Data-Driven Computation of the Accessibility Provided by Demand-Responsive Transport

Abstract

Conventional Public Transport (PT) cannot support the mobility needs in weak demand areas. Such areas could be better served by integrating, within PT, Demand-Responsive Transport (DRT), in which bus routes dynamically adapt to user demand. While the literature has focused on the level of service of DRT, it has overlooked its contribution to accessibility, which measures the ease of accessing opportunities (e.g, schools, jobs, other residents). Therefore, the following simple question remains unanswered: How many additional opportunities per hour can be reached when DRT is deployed? However, no method exists to quantify the accessibility resulting from the integration of conventional PT and DRT. We propose a novel method to compute isochrone-based accessibility. The main challenge is that, while accessibility isochrones are computed on top of a graph-model of the transport system, no graph can model DRT, since its routes are dynamic and stochastic. To overcome this issue, we propose a data-driven method, based on the analysis of multiple days of DRT operation. The methodology is tested on a case study in Acireale (Italy) simulated in Visum, where a many-to-many DRT service is integrated with a metropolitan mass transit. We show that, regarding the currently deployed conventional bus lines, the accessibility provided by DRT is much higher and its geographical distribution more equal. While current DRT planning is based exclusively on level of service and cost, our approach allows DRT planners and operators to shift their focus on accessibility.