AIthena user needs assessment survey

To establish AIthena’s methodology to involve trustworthiness to AI development, explainability, ethics and privacy dimensions should be incorporated in a human-centric approach. Supporting this, user needs, expectations and concerns were gathered in 2023 through a survey designed by Rupprecht Consult in collaboration with several other AIthena partners.

Prior to designing the survey, research gaps were identified through a literature review. These included:

  • lack of a comprehensive understanding of user acceptance of autonomous vehicles and CCAM (i.e. Cooperative, Connected and Automated Mobility) technologies and AI,
  • need to explore the user interaction with the automated vehicles and AI functionalities, and
  • identification of factors required to build trust and transparency in an automated mobility environment involving AI.

To address the identified research gaps, a survey structure was designed involving the four use case partners in the AIthena project. The survey concept was presented in a workshop, followed by discussion on the needs of the use cases and classification of target groups. The main sections of the survey focused on the ‘mobility needs’, ‘expectations’, and ‘concerns’ of the user. While the user profile in the survey addressed sociodemographic and situational factors (such as their living environment, travel mode choices), it also focused on the user’s present level of understanding and interaction with the self-driving vehicles. For the understanding of survey respondents, self-driving vehicles were defined as ‘vehicles in which the occupants of the vehicle do not participate in driving activities but must be prepared to take over control if requested by the vehicle for safety, involving communication on an ongoing basis with other vehicles on the road and with their surroundings (e.g. with traffic lights), and obeying all traffic regulations (including speed limits)’.

Following the literature review and survey design, the survey was shared widely by all project partners. The Qualtrics platform was utilized from where the data was exported and later cleaned to rectify any errors (such as incomplete surveys). The survey analysis involved methods of descriptive statistics and simple correlations, along with ordered logistic regression for a selection of specific questions. The survey analysis included identification of interest areas of the four AIthena use case through a second workshop.  

Approximately 450 responses were collected in ten languages (including incomplete responses); many of the responses had a completion rate of at least 70%. The respondent profiles mainly reflected urban perspectives of young and middle-aged respondents, where a majority had a university level of education. In addition, most of the respondents had a driving licence and access to cars.

Based on the relationship of users to the self-driving vehicles at present or in the future, the identified respondent profiles included:

  • active mobility users such as pedestrians and cyclists,
  • users of a self-driving vehicle or a vehicle with automated functions,
  • decision makers or planners involved in decisions as to whether self-driving vehicles will be allowed, and
  • developers of self-driving vehicle technologies or services.  

The survey results reflect upon the mobility needs of the respondents by identifying:

  • travel behaviour through preferred travel modes for different activities (e.g., door-to-door commute, errands, first and last mile activities),
  • potential frequency of using self-driving vehicles for identified activities (such as errands, leisure activities, and first and last mile activities),
  • criteria to choose a mode of travel,
  • level of trust towards self-driving vehicles in different living environments (e.g., urban environment, suburban environment, intercity highways, rural areas), and
  • use of self-driving vehicles in different scenarios (e.g., owning a self-driving car, booking a self-driving shuttle service on a short notice, using automated public transport vehicle on fixed routes).

Considering the expectations of self-driving vehicles, survey results reflect:

  • priority of access to information for the occupants of self-driving vehicles, passengers of a self-driving shuttle/bus, and pedestrians or cyclists,
  • preferred activities in a self-driving vehicle,
  • opinion of respondents towards impact of self-driving vehicles in comparison to human-driven vehicles (e.g., expecting fewer accidents, less congestion, and shorter travel time), and
  • opinion towards vehicle certification.

Following mobility needs, and expectations, the survey addressed user concerns by tackling level of comfort, trust, and confidence towards self-driving vehicles in diverse situations (e.g., feeling comfortable with the ability of self-driving vehicles to interact with human-driven vehicles, trusting self-driving vehicle to be safe and reliable, or being confident of protection from cybersecurity breaches or data-theft). This was followed by identification of factors discouraging use of self-driving vehicles.

A detailed description of the survey results along with use case descriptions will be reflected in the upcoming AIthena project deliverable titled User Group Needs Report and Technical Use Case Definition in July 2024.