Impact of urban morphology on the reliability of electric on-demand feeder services

Urban morphology, the physical form and layout of the city including streets, transport infrastructure and land use patterns, highly in"uences the spatial distribution of travel demand. As urban morphology evolves (e.g., urban sprawl), so the travel demand patterns that tend to arise will change. This may impact upon the performance of Mobility on-Demand services such as demand responsive feeder services (DRFS). Electric DRFS (E-DRFS) may be particularly susceptible to perturbations in the distribution of demand; due to battery capacity constraints, the con!guration of charging infrastructure determines service performance, and it cannot easily be modi!ed to respond to demand changes. Dayto-day spatiotemporal variability in demand impacts vehicle routing and hence energy consumption, making "eet planning and charging management a challenging problem. To design and provide a reliable E-DFRS, this study aims to understand how urban morphology in"uences the reliability of E-DRFS given stochastic demand characteristics. Previous research has focused on the performance of mobilityon-demand services in terms of service area characteristics and perturbation in demand patterns. However, some only focus on the temporal perturbation while others investigate highly simpli!es case with one vehicle. Besides, none have focused speci!cally on the context of E-DRFS. To address these issues, this research aims to understand which dimensions of urban morphology bring systematic changes in the performance of E-DRFS, assuming synthetic but somewhat realistic scenarios. The main focus is the impact of urban morphology on service reliability under stochastic demand. The ensembles of simulation experiments are conducted to investigate the impact of changes in the distribution pattern and density of demand on the reliability of E-DRFS. Preliminary results suggest that systematic changes in some key performance indicators (KPIs) can be identi!ed, and the study is currently conducting additional simulation experiments to gain a better understanding of these complex interactions and identify systematic trends.