Balancing convenience and sustainability in public transport through dynamic transit bus networks✩

The rapid transformation of urban mobility fueled by the evolution of digital technologies has enabled smarter and more convenient modes of transportation. However, these new modes (e.g., ride-hailing and ridesharing) are not always more sustainable than traditional modes, such as public transport networks. In this work, we propose a dynamic transit bus system that aims at combining the higher convenience provided to passengers in on-demand systems with the sustainability of public transport. The goal is to attract more passengers to the transit service by offering reduced walk-to-station distance and total travel time compared to the fixed lines. Our proposed system plans the routes and intermediate stops of the transit lines based on the actual observed demand. Due to the complexity of the resulting optimization problem under investigation, we propose a solution method which splits the problem into four stages: clustering, initialization, optimization, and merging. We compare our proposed system in two settings: flexible (i.e., all intermediate stations are dynamic), and semi-flexible (i.e., original fixed lines intermediate stations are kept at the route and more dynamic stations are added) to the fixed static line. Our results show that our proposed system offers a more sustainable option and can reduce the system-wide travel distance by attracting more on-demand passengers to the transit service by offering them reduced walking distance and travel duration compared to the static lines. We also quantify the positive impact of passengers' willingness to act more sustainably by accepting to walk more or arrive later to their destinations on the system-wide sustainability.