Dial-a-Ride Problem Considering Urban Traffic Dynamics in Multi-Reservoir Systems

Although on-demand mobility services can provide more efficient and sustainable door-to-door transport options, they typically face challenges due to the complexity of incorporating traffic conditions and stakeholders' involvement in the decision-making process, particularly under dynamic traffic settings. This study proposes a multi-reservoir frame-work to maximize the overall profit of the dynamic dial-a-ride (DAR) problem while considering the dynamic nature of traffic conditions, real-time passenger requests and their sharing preferences. A dynamic DAR algorithm is developed and evaluated numerically through a case study, considering specific traffic scenarios. Numerical results demonstrate the practical applicability and operational efficiency of the proposed system. Managerial insights and potential research directions are also provided.