Development of an EEG Headband for Stress Measurement on Driving Simulators

Area of Science:

• Neuroscience
• Wearable Technology
• Human-Computer Interaction

Background:

• Stress monitoring during driving is crucial for safety.
• Existing Electroencephalography (EEG) sensors often lack detailed accuracy specifications.
• Autonomous driving systems require evaluation of driver acceptance and cognitive load.

Purpose of the Study:

• To design, realize, and characterize a novel six-channel EEG wearable headband for measuring stress.
• To assess the impact of different autonomous driving algorithms on driver brain activity.
• To compare driver stress levels between manual and autonomous driving scenarios using EEG.

Main Methods:

• Developed a custom six-channel EEG wearable headband with WiFi transmission and 10-hour battery life.
• Performed full metrological characterization of the EEG sensor, noting a 6 μV error and 50 nV resolution.
• Conducted a driving simulator experiment with ten volunteers across manual, gentle autonomous, and aggressive autonomous driving scenarios.

Main Results:

• The EEG headband demonstrated a measurement error of 6 μV and a resolution of 50 nV.
• Beta wave power, indicative of stress, was significantly lower during autonomous driving (both gentle and aggressive) compared to manual driving.
• This study is the first to compare autonomous driving algorithms using EEG signals to assess driver acceptability.

Conclusions:

• The developed EEG headband provides accurate and reliable measurements of brain activity during driving.
• Autonomous driving algorithms appear to reduce driver stress compared to manual driving.
• The findings suggest higher driver acceptability for autonomous vehicles over traditional manual driving.