On-demand transport service evaluation considering the impact of traffic congestion: a real case study

Demand Responsive Transport (DRT) has been deeply analyzed as a promising solution to fill gaps of existing public transport systems. In particular, its flexibility allows to replace traditional transit services, for people with specific needs, in low-demand periods or areas, and to complement it, serving as transit feeders. Recently, increased attention for DRT has been registered, because of the emerging diffusion of automated driving technologies, which can be implemented for a service operating in urban areas with autonomous vehicles. Despite the thriving literature about DRT systems, the effects of real traffic congestion on service efficiency and quality have not been explicitly modelled. In this work, a methodological/experimental approach is developed and applied to a real case study to evaluate the effectiveness of a DRT service for people with reduced mobility, considering the dynamic effects of traffic conditions in an urban area. The procedure is applied to the city of Padova (Italy), in particular, real requests and the static traffic simulation model matrix are used as input for a DRT planner. Then, the reassessment of the derived plan is made considering the output of a dynamic traffic simulation model. The results highlight that neglecting congestion effects could lead to biased results. In addition, including traffic dynamics to estimate route travel time of DRT vehicles allows to obtain a more effective and reliable assessment of the quality of the service.