Autonomous vehicle fleets for public transport: scenarios and comparisons

Abstract

Autonomous vehicles (AVs) are becoming a reality and may integrate with existing public transport systems to enable the new generation of autonomous public transport. It is vital to understand what are the alternatives for AV integration from different angles such as costs, emissions, and transport performance. With the aim to support AV integration in public transport, this paper takes a typical European city as a case study for analyzing the impacts of different AV integration alternatives. A transport planning model considering AVs is developed and implemented, with a methodology to estimate the costs of the transport network. Traffic simulations are conducted to derive key variables related to AVs. An optimization process is introduced for identifying the optimal network configuration based on a given AV integration strategy, followed by the design of different AV integration scenarios, simulation, and analyses. With the proposed method, a case study is done for the city of Uppsala with presentation of detailed cost results together with key traffic statistics such as mode share. The results show that integrating AVs into public transport does not necessarily improve the overall cost efficiency. Based on the results and considering the long transition period to fully autonomous vehicles, it is recommended that public transport should consider a gradual introduction of AVs with more detailed analysis on different combination and integration alternatives of bus services and AVs. • A transport planning model with integration of autonomous vehicles. • Impact analysis of autonomous fleets integration in public transport. • Based on a typical European city with realistic traffic data. • Simulation study on different integration alternatives of autonomous vehicles. • Cost estimation and sensitivity analysis of different integration scenarios. • A mixed-mode with autonomous fleets and buses is better compared with a single mode. • The value of time of autonomous vehicles has big impact on the overall network costs.