Related Experiment Video
Updated: Jun 20, 2026
07:08
Estimate the Cognitive Load Using Electrocardiographic Measure: A Human-AI Collaborative Task
Published on: December 5, 2025
Predicting driver fatigue using HRV measures and machine learning
Maya Arlini Puspasari1, Annisa Marlin Masbar Rus1, Danu Hadi Syaifullah1
1Department of Industrial Engineering, Faculty of Engineering, Universitas Indonesia, Depok, Indonesia.
Frontiers in Sports and Active Living
|June 19, 2026
Summary
Prolonged driving increases physiological fatigue, impacting autonomic balance. Combining heart rate variability (HRV) with machine learning accurately detects driver fatigue, aiding real-time monitoring systems.
Area of Science:
- Physiology
- Transportation Safety
- Machine Learning
Background:
- Driver fatigue is a major cause of traffic accidents, linked to sleep deprivation and extended driving.
- Objective physiological measures and subjective fatigue scales are crucial for assessing driver alertness.
Purpose of the Study:
- To investigate physiological and subjective fatigue responses during simulated driving.
- To evaluate the effectiveness of Heart Rate Variability (HRV) combined with machine learning for fatigue detection.
Main Methods:
- Forty participants underwent simulated driving sessions under normal and sleep-deprived conditions.
- Heart Rate Variability (HRV) metrics, Karolinska Sleepiness Scale (KSS), and Rating of Fatigue (ROF) were recorded.
- Logistic Regression and ensemble learning (XGBoost) were used for fatigue classification.
Main Results:
- Increased driving duration correlated with autonomic imbalance (lower mean RR, mean HR, RMSSD, HF) and higher fatigue levels.
- Sleep duration significantly impacted subjective fatigue (ROF, KSS) but not most HRV indices.
- XGBoost model achieved 77% accuracy in fatigue classification, identifying Mean RR, mean HR, and LF as key predictors.
Conclusions:
- Integrating HRV metrics and machine learning enhances driver fatigue detection.
- Objective and subjective measures are vital for evaluating driver fatigue, especially under sleep restriction.
- Findings support the development of real-time driver fatigue monitoring systems.
