Abstract: A modeling approach for predictive maintenance (PdM) scheduling model based on decision variable conversion was proposed. Based on the proposed approach, a PdM scheduling model could be established by replacing the time decision variable in the corresponding classical time-based preventive maintenance (PM) scheduling model with product condition variable(s) and checking the correlation of the model parameters, maintenance effect and maintenance constraints with product condition. Based on the proposed modeling approach, the application case study considered an unreported predictive maintenance model in which system benefit was correlated with its condition. The proposed approach used the existing PM scheduling models for reference and simplified the modeling process for PdM scheduling models. The proposed approach can hopefully advance wide application of PdM policies.

Key words: preventive maintenance, predictive maintenance, maintenance modeling, maintenance time optimization