Robust rescheduling and holding of autonomous buses intertwined with collector transit lines

The use of autonomous buses increases, and their operations increasingly intertwine with other public transport lines. At the same time, the use of autonomous buses in a mixed traffic environment induces new challenges with respect to their service synchronization with other lines. In fact, due to their cautious driving behavior in a mixed environment, the operations of autonomous buses show a high degree of variation in inter-station travel times, which leads to many unsuccessful transfers. Many tactical planning problems, however, naively assume that this uncertainty can be reduced to a single deterministic value (e.g., a travel time expectation). In this study, we propose a technique to find a robust schedule of dispatching and holding times that incorporates the inherent uncertainty in the operations. We discuss the complexity and applicability of such robust problems in public transit operations in general, and solve the optimization problem for an autonomous bus system that serves as a feeder line to a collector line. The numerical experiments show that the robust rescheduling and holding solution is resilient to variations in travel times: the operations maintain a high level of service regularity and avoid missed passenger transfers even under extreme scenarios.