Considering human resource constraints for real joint production and maintenance schedules

•We propose new method for inserting a condition based maintenance (CBM) activity.•We take into account the human resources constraints when planning the CBM activity.•We propose fuzzy estimation for both processing time of CBM and RUL of the machine.•We propose a multi-agent architecture for the workshop.•We propose two phases: (1) a one-machine and (2) a multi-machine resolution.

This study deals with the problem of scheduling jobs and preventive maintenance (PM) activities in production workshops jointly in order to reduce failure occurrence. PM activities can be planned at fixed periods (time-based PM) or in response to a signal provided from a machine’s captor after anomaly detection (condition-based PM). Maintenance activities’ planning depends on different constraints like human resources, spare parts availability and past operating duration of the machine.Proposed related works evaluate possible arrangements of both production and maintenance activities according to different criteria (makespan (Cmax: completion date of the last operation on the last machine), machine’s failure risk, jobs’ lateness, etc.). However, maintenance activities are frequently considered as production jobs. In fact, maintenance activities present different constraints and cannot be planned in the same manner as production ones. In addition to that, most of current works deal with the time-based maintenance planning even so condition-based maintenance is qualified as more realistic and more economic than the former.This work discusses a new approach to integrate the scheduling of production and maintenance operations. Proposed approach takes explicitly into account human resources availability and skills when updating integrated production and maintenance schedules. It is based on multi-agent systems for modeling the production workshop.