Explore the scholarly record.

Abstract

Integrating passenger and freight services offers an opportunity to improve public transport resource utilization. However, both demands exhibit distinct spatiotemporal patterns and service requirements, making it difficult for conventional fixed-capacity vehicles to accommodate both demand types flexibly. To address this limitation, this study proposes a modular autonomous vehicle based passenger–freight co-modal service (MAV-PF) that physically separates passenger and freight flows in dedicated MAVs while coordinating their operations through joint formation and scheduling. A mixed-integer programming model is formulated to integrate service timetabling with MAV movements, formations, and demand allocation. A tailored matheuristic based on Adaptive Large Neighborhood Search (ALNS) is developed to solve the studied problem. Extensive numerical experiments using Beijing transit data are conducted to evaluate the performance of the proposed MAV-PF. The results show that MAV-PF reduces operating costs by up to 65.75% compared with conventional co-modal services. Moreover, relative to separate passenger and freight MAV operations, integrated operation reduces the required passenger and freight fleet sizes by up to 22.22% and 33.33%, respectively. These findings show that flexible capacity adjustment and integrated MAV operation can improve capacity utilization while achieving a better balance among passenger service quality, freight delivery performance, and operating efficiency.