Driving Simulator Study Found Participants Equipped with Level 3 Automation Experienced a Lower Likelihood of a Crash Occurring During Lane Change Maneuvers Compared to Drivers Without These Capabilities.
Researchers Examined the Lane-Changing Behavior of Drivers in Response to a Lane Closure in a Driving Simulation.
McLean, Virginia, United States
Lane-Change Response to Infrastructure Warning About Lane Closure in a Mixed Vehicle Fleet
Summary Information
Changeable message signs (CMS) are programable, electronic signs used to convey traveler information to drivers, but they may be problematic for automated driving systems (ADSs) to interpret. CMSs are frequently used to convey information about incidents and can help inform drivers about unexpected events. If ADS-equipped vehicles cannot interpret CMS messages, an information mismatch occurs, with the ADS having less information to act on than a human driver. The goal of the study was to examine the potential benefit of using cooperative driving automation (CDA) alerts to communicate messages to drivers that are traditionally disseminated via CMSs. This study recruited 96 participants and used a driving simulator, eye-tracking software, and questionnaire data to assess the potential benefits of CDA alerts on driver behavior. The study occurred at the Federal Highway Administration Turner-Fairbank Highway Research Center in May and June 2024.
METHODOLOGY
The simulated roadway was a 4-mile highway segment in which participants encountered a traffic incident that blocked the right lane. It had 12 scenarios based on vehicle automation, messaging source, and message type:
- No vehicle automation or SAE Level 3 (L3) automation (lane centering, adaptive cruise control, and automated lane changes).
- Messages by overhead CMS only, in-vehicle CDA alert only, or both.
- Instructional or informational messages (“Traffic Incident Ahead Slow Down” versus “Traffic Incident Ahead Right Lane Closed”).
The research team used questionnaires to collect participant demographics, familiarity with vehicle features, trust between people and automation, and post-drive feedback. In terms of limitations, the researchers noted that driving simulators may not always capture real-world driving behavior and the findings from the eye-tracking data were limited because of a small sample size (many participants did not view particular areas of interest).
FINDINGS
- Of the 96 participants, 12 (12.5 percent) experienced a collision with another vehicle during their lane change in the left lane. L3 automation was associated with 84 percent lower odds of a crash occurring during the lane-change event (odds ratio = 0.16).
- Compared to participants who only received CMS information, participants who received information from both message sources (CDA and CMS):
- Traveled a longer distance during the lane change (158.17 feet longer).
- Took longer to complete the lane change (2.46 seconds longer).
- Displayed less variability in steering wheel angle during the lane change.
These findings suggested smoother and possibly safer lane-change behaviors by drivers experiencing both CMS and CDA alerts, supporting the hypothesis that these drivers were more likely to modify their driving behavior in response to the messaging.
