Physiological and Physical Sensors for Stress Level, Drowsiness Detection, and Behaviour Analysis

Driver state detection system is designed with physiological and physical sensors. The proposed system conveys a way in which sensor data can be integrated to obtain the driver’s stress level, drowsiness, and behaviour analysis. Physiological and physical sensors data used are: Electrocardiogram (ECG) recordings of driver, head movements of driver which are recorded by accelerometer sensor, and hand touch positions of driver on steering wheel which are recorded with interdigitated (IDT) capacitive sensor. ECG recordings are decomposed through the synchrosqueezing-continuous wavelet transform (CWT-SST) based time-frequency technique for estimating the heart-rate variability which is one of the possible parameters for detecting the stress level of the driver. Accelerometer sensor data is used for drowsiness detection through CWT-SST. A flexible IDT sensor is designed for hand touch and position sensing on steering wheel rim. This data is analyzed for understanding the behaviour of driver while driving. The driver stress level detection accuracy achieved on stress recognition using automobile driver database is 91.82%, and drowsiness state detection on recorded test subjects is 97.30%. The proposed system shows the good capability in detecting different state of the driver than existing methods.