Impact of Information on Consumer Understanding of a Partially Automated Driving System

Abstract

Introduction

Partial driving automation technologies, also known as active driving assistance or Level 2 (L2), allow vehicles to perform some aspects of the driving task, such as maintaining speed, lane position, and distance from the vehicle ahead, without direct input from the driver. The safe use of these systems depends on the driver having an accurate understanding of the capabilities and limitations of the system

There is growing concern that marketing materials and other information provided by automobile manufacturers, the media, and other sources may lead consumers to develop unrealistic expectations of the extent to which a vehicle with such technology is capable of safely “driving by itself.”

The purpose of this study was to explore how the information given to drivers about an active driving assistance technology influences their initial beliefs about and expectations of the system, and by extension, how they interact with it.

In this study, researchers introduced study participants to an actual active driving assistance system with a fictitious name, provided brief training about how to use the system, and then examined how the training influenced participants’ beliefs and behaviors. Half of the participants received information that called the system “AutonoDrive” and emphasized the system’s capabilities and driver convenience; the other half received information that called the same system “DriveAssist” and emphasized the system’s limitations and driver responsibility.

Key Findings

The training that emphasized system capabilities and driver convenience (“AutonoDrive”) led to greater confidence—and in some cases overconfidence—in the system, relative to the training that emphasized system limitations and driver responsibility (“DriveAssist”). For example, participants trained about AutonoDrive were significantly more likely to believe incorrectly that the system could detect and respond to a variety of hazards, such as a construction worker standing in the road or a vehicle directly to the side encroaching into its lane, even though all training materials explicitly stated that it could not.

Participants who received training about AutonoDrive were also more likely to report willingness to engage in potentially distracting or risky behaviors (e.g., at talk on a cell phone) while using the system, compared with those who learned about DriveAssist.

When actually driving a vehicle equipped with the system, participants who received the training that emphasized the system’s capabilities spent more time with their hands away from the steering wheel and feet away from the pedals, compared with those who received the training that emphasized limitations. They were also more likely to take a very long time (5+ seconds) to respond when the system disengaged unexpectedly and returned speed and steering control to the driver.

Many of the differences in participants’ understanding of and confidence in the system observed before the drive persisted, and in some cases even increased, after driving the vehicle.

Results underscore the importance of providing consumer-oriented information that is not only technically accurate but also balanced, with appropriate emphasis given to the limitations of technology and the importance of driver engagement.

Methodology

Researchers recruited 90 licensed drivers, with no prior experience driving a vehicle with a Level 2 partially automated driving system, from the Washington, D.C. metropolitan area.

For the study, researchers gave study participants brief training on how to use an actual SAE Level 2 active driving assistance system with a fictitious name, using training materials developed by the researchers. Half of the participants received information that called the system “AutonoDrive” and emphasized the system’s capabilities and driver convenience; the other half received information that called the same system “DriveAssist” and emphasized the system’s limitations and driver responsibility. Despite the differences in emphasis, neither version omitted any important safety information nor presented any false information.

After completing the training, participants completed a questionnaire about their understanding of the system, its capabilities and limitations, and their willingness to drive with versus without it under various scenarios. After completing the questionnaire, respondents drove a vehicle equipped with the system on which the training materials were based (Cadillac Super Cruise) approximately 31 miles on a divided, limited access highway with a 60 mph speed limit with a researcher present in the front passenger seat to instruct participants, take notes, and ensure safety. Various driver behavior and performance data, such as the vehicle’s speed and the position of the drivers’ hands and feet, were recorded by cameras and sensors installed in the vehicle. After completing the drive, participants completed another questionnaire that asked many of the same questions as the previous questionnaire regarding their understanding of the system, as well as some questions about their overall impression of the system after having driven with it. Questionnaire responses and driving data were compared between respondents who were given information about “DriveAssist” versus “AutonoDrive.”