Citation

Wang M, Chen S, Meng Q. Transp. Res. B Methodol. 2024; 184: e102965.

Copyright

(Copyright © 2024, Elsevier Publishing)

DOI

10.1016/j.trb.2024.102965

PMID

unavailable

Abstract

Autonomous buses offer a promising solution to the first- and last-mile problems, but their initial deployment requires safety drivers to supervise bus operations. Compared to traditional bus driver scheduling, the safety driver scheduling is more flexible and allows safety drivers to start/end their tasks at any bus stop/terminal. However, uncertainties such as delays in public bus services can bring new challenges to the flexible safety driver scheduling. To address these challenges, we put forward a robust safety driver scheduling problem (RSDSP) for autonomous bus services in order to achieve the trade-off between cost reduction and robustness against possible delays. The proposed robust optimization model takes into account the propagation, uncertainty, and correlation of delays. Besides, we develop both exact and heuristic solution methods to solve the RSDSP efficiently. We demonstrate the effectiveness of the proposed model and solution methods through the extensive numerical experiments in which the instances are generated based on the historical operational data of Singapore's bus lines. The results highlight the importance of incorporating robust optimization into the safety driver scheduling for autonomous bus services. The proposed methods can also be applied to other robust driver/crew scheduling problems in public transportation.

Keywords

Autonomous buses; Branch-price-and-cut; Delay propagation; Driver scheduling; Robust optimization