Fatigue is a problem with severe consequences for drivers and people who work in high-risk situations such as mining, operation of heavy machinery, and for workers in factories, chemical/nuclear plants, etc. Current methods to implement fatigue detection in automotive sector include techniques such as blink frequency detection, posture/movement analysis etc. However, the reliability of these methods is not very high due to varying lighting conditions and inherent person to person variation. Unlike factors like drunkenness which can easily be estimated using breath analysers, fatigue and drowsiness resulting from fatigue are relatively hard to detect. Hence, it is vital to have devices/systems to check a driver’s fatigue levels continuously.

Heart Rate Variability (HRV) is an emerging field of research for physiological monitoring. Variations in the heart rate can indicate the activity of the nervous system which can be a potential indicator of fatigue. As a part of sponsored project by M/s. Tereso Ventures Private Limited, Pune, CSIR-CEERI has developed a mathematical model to estimate fatigue from HRV. Subjects indulged in physical and mental activities and their Electro Cardiogram (ECG) waveforms were recorded using ECG sensors attached to a hardware board during the activity (Fig. 1). Data from an Emotiv EPOC Electroencephalogram (EEG) headset was used to provide an independent validation for fatigue levels. The data showed a rise in the Low Frequency to High Frequency Ratio (LF/HF ratio) of the heart rate variability when fatigue levels rose and hence this parameter was established as an indicator for the fatigue. Detailed validation of the technology for detecting fatigue of motorbike riders is currently being carried out by M/s. Tereso Ventures and the developed mathematical model for the estimation of fatigue by CSIR-CEERI is being transferred to the company.

Fig. 1: Prototype of fatigue detection system using captured ECG signals and Heart rate variability