Driving automation state-of-mind: Using training to instigate rapid mental model development

Area of Science:

• Human-computer interaction
• Automotive engineering
• Driver behavior analysis

Background:

• The automotive industry is progressing towards full autonomy, with partial driving automation representing an intermediate stage.
• Partial automation systems offer potential safety benefits but present challenges due to inherent limitations and the driver's evolving role.
• Safe operation requires drivers to fully comprehend the automation's capabilities and their responsibilities.

Purpose of the Study:

• To develop and evaluate a training program for drivers interacting with partial automation systems.
• To assess the impact of training on driver understanding and behavior in human-automation interaction.
• To investigate the effect of training on drivers' willingness to retake control in critical scenarios.

Main Methods:

• A proof-of-concept training program was designed for partial automation systems.
• A between-group mixed-methods simulator experiment was conducted to evaluate training effectiveness.
• Trained and untrained driver groups were compared on understanding and control-taking behavior.

Main Results:

• Trained drivers demonstrated a significantly improved self-reported and exhibited understanding of the automation system.
• Trained drivers showed a higher inclination to retake control in critical situations compared to untrained drivers.
• The study highlights the positive impact of targeted training on driver performance with partial automation.

Conclusions:

• Training is an effective method for enhancing driver comprehension of partial automation systems.
• Improved understanding through training can lead to safer human-automation interaction.
• Driver training is crucial for mitigating risks associated with intermediate levels of driving automation.