Path Choice Calibration Across Urban Road Networks using Routing Services

In the field of transportation modelling, achieving cost-effective routing solutions is fundamental. This study presents a pioneering approach to path choice calibration in urban road networks using OpenStreetMap (OSM) data, guided by benchmark routes from a commercial routing service. Given the significant costs associated with commercial routing services and the extensive data required for traditional path choice behavior analysis, our research introduces a novel set of parameters termed "routing factors" that adjust the transport costs on specific link types to align more closely with conventional routing choices. By exploring the range of these factors for different link types and determining the minimal input points required for stable calibration across several cities, we provide a cost-effective solution that reduces both computation time and financial expenditure. The calibration process utilises a random search approach alongside the Root Mean Square Error (RMSE) metric to evaluate the similarity between OSM-generated routes and the benchmark routes. The findings has proven adaptable across different network types, as confirmed through case studies in Brescia, Turin, Rome, and Milan. Our key findings was that using open-source data for routing problems requires multiplying the transport cost by a routing factor in the range of 1.5 to 4 for certain link types to match commercial services. This emphasises the importance of our calibration approach for accurate and reliable routing solutions. By extending our analysis to various network sizes, we aim to inform the decision-making processes of small logistics companies, potentially helping them reduce costs and optimise resource allocation.