On-Demand: Adaptive Routes for Public Transportation
Abstract
This article considers the concepts and principles of adaptive public transport: shuttles carrying passengers on request, without fixed stops, with the ability to adjust their routes on demand. The route calculation is based on a multifactor model, where the goal is to minimize the distance traveled by shuttle, waiting time and travel time. We describe algorithms to balance these parameters with data collected in Innopolis city (a satellite city of Kazan, the capital of the Republic of Tatarstan) where an on-demand shuttle has operated since June 2018. The results of comparative modeling of on-demand transport systems with different degrees of flexibility and specific road networks are presented. We compared technical and economic indicators and the quality-of-service parameters of models with different degrees of flexibility but with the same parameters for the number of vehicles and passenger traffic. The results show the efficiency of the combination of adaptive routes with varying degrees of flexibility within the service operation area. The optimal combination of routes was where some of the vehicles are allocated to serve well-predicted passenger flows, with small deviations from the main route; the rest of the vehicles are without route or schedule restrictions, moving to the areas and in the directions where demand increases at different periods.