Production scheduling and lot streaming at flexible job-shops environments using constraint programming

Abstract

In this contribution, the scheduling at flexible job-shops and the lot streaming problem are simultaneously addressed, by means of a novel Constraint Programming (CP) approach. The proposed CP model can efficiently tackle both the (i) lots splitting, deciding the number of sublots for each lot and the number of parts that belongs to each sublot, and (ii) the scheduling of production tasks, assigning the operations on sublots to machines and defining the start and completion times of those activities. The novelty of the proposal relies on a formulation that (i) can easily be adapted to cope with different operational policies, such as no-wait or wait schedules, idling or intermitted idling, sequence-dependent setup times, among other alternatives, (ii) introduce to new features that impact on the feasibility of the schedules, such as an extension to the inter-operation waiting policy, to consider whether a sublot can wait an unlimited or a limited period of time, as well as an intermediate storage policy, to indicate if the area to hold work-in-progress is constrained or not. Moreover, a set of operational modes found in real industrial settings are addressed. A framework to classify scenarios based on different problem characteristics is also introduced. Different case studies were solved, and good quality solutions were found when minimizing makespan.