A systematic review of physiological signals based driver drowsiness detection systems

Area of Science:

• Biomedical Engineering
• Transportation Safety
• Machine Learning

Background:

• Drowsiness significantly impairs driving ability, leading to accidents, injuries, and fatalities, particularly among young drivers and night shift workers.
• Accurate, real-time driver drowsiness detection is essential to mitigate risks associated with drowsy driving.
• Existing detection methods often analyze vehicle dynamics, driver behavior, or physiological measures.

Purpose of the Study:

• To systematically review recent techniques for driver drowsiness detection using physiological signals.
• To analyze the advantages and disadvantages of various sensor types and experimental setups.
• To identify research gaps and propose future research directions in physiological signal-based drowsiness detection.

Main Methods:

• Comprehensive literature review of recent studies on driver drowsiness detection.
• Analysis of techniques employing physiological signals and machine learning algorithms.
• Evaluation of studies based on sensor types, data collection environments (controlled vs. dynamic), and experimental setups (real traffic vs. simulators).

Main Results:

• Physiological signals combined with machine learning show promising results for drowsiness detection.
• Different sensors and experimental conditions impact the effectiveness and applicability of detection techniques.
• Existing studies reveal limitations in sensor technology, data diversity, and real-world validation.

Conclusions:

• Physiological signal-based drowsiness detection is a rapidly advancing field with significant potential for improving road safety.
• Further research is needed to address current limitations and develop robust, real-world applicable systems.
• Future directions include exploring novel sensors, advanced machine learning models, and diverse driving scenarios.