Which factors affect willingness-to-pay for automated vehicle services? Evidence from public road deployment in Stockholm, Sweden

• Main Authors: Chee, Esther Pei Nen, Susilo, Yusak O., Wong, Yiik Diew, Pernestål, Anna
• Other Authors: School of Civil and Environmental Engineering
• Format: Article
• Language: English
• Published: 2021
• Subjects: Engineering::Civil engineering; Automated Vehicle; Demand Responsive
• Online Access: https://hdl.handle.net/10356/147007
• Institution: Nanyang Technological University

Introduction: Travel demand and travel satisfaction of a transport service are affected by user perceptions of the service quality attributes, and such perceptions should be included in studying user willingness-to-pay (WTP) for automated vehicle (AV) services. This study applied structural equation modelling with service quality attribute perceptions as latent variables affecting WTP. Objectives: We investigated how WTP AV services are affected by socio-demographic characteristics, knowledge and experiences with AV, existing travel modes and particularly, perceptions of the associated service quality attributes. The AV services are: 1) on-demand personalised AV (PAV) service, 2) demand responsive shared AV (SAV) service, and 3) first−/last-mile automated bus (AB) service. Methods: The data were collected from 584 potential users of a first−/last-mile AB service trial operated in Kista, Stockholm. Results: Results show people hold different expectations towards each type of AV service. These expectations act as the minimum requirements for people to pay for the AV services. Respondents are found to be willing to pay more for PAV service if it is safe, provides good ride comfort, and is competitively priced relative to the price travelling by metro and train over a same distance. Other than service quality attribute perceptions, income level, existing travel modes for daily trips, familiarity with automated driving technology and AB ride experience are important factors affecting WTP for the AV services. Conclusion: The developed model can be applied to understand expectations of potential users towards a new AV service, and to identify user groups who are willing to pay the service. New AV services can thus be designed sensibly according to users’ actual needs.