Driver Fatigue Detection for Microsleep Prevention
The aim of this project is to refine techniques for fatigue prediction that are based on pupil fatigue oscillation and EEG sleep spindle signals. Further, fatigue intervention techniques that alarm fatigued drivers using blue enriched light and trigeminus stimulation are to be refined. The hypothesis is that at least one out of the two techniques of fatigue prediction can forewarn the onset of a first 1-second sleep at least 8 minutes in advance. At least one out of the two interventions or a combination of them again can prevent the occurance of predicited microsleep in advance.
Therefore this study is relevant for driving safety. Increased safety as a result of technologies for a better fatique decrease reduce traffic accidents contigent of fatigue at the same time. In the scope of this project 172 subject were recruted which were under sleep deprivition for 27 hours. The subjects were devided into two controll groups and four different intervention groups. During the driving in a test car that has been developed and equipped with measurement techniques the subjects experienced real microsleeps. The according biosignals were analysed using ECG, OFT, wearables, eyetrackig, and speech-based methods.