Bayesian Network-based Proactive Maintenance

Today, the degradation mastery of critical components is one of the efficient means to minimise the non expected stops of production system and reduce its costs. Consequently, a prognosis process-based proactive maintenance system is proposed and formalized in this paper. This process allows to follow the system degradation and to implement adequate preventive actions for anticipating failures. Its development combines both probabilistic and event approaches for degradation mechanism modelling and dynamical monitoring.