How does training affect expectations and use of a vehicle self-driving feature?
Evaluating consumers’ understanding of partially automated driving systems
Challenge
New vehicles increasingly include features that allow the vehicle to control its speed and lane position without the driver doing anything. Drivers do not always receive accurate and complete information about these features’ capabilities and limitations, which can lead to unsafe behaviors and overreliance.
For the AAA Foundation for Traffic Safety, Westat conducted research on consumers’ understanding of partially automated driving systems.
Solution
Westat randomly assigned 90 participants to 1 of 6 training conditions. Each condition was a combination of a marketing/branding approach (to either emphasize the technology’s limitations or capabilities) and a training mode (printed quick-start guide; training video; or in-person, on-road demonstration).
The study participants
- Received training on use of a self-driving feature present as original equipment in a production vehicle
- Completed a post-training questionnaire assessing their mental model of system operation and capabilities
- Drove the vehicle on a freeway route using the feature for 35 minutes
- Completed a final questionnaire after the drive
The Results
A branding approach that emphasizes feature capabilities, compared to an approach that emphasizes limitations, led to greater confidence in the capabilities of the feature in ways that might lead drivers to over-rely on it or use it unsafely.
Participants were more likely to report willingness to engage in potentially distracting or risky behaviors in the condition that emphasized capabilities. The differences observed in drivers’ initial mental models often persisted after participants used the feature on the road.
The results show the importance of providing information about self-driving technologies that accurately reflect their capabilities and limitations.
