Methodology and System of Evaluating the Driver's Vigilance Level in an Automobile Transportation Examining Both Physiological and Mechanical Data

This paper deals with the methodology followed in order to design a new intelligent system to improve the driver’s safety in an automobile transportation and the actual realization of a first prototype. The results of the study are reported. A simulator system has been developed at the Robotics Laboratory of the “Politecnico di Milano”. A description of the necessary hardware and architecture is made in detail. Driver’s physiological data, acquired from sensors on the wheel, is correlated, using statistical multivariate analysis, with his/her vigilance level evaluated using polysomnography. This statistical model is applied on the data off-line in order to define a controller, to be applied on real time acquired data. The platform’s mechanical data is also acquired and studied. All the elaboration of the data results in one vigilance level index for the current driver and situation. Future steps and possibilities are also discussed.