Users’ Understanding of Automated Vehicles and Perception to Improve Traffic Safety –Results from a National Survey

This examines people’s perceptions and expectations of automated vehicle (AV) technologies. Although there is a relatively large volume of literature studying people’s perceptions of AVs, there has been limited attention paid to the source of their distrust and discomfort.

December 2019

Suggested Citation

Authors

Ipek Sener

Johanna Zmud

Madison Graham

Stanislaw Kolek

Abstract

Introduction

As development and testing of automated vehicle (AV) technologies grow, many advanced driver assistance systems are rapidly being deployed (Zmud & Reed, 2019). This movement is largely due to their potential benefits in reducing crashes and crash severity (e.g., Fagnant and Kockelman, 2015; Benson et al., 2018). However, empirical evidence for these safety benefits remains inconclusive (Sivak and Schoette, 2015; Strayer et al., 2017; Noy et al., 2018), and the public remains uneasy about these technologies.

Hence, the AAA Foundation for Traffic Safety conducted a nationwide survey assessing:

(1) People’s understanding of AVs;

(2) Their expectations and concerns about AVs, and;

(3) Rationales behind their distrust and discomfort toward AVs.

Key Findings

  • People generally perceived higher levels of vehicle automation as more effective in preventing crashes related to specific driving behaviors and situations
  • A degree of concern with AV technology increased as levels of automation increased
  • People’s concerns and distrust stem from unfamiliarity with the technology and perceived unreliability with current AV technologies

Figure 2. Trust in AV Levels to Reduce Likelihood of a Crash

Methodology

Data were collected in three phases: 1) pre-survey focus groups, 2) survey implementation and 3) post-survey follow-up interviews. For the pre-survey, focus group discussions gathered information regarding perceptions about AV technologies, in-vehicle information systems and traffic safety. The groups included 12 participants in Austin, Texas and 11 in Bethesda, Maryland that were equally balanced in age (18+), sex and ownership of vehicles with advanced driver assistance systems (e.g., adaptive cruise control, lane-keeping assist, blind spot warning, etc.). A two-hour guided discussion addressed topics including: 1) perceptions of driving safety, 2) awareness, knowledge, and perceptions of AV technologies and in-vehicle infotainment systems, and 3) reactions to descriptive videos about AV technologies.

For the survey implementation, it was administered as a part of the 2018 Traffic Safety Culture Index (TSCI), a nationally representative survey that identifies and assesses Americans’ attitudes and behaviors concerning traffic safety (AAA Foundation for Traffic Safety, 2019). As part of the 2018 survey revisions, a newly developed set of questions regarding AVs was included. A total of 3,349 respondents over the age of 16 (2,432 adults and 917 teens) were recruited using KnowledgePanel®, a probability-based online panel designed to be representative of the United States’ population (GfK, 2016).

Lastly, for the post-survey follow-up interview based on the preliminary results from the completed survey in Phase 2 (survey implementation), follow-up phone interviews were conducted on a subset of respondents (n = 93). Twelve analytical interview groups were formed to represent the overall sample in terms of geographical distribution, age, and sex. Interview questions probed for additional information about particular responses to previous questions, such as why an individual did not trust Level 5 automation in avoiding crashes.

Results

Although people generally perceived higher levels of vehicle automation as more effective in preventing crashes related to specific driving behaviors and situations, their degree of concern with AV technology increased as levels of automation increased. During post-survey interviews respondents indicated distrust and concern due to the demonstration of their unreliable performance and insufficient testing on today’s roads. Respondents noted that there were too many unknown road hazards that AV technologies would not be fully capable of detecting. Additionally, they expressed concern about system glitches or malfunctions, which could potentially cost someone’s life.

However, the present study results also showed that many respondents recognize the potential safety benefits when Level 5 automation is fully realized. Trust in higher levels of AV technology in reducing the likelihood of a crash increased as respondents’ level of understanding and experience with AV technology increased.

Suggested Citation

Authors

Woon Kim

Senior Analyst, Data & Information Group

AAA Foundation for Traffic Safety

Tara Kelley-Baker

Group Leader, Data & Information Group

AAA Foundation for Traffic Safety

Ipek Sener

Johanna Zmud

Madison Graham

Stanislaw Kolek