Towards Rolling Stock Preventive Maintenance Scheduling - Short-term Scheduling Optimisation

Rolling stock preventive maintenance (PM) is a key element in ensuring a safe and reliable railway system. PM, also known as periodic maintenance, is based on the age (distance or mileage) of the train, where the intervals are specified carefully for each type of maintenance. PM scheduling optimisation is required to maintain the planned level of safety and a reasonable maintenance cost. Despite the benefits of such optimisation, this process, as of today, is still largely a manual process relying on human expertise. However, with the increased complexity of the PM scheduling process, the human brain reaches its natural limits causing increased PM costs and sometimes less safe and reliable railway systems. Existing relevant academic research contributions are more theoretical than practical studies, and most of them were attempting to change railway practices while the industry needs a smooth transition from manual to automated processes and then to more sophisticated smart maintenance. To overcome the above limitations, this paper describes the rolling stock PM scheduling process along with a short-term PM scheduling model. The proposed MIP model aims to automate the manual short-term PM scheduling process while maintaining a good level of safety, a reduced maintenance cost, and acceptable fleet availability.