Integrated planning of maintenance operations and workload allocation

Abstract

Motivated by a practical problem, this paper investigates the integrated planning of maintenanceoperations and workload allocation on a set of machines in a workshop. Given quantities of productsto be produced per period on a planning horizon must be processed on unrelated flexible machines.Moreover, each machine has to undergo one or more maintenance operations that must be plannedwithin a given time window and impact products differently. The main goal is to find a feasible planthat satisfies the machine capacity by allocating the production quantities to machines and assign-ing maintenance operations as late as possible in their time windows. Various original mathematicalmodels are presented. In particular, we propose models that allow maintenance operations andsome production quantities to overlap two consecutive periods. Computational experiments basedon industrial data show that allowing this overlapping helps the earliness of maintenance opera-tions to be significantly reduced in the most difficult instances, going for example from a total of 14periods to only 1 period, and by more than 35% on average.