Exploring how physio-psychological states affect drivers' takeover performance in conditional automated vehicles.

Area of Science:

• Human-Computer Interaction
• Automotive Engineering
• Cognitive Psychology

Background:

• Automated driving systems require drivers to retake control in emergencies.
• Estimating driver takeover performance is crucial for safety and adaptive automation.
• Current methods lack a full understanding of the interplay between driver metrics and takeover behavior.

Purpose of the Study:

• To investigate the complex relationships between physiological, eye-tracking, and psychological metrics and driver takeover performance.
• To explore how trust and workload mediate the effects of physiological and eye-tracking data on takeover outcomes.
• To provide insights for developing advanced driver monitoring systems for automated vehicles.

Main Methods:

• A driving simulation experiment with 42 participants experiencing takeover scenarios.
• Collection of physiological (cardiac, respiratory, electrodermal activity) and eye-tracking data.
• Utilized structural equation modeling to analyze interactions among metrics, psychological states, and takeover performance (time and quality).

Main Results:

• Driver trust in automation positively correlated with higher takeover quality.
• Perceived workload was positively associated with longer takeover times.
• Physiological and eye-tracking metrics indirectly influenced takeover quality through psychological states like trust and workload.

Conclusions:

• Driver takeover performance is hierarchically influenced by psychological states, which are in turn affected by physiological and eye-tracking measures.
• Findings support the design of driver monitoring systems that integrate multiple data streams for accurate takeover performance estimation.
• This research contributes to enhancing driving safety in conditionally automated vehicles through better understanding of driver behavior during takeovers.