Redesigning Today's Driving Automation Toward Adaptive Backup Control With Context-Based and Invisible Interfaces

Area of Science:

• Human-computer interaction
• Automotive safety
• Driver monitoring systems

Background:

• Human supervision of automated driving is challenging due to limitations in attention and vigilance.
• Driver monitoring systems (DMS) can act as a backup, retaining human control while engaging automation when necessary.
• Minimizing false alarms and unannounced backup status in DMS may mitigate distrust and overreliance issues.

Purpose of the Study:

• To investigate the feasibility of adaptive driver monitoring systems (DMS) for backing up distracted drivers during automated driving.
• To evaluate different DMS backup control strategies, including eyes-only versus eyes-plus-context and visible versus invisible backup.
• To assess the impact of these systems on driver distraction, performance, and satisfaction.

Main Methods:

• A driving simulator study with 91 participants was conducted.
• Participants experienced various conditions: supervised automated driving (hands-on/off-wheel), DMS-induced backup control (eyes-only/eyes-plus-context, visible/invisible), and conventional driving.
• A continuous visual N-back task was used to induce and measure distraction.

Main Results:

• Supervised automated driving increased distraction and hazard non-responses compared to backup and conventional driving.
• Eyes-plus-context backup improved control and satisfaction compared to eyes-only backup, without negatively impacting safety.
• Invisible backup showed no appreciable negative consequences, while eyes-plus-context backup achieved satisfaction levels comparable to full automation.

Conclusions:

• Driver monitoring systems (DMS) can be designed to adaptively back up distracted drivers in automated driving scenarios.
• Incorporating driving context for distraction assessment and suppressing status indication are feasible design elements.
• Appropriately designed DMS can enhance the benefits and safety of automated driving systems.