Planning rolling stock maintenance based on prognostic data (Condition Based Maintenance, CBM) is a trend since the ideal timing for maintenance before failure can be planned. With the use of prognostics, a failure can be predicted so Corrective Maintenance (CM) can be avoided. T
...
Planning rolling stock maintenance based on prognostic data (Condition Based Maintenance, CBM) is a trend since the ideal timing for maintenance before failure can be planned. With the use of prognostics, a failure can be predicted so Corrective Maintenance (CM) can be avoided. Traditional rolling stock decision-making for Preventive Maintenance (PM) is based on the time and/or mileage since last PM routine at the maintenance depot. A sufficient amount of literature is available that considers rolling stock PM planning optimization methods. Planning rolling stock maintenance is constrained by the required availability for passenger operations, the conditions for PM and the maintenance depot capacity. However, the integration of CBM with the rolling stock PM planning has not been researched. CM also needs to be performed and is considered a disruptive element in the maintenance planning. It is expected that integrating CBM in the rolling stock PM planning is less disruptive than CM. This study investigates how CBM impacts the rolling stock PM planning by formulating a deterministic MILP that uses a rolling horizon framework that minimizes the maintenance costs. An approach that optimizes the rolling stock PM planning while being disrupted by unexpected failures that lead to CM is compared with an approach that integrates CBM that is disrupted by predicted failures that can be planned in advance. The outcome of the model demonstrates that CBM is less disruptive to the maintenance planning than CM, because the time to failure gives the model flexibility to find the ideal moment to perform CBM. Conclusions and recommendations of this study can be used for implementing CBM approaches for rolling stock.