MODELISATION D’UN OUTIL DE PRONOSTIC POUR L’OPTIMISATION D’UNE POLITIQUE DE MAINTENANCE PREVENTIVE PREVISIONNELLE

Nowadays safety systems are designed to warn the operator in case of the system incoming failure or danger. Unfortunately those systems can fail because the warning information came late. Then the operator is exposed to hazardous situations and could not have means and time to face them. In this paper from a diagnostic Fault Tree Analysis (FTA) and safety/hazard analysis like FMECA (Failure Mode, Effects, and Criticality Analysis of the system, the EPF (Episodic Probability Function) concept is developed to assess the probability and/or the time of shortcomings occurrence in the process FTA with the objective of anticipating necessary maintenance actions, identifying potential resource requirements (equipment, skill levels, tools, etc.), and estimating maintenance task completion times