The US Department of Transportation has released a paper summarizing the results and implications of past research on the dynamics of consumer adoption of automated vehicles.

The paper, “Understanding Surveys of Public Sentiment Regarding Automated Vehicles: Summary of Results to Date and Implications of Past Research on the Dynamics of Consumer Adoption,” explores the public’s outlook on automated vehicles, focusing specifically on attitudes regarding safety, trust, and willingness to try, as well as the factors that influence those opinions.

It compiles the results of numerous surveys and studies conducted over the past 4 years and tracks consumer attitudes over time, against the backdrop of newsworthy events in the development, testing, and early deployment of automated vehicles.

The paper also presents an overview of research on the factors that affect consumer technology adoption, particularly in the context of innovative technologies, and identifies implications of this research for future assessments of the public’s interest in automated vehicles.

The authors reviewed 22 surveys conducted by 15 entities/partnerships, several of which recur annually. The authors noted the following general observations based on this review:

• Survey results should be assessed cautiously – Survey methods vary, and in some cases may provide insufficient and/or biased context to respondents. The authors encountered a range of terms used to describe automated vehicles to survey respondents, as well as definitions of those terms that were, in some cases, sparse, incomplete, inaccurate, or potentially biasing. Surveys that present minimal or no descriptions or definitions of the terms used risk respondents interpreting them inaccurately. On the other hand, some surveys that present more thorough definitions risk biasing respondents’ views on automated vehicles.

• Vehicle automation may be too novel to reveal reliable survey results – The authors reviewed several studies on relevant considerations for gauging public interest in a novel technology. Prevailing literature suggests that a priori preferences for a “really new product” (in contrast to an “incrementally new product”) tend to be unstable and easily changed by small adjustments in the measurement procedure. Specifically in the context of automated vehicles, one study notes: “… due to the lack of wide-scale deployment and use of AVs by the general public, [large-scale surveys] cannot provide survey respondents with an accurate overview of the capabilities and limitations of the new technologies in question. It is also difficult for the surveys to capture an accurate understanding of users’ actual use and impression of these vehicles. Results are therefore likely to be influenced by the interviewer’s descriptions, or rely on respondents’ impressions, which are, for example, shaped by promotional videos.”

• The surveys reviewed suggest that interest in automated vehicles is mixed, and leans negative – While potentially subject to biases as discussed above, surveys of consumer sentiment regarding automated vehicles still provide useful insight. Across most surveys that asked respondents about their willingness to try riding in an automated vehicle, slightly more than half of responses were negative (i.e., respondents were not interested in or were concerned about trying automated vehicles). The American Automobile Association (AAA) has been surveying the public since 2016 about their willingness to ride in an automated vehicle, and across five instances of this survey, between 63 and 78 percent of respondents have indicated that they would be “too afraid to ride in a fully self-driving vehicle.” Interest levels in buying or regularly using an automated vehicle are fairly consistent with interest in trying them. Several surveys have provided context to these findings, discovering that more than half of respondents perceive that automated vehicles are not safe.