches per lot. Due to perishability and quarantine issues, time windows for the production are given. We formulate a big-bucket lot sizing model and apply MIP-based heuristics to two industrial data sets.