New Algorithm Improves Public Transport Routing Efficiency

A novel method known as Early Pruning has been introduced to enhance the efficiency of routing algorithms used in public transport systems. This approach specifically addresses performance issues in commonly utilized algorithms, such as RAPTOR and its variations, which often face challenges during the transfer relaxation stage on dense transfer graphs. Such challenges arise when accommodating unlimited transfers, as algorithms must evaluate numerous inter-stop connections, including walking, biking, and e-scooter options. To ensure satisfactory performance, many practitioners limit transfer distances or omit certain options, potentially diminishing path optimality and restricting multimodal choices for travelers. Early Pruning mitigates this by pre-sorting transfer connections based on duration and implementing a pruning rule in the transfer loop, enabling the elimination of longer transfers that won’t provide an earlier arrival than the best current solution. This low-overhead technique accelerates routing while maintaining optimality. The findings are detailed in the paper arXiv:2603.12592v2, which was released as a replace-cross type publication. The research aims to enhance the efficiency of public transport routing algorithms managing intricate multimodal networks.