Mixed fleets of automated and human-driven vehicles in public transport systems: An evaluation of feeder line services

Abstract

This study focuses on the transitioning period of operating mixed fleets of both automated and human-driven vehicles for public transit services. The type of service investigated here is flexible, including elements of both fixed route and on-demand systems. The operation of the mixed fleet is optimized with analytical methods leading to models for optimal service headway and stop spacing for the two types of vehicles. Analytical models for optimal passenger capacity per vehicle and required fleet size for each type of vehicles are also derived. Four operational strategies are considered, referring to whether the two types of vehicles operate jointly or independently in terms of optimal service headway and stop spacing within the mixed fleet. Numerical analyses indicate that automated vehicles operate optimally with less frequent vehicle dispatches and more fixed stop locations compared to human-driven vehicles. They also require greater fleet size and similar passenger capacity per vehicle. The four operational strategies perform similarly in terms of total generalized costs for the input values considered here. However, sensitivity analyses showed that the operational characteristics of the two types of vehicles in a mixed fleet and the performance of the four operational strategies depend significantly on the percentage of total demand that each type of vehicle serves, as well as on the automated vehicles’ speed and in-vehicle travel time cost for users. The mixed fleets represent the transitioning period towards transit fleets of automated vehicles only and it is shown to be the costliest period for both users and operators.