A statistical analysis of consumer perceptions towards automated vehicles and their intended adoption

ABSTRACT

While automated vehicle (AV) development continues to progress rapidly, how the public will accept and adopt automated vehicles remains an open question. Using extensive survey data, we apply cluster analysis to better understand consumer perceptions toward potential benefits and concerns related to AVs with regard to factors influencing their AV adoption likelihood. Four market segments are identified – ‘benefits-dominated,’ ‘concerns-dominated,’ ‘uncertain,’ and ‘well-informed.’ A random parameters multinomial logit model is then estimated to identify factors influencing the probability of respondents belonging to one of these four market segments. Among other influences (such as socio-economic and current travel characteristics), it is found that ‘Millennials’ have a higher probability of belonging to the well-informed market segment, ‘Gen-Xers’ with a lower probability to the uncertain market segment, and ‘Baby Boomers’ with a higher probability to the concerns-dominated market (relative to the ‘Great Generation’). We also study the individuals’ expressed likelihood of AV adoption using separate random parameters ordered probit estimations for each of the four market segments. The substantial and statistically significant differences across each AV consumer market segment underscore the potentially large impact that different consumer demographics may have on AV adoption and the need for targeted marketing to achieve better market-penetration outcomes.

KEYWORDS:

• Autonomous vehicles
• consumer adoption
• random parameters ordered probit model
• AV users market segmentation

Acknowledgments

The contents of this report reflect the views of the authors, who are responsible for the facts and the accuracy of the information presented herein. This document is disseminated in the interest of information exchange. The report is funded, partially or entirely, by a grant from the U.S. Department of Transportation’s University Transportation Centers Program. However, the U.S. Government assumes no liability for the contents or use thereof.

Disclosure statement

No potential conflict of interest was reported by the author(s).

ORCID

Nikhil Menon http://orcid.org/0000-0002-7035-204X

Yu Zhang http://orcid.org/0000-0003-1202-626X

Abdul Rawoof Pinjari http://orcid.org/0000-0002-3056-4259

Fred Mannering http://orcid.org/0000-0002-2803-4582

Notes

1 Although data from these two sources does not represent a national sample, our intent was exploratory in nature. It is important to note that even if the analysis was based on a national sample, issues would arise because consumer preferences for new technologies tend to be highly volatile. In fact, Menon et al. (2016) and Sheela and Mannering (2018) show that preferences for autonomous-vehicle adoption have a high degree of temporal instability that would mitigate the benefits of a more representative survey. Thus, the focus on this selected sub-sample of potential consumers was intended to provide some initial insights and a demonstration of a methodological approach that is simple and easily accessible for empirical researchers and yet can be used to guide future studies on the subject.

2 This is supported by the empirical work of Menon et al. (2016) and Sheela and Mannering (2018), who demonstrated the importance of initial perceptions with regard to autonomous vehicle adoption.

3 It is important to note that our sample tends to be more highly educated and wealthier than the population as a whole. The fact that our sample is not representative of the population as a whole does not present an issue for model estimation since having explanatory variables not reflective of the population does not adversely affect parameter estimates. However, it is important to note that using a non-representative sample for market forecasting would result in incorrect adoption-behavior forecasts.