On the robustness of joint production and maintenance scheduling in presence of uncertainties

Production and preventive maintenance are very important functions in industry which act on the same resources. However, in most real workshops, the scheduling of their respective activities is independent and the constraint that they cannot be accomplished at the same time is rarely considered. Therefore, we are facing a joint scheduling problem of production and preventive maintenance tasks. In addition, this joint scheduling risks at any moment to deviate from the theoretical desired performances when facing disturbances due to various causes. Thus, we must still seek the most robust scheduling, i.e. the one that resists to uncertainties. This paper proposes a new approach to study robustness of joint production and maintenance scheduling in permutation flow shop workshops. The studied scheduling are generated according to two strategies: sequential and integrated. As methods of scheduling resolution, we will consider the well-known ants colony optimization, genetic algorithm, tabu search and some hybridizations of these methods. Our approach can be applied to other joint scheduling generating methods. In particular, we study how insertion of maintenance tasks can contribute to the robustness of production scheduling and how some scheduling strategies and methods are more robust than others. Several experimental results show the merits of our approach.