A predictive approach to jointly schedule missions and maintenances for a deteriorating vehicle

Abstract

Both for the manufacturer and the user of industrial vehicles, optimizing simultaneously the maintenance and missions schedule at the fleet level becomes a necessity to improve the profitability. Lots of researches have been realized to optimize either the preventive maintenance schedule or the production planning. Integrating both activities in the same schedule has become a new hotspot. Few researches have been led to schedule both maintenance operations and missions for a fleet of systems which deteriorate over time in a comprehensive predictive approach. As a first step to reach this objective, a single vehicle is considered and we propose a method to jointly optimize its predictive maintenance and its mission planning using the remaining useful life and deterioration information. The vehicle has a set of missions to complete. The aim is to group the missions in blocks and interpose maintenance operations between these blocks. However, the vehicle deteriorates over time and each mission impacts differently the deterioration according to its severity. A stochastic process is used to model the deterioration phenomenon and its parameters are changed for each mission. To obtain the best arrangement between missions and maintenance operations, a genetic algorithm is used. A criterion to minimize the maintenance cost is defined, which takes into account preventive and corrective maintenance costs associated with the missions failure probabilities. The genetic algorithm enables to approach the optimal solution in a reasonable time and avoid considering every possible arrangement.