Reducing vehicle maintenance using condition monitoring and dynamic planning

Abstract

It is a common fear in industry that introducing condition based maintenance (CBM), with its constant monitoring of several subsystems, will lead to more frequent service interventions compared to traditional cyclic maintenance, effectively countering the potential value of implementing CBM. Because of this, adoption of CBM must be done with great care, and the maintenance organization and planning process needs to be geared for more flexibility. To harvest the potential value in CBM for rail vehicles, we propose to combine condition monitoring with online maintenance planning. We use an adaptive planning software module to quickly find new suitable vehicle movement plans, and a heuristic packing module to reconstruct maintenance packages with as few maintenance stops as possible. This prevents vehicles from visiting the maintenance depot too frequently. At the same time, we actively keep the risk of breakdowns low. Evaluation of our methods in a simulated environment for train operation and maintenance, using real-world time tables and vehicle plans, show that by taking the operation times of individual components into account, it is indeed possible to reduce the amount of maintenance as well as the number of service interventions significantly compared to traditional cyclic maintenance.