In recent years, auto manufacturers have begun to introduce Level 2 automation systems such as lane centering assistance, automatic braking, and automated cruise control into new cars. Despite the fact that these systems require drivers to constantly monitor the vehicle, some auto manufacturers may be inappropriately marketing these systems using terms such as “autopilot.” Researchers and safety advocates have raised concerns that such marketing may give drivers a false impression of these systems' capabilities and lead to risky driving behavior.

A research team at the American Automobile Association (AAA) Foundation for Traffic Safety, based in Washington D.C., sought to test if marketing and branding changed how drivers perceived and used Level 2 (L2) automation systems.  

The team recruited ninety participants between the ages of 20 and 70 in the Washington D.C. Metro Area for this study. The team then randomly assigned participants to different experimental groups, where they received varying types of training on the L2 automation systems in use on the test vehicle. Participants either received training via a printed “quick-start guide,” a video, or an in-person training.

Furthermore, participants received one of two versions of each training. One type of training was less optimistic about the L2 system and emphasized “driver responsibility” and the L2 system’s limitations. The other type of training was more optimistic and emphasized the L2 system’s capabilities and “driver workload reduction.” The team then surveyed the drivers about their perceptions of the L2 systems after training but before driving.

Finally, participants drove the L2 equipped study vehicle along a highway in Maryland while a research team member recorded their actions. Along the route, a certain segment of the highway is known to cause problems with L2 automation systems, and thus all drivers experienced an unexpected failure of the L2 system during the test run. The team then surveyed drivers again about the systems after the test run.