Recently, Prognostics and Health Management (PHM) reveals to be a key feature for industrials as it should improve manufacturing system availability and allow avoiding downtime as well as inopportune maintenance spending. In this context, we investigate the problem of scheduling several jobs on a single machine subjected to predictive maintenance based on PHM. We propose a new hybrid genetic immune algorithm, called IPro-HGIA, which incorporates an emulation of GA with artificial immune system, to create an integrated prognostic based scheduling for planning both production and predictive maintenance interventions under the total interventions cost minimization criterion. Furthermore, we use the principals of vaccination and receptor editing in order to strengthen search ability. Computational results show the efficiency of our hybrid scheme.

We tackle in this paper the Permutation Flow-shop Scheduling Problem (PFSP) with predictive maintenance interventions. The objective is to propose an integrated model that coordinates production schedule and predictive maintenance planning so that the total time to complete the schedule after predictive maintenance insertion is minimized. Predictive maintenance interventions are scheduled based on Prognostics and Health Management (PHM) results using a new proposed heuristic. To jointly establish an integrated scheduling of production jobs and predictive maintenance actions, we propose a tailored genetic algorithm incorporating properly designed operators. Computational experiments carried out on Taillard well known benchmarks, to which we add both PHM and maintenance data, show the efficiency of the newly proposed maintenance planning heuristic and genetic algorithm.