Publication Detail
User Acceptance of AI in Transport: The Case of SAE Level 3 Conditional Automated Driving
|
UCD-ITS-RP-25-97 Journal Article Electric Vehicle Research Center, UC ITS Publications |
Suggested Citation:
Nordhoff, Sina, Simeon C Calvert, Marjan P. Hagenzieker, Y. Lee, Natasha Merat (2025)
User Acceptance of AI in Transport: The Case of SAE Level 3 Conditional Automated Driving
. Transportation Research Part F 116This study applies an extended version of one of the most popular technology acceptance models, the Unified Theory of Acceptance and Use of Technology (UTAUT2), to predict user acceptance of SAE Level 3 conditional automated driving among more than 9,000 car drivers from nine European and non-European countries. We extend the model by two factors, trust and teaming, that we consider pivotal for user acceptance of conditional automated driving. We also investigate the factors impacting the determinants of acceptance and use of conditional automated driving, addressing a well-known gap in research. In this study we find that 40% of respondents did not intend to buy, and 39% of respondents did not express the intention to use conditional automated driving when available. 71% of respondents indicated a preference to stay engaged in the driving task to respond to requests from the car to resume manual control. The structural equation modeling analysis revealed that performance expectancy is the strongest predictor of driver’s behavioral intentions to use conditional automated driving, followed by trust and social influence. Contrary to common beliefs positioning trust as one of the most influential drivers of user acceptance of automated vehicles, the influence of trust on behavioral intention to use conditional automated driving is small. The availability of facilitating conditions supporting the use conditional automated driving (e.g., knowledge, getting help from friends, family, or car dealers) has a small influence on user acceptance. We also found significant effects of the factors impacting the determinants of acceptance and use. The effect of performance expectancy on hedonic motivation is positive, suggesting that the perceived usefulness positively enhances the perceived enjoyment. Similarily, the effect of social influence on performance expectancy and trust is positive, suggesting the social network of the individual plays an important role in promoting positive beliefs about the effectiveness of the technology and trust in the technology. Access to participation in the questionnaire was limited to respondents with access to internet, which is why future research should be performed with respondents without internet accessibility to examine differences in attitudes and acceptance between these internet-affine and less internet-affine groups.
Key words:
user acceptance, AI in transport, conditional automated driving, teaming, trust