A Respiration Rate-based algorithm for drowsiness detection

Drowsy driving can lead to road accidents. Therefore, it is essential to find methods to prevent drowsy driving. The aim of this thesis is the implementation of a method for detecting drowsy driving using the respiration rate (RR). Data have been collected through a dynamic driving simulator. RR data has been measured with a thoracic respiratory band for the Polysomnographic test. Two real-time algorithms have been developed to detect drowsy driving from the dsRR (differential of the standard deviation) and the dmRR (medium value of the respiration rate) of a certain window of RR signal samples. The needed system transfer functions have been identified through the set membership identification approach. The algorithms can predict the first falling asleep in most of the cases.