Simulation based distracted driver detection system using alcohol sensor and fusion of bioelectric signals to disseminate them through V2V/V2I to alleviate risky driving

Abstract

The rampant road accidents due to drowsy or drunk driving and its dire consequences of losing countless life including making million others maimed each year in the whole world, which claims the attention of all concerned to address this issue seriously. The fatalities occur as consequences of frequent road crashes due to drowsy driving in Bangladesh are also increasing in an unbridled way. Undoubtedly, the ubiquity of road accidents has become a perennial problem from which we need a feasible way out. Current research on using Body Sensor Network inspired us in developing a system that can disseminate the physiological state of the driver while driving to other nearby vehicles or emergency unit who can respond immediately to halt the affected vehicle. Prior to communication, detection of drunk or drowsy state is a necessity. Bioelectric signals like ECG and EEG contain the feature for different physical and mental states which can be interpreted through some parameters with accuracy. Extraction of the desired feature from the driver’s ECG and EEG would certainly be capable of detecting the physical and the mental state of the subject. Although alcohol intoxication changes the ECG pattern, there are more convenient ways like using alcohol sensors to detect drunken state. Accumulating all the features collected from the subject, it can be sent to other vehicles and the emergency unit using VANET system. The proposed developed system generated ECG and EEG in simulation and extracted the necessary features from the subject and successfully combined these signals for communication. In addition, drunken state of the driver is also detected in this model. The idea of fusing bioelectric signals for distracted driving detection may contribute to reduce the number of uncontrollable road accidents.