Crew Rostering in Long‐Distance Freight Railways: A Multi‐Depot <scp>VRP</scp> ‐Based Heuristic Approach

ABSTRACT Efficient crew management is a critical challenge in long‐distance freight rail operations, where shift limits, depot logistics, and labor regulations must be jointly considered. This paper addresses the Railway Crew Rostering Problem by proposing a novel mathematical model inspired by the Vehicle Routing Problem with Multiple Depots and Multiple Trips (VRP‐MD‐MT). The model incorporates depot assignment, shift structure, and overtime constraints. A tailored heuristic algorithm is developed and applied to real data from a major Brazilian freight railway. Compared to the company's manual planning, the proposed approach reduced the number of required drivers and significantly decreased overtime in several scenarios. The analysis indicates the potential for meaningful annual cost savings, achieved without reducing the workforce. Beyond cost efficiency, the method offers a standardized planning tool aligned with legal and operational constraints. These results demonstrate the model's practical relevance and computational effectiveness in complex railway environments.