Improving conventional transit services with modular autonomous vehicles: A bi-level programming approach

Modular autonomous bus technology is highly anticipated to be introduced into public transit systems with a gradually increasing market penetration, and it will be operated within a multimodal public transport system with a mixed fleet of autonomous buses and conventional human-driving buses. This study establishes a bi-level programming model for the coordinated optimization of hybrid transit system, considering transit company and passengers with conflicting objectives in passenger flow assignment and minimizing the systematic cost from operational and environmental perspectives. Results show that the integration of autonomous buses significantly improves the sustainability of the transit system, reducing operating costs by 25.1% and environmental costs by 27.72% compared to the conventional transit system. While reducing passenger waiting times, it also decreases carbon emissions of the current transit system. This study concludes with some managerial insights into the introduction of autonomous buses.