This study evaluated four automated shuttle projects in the U.S., Germany and Estonia, using corresponding surveys to gain insight into three aspects of survey development: survey population, survey approach, and questionnaire design. All projects were conducted between 2018 to 2020, with surveys administered either during or after the related project. The surveys were evaluated on the previously mentioned aspects of survey development. The projects were as follows:
- The first project was a 90-day automated shuttle pilot at Joint Base Myer-Henderson Hall in Washington, D.C., led by the U.S. Army Corps of Engineers Research and Development Center. A total of 154 paper and online surveys were distributed to riders and non-riders with questions about perceived safety.
- The second project piloted a 5-month automated shuttle service in Schöneberg, Berlin at EUREF's Office Campus. Passengers were given surveys concerning demographics, shuttle and service characteristics (e.g., perceived safety and enjoyment), environmental attitudes, and acceptance.
- The third project involved a one-week demonstration of an automated shuttle on the campus of the University of South Florida. Paper surveys were distributed to passengers on-board; passengers were asked about their trip experience, level of trust and comfort, and how future service would impact travel choices.
- The fourth project was a four-month pilot of a circulator service at a public park in Tallinn, Estonia. Passengers were directed to an online survey while non-passengers were recruited via the Tallinn University of Technology. Survey questions asked about perceived safety and security, overall trip experience, and ability to use the service.
The following are several lessons learned from the projects for the development of future surveys about automated shuttles as they relate to survey population, survey approach, and questionnaire design:
- Survey both passengers and non-passengers to gain insights on automated shuttle service. Non-passengers can offer insights on why they did not ride or on their experience sharing road space with automated shuttles, which may travel at very slow speeds and stop unpredictably.
- Tailor the survey approach to the analytical needs and enrich the analysis by use of mixed methods (e.g., focus group, supplemental data).
- Avoid asking about multiple factors (e.g., perceived safety and comfort) at the same time and clearly ask about one item per question.
- Clearly distinguish the questions related to the willingness to ride an automated shuttle in survey design as it may be influenced by several different elements. Describe the proposed service in detail when asking about proposed or future service(s).
- Ask questions about speed in several different ways. This may help improve the usefulness of results, as low speeds have a complicated relation to user acceptance.
- Include questions that allow for an assessment of the representativeness of the sample because the volunteer drivers/riders are often those who are interested in novel technology and are not a representative sample of the general public.