This paper reports the work carried out under a European project - CEMVOCAS(Centralised Management of VOCal interfaces aiming at a better Automotive Safety) - centred on the development of a system to manage vocal communications to or from the driver according to the driver's mental availability to process additional information. Most studies centred on road accident analysis stress the importance of attention, particularly divided and sustained attention. Moreover, a lack of attention due to a low vigilance (monotonous driving, night driving) or the performance of an additional task while driving in a complex environment are the basis of many road accidents. Therefore, the same methodology that has been used to define the three levels of driving situation complexity and develop the Driving Situation Recognition Unity (DSRU) could be used to identify the signs of drowsiness and decline of the driving task performance on the same basis: the driver's actions or lack of action. Then, additional sensors to identify the alertness related parameters should be defined, aiming for the development of a DSRU able to identify driver attentional demands in extreme situations and to generate relevant actions to ensure road safety. (A)
Abstract