Should optimal stop spacing vary by land use type? New methodology

This study aimed to establish whether optimal stop spacing should vary by land use type. A new stop clustering method was proposed with stops divided into catchments on the basis of land use. A bilevel optimization model was then used to suggest optimal stop spacing for these catchments. The upper-level problem minimized a cost function that represented the perspectives of the passengers, the operators, and the local authority, and the lower-level problem was the transit assignment problem. Results show that stop spacing in the catchments with residential usage is generally larger than in those catchments with nonresidential or mixed land use. Near-optimal solutions were proposed to accommodate local considerations of stop placement. The spacing interval created by near-optimal solutions could be used to adjust existing stop placement. A case study of Tram Route 109 in Melbourne, Australia, revealed that the practical implication of the proposed analytical approach indeed improved the practical operations of the tram service. In future studies, the model could be extended to a network of tram routes. Overall, the results suggest that optimal stop spacing on public transport should be a function of land use type.