Predicting Driver Stress Levels with a Sensor-Equipped Steering Wheel and a Quality-Aware Heart Rate Measurement...

Area of Science:

• Automotive Safety
• Biomedical Engineering
• Human-Computer Interaction

Background:

• Camera-based driver monitoring systems face limitations due to lighting and accessories.
• Unobtrusive monitoring of driver mental states is crucial for transportation safety.

Purpose of the Study:

• To evaluate the use of physiological signals from steering wheel sensors for monitoring driver stress.
• To develop a method for extracting and analyzing electrocardiogram (ECG) signals from steering wheel sensors.

Main Methods:

• Proposed a modulation spectral signal representation to extract ECG signals from steering wheel sensors.
• Computed heart rate and heart rate variability features from the extracted ECG signals.
• Utilized a logistic regression classifier to discriminate between stressful and non-stressful driving conditions.

Main Results:

• Achieved up to 72% accuracy in discriminating between stressful and non-stressful driving conditions.
• The proposed method enables direct quality assessment of heart rate information.
• Demonstrated the feasibility of using steering wheel sensors for driver stress monitoring.

Conclusions:

• Physiological signals from steering wheel sensors can effectively monitor driver stress levels.
• The modulation spectral signal representation enhances the reliability of heart rate extraction.
• This approach offers a promising, unobtrusive solution for intelligent driver monitoring systems.