Multi-parametric analysis of sensory data collected from automotive drivers for building a safety-critical wearable computing system

We present a methodology and an off-line analysis of data collected from several body-mounted sensors for automotive drivers during multiple fixed-route, segmented driving on real roads. This analysis was used for a range of preliminary processes including those related but not limited to stress detection. It also attempts to illustrate cases of several drivers for whom the real-time data acquisition was carried out through online physiological sensors. This analysis is to be used to designing and building a wearable computer involving Body Sensor Network for avoiding road accidents. In order to estimate the mental and physical fatigue to which a driver may be subjected to, we collected GSR, SpO2, Respiration, and ECG signals during relaxed and stressful driving scenarios.