This paper describes a part of the European project CEMVOCAS (CEntralised Management of VOCal interfaces aiming at a better Automotive Safety) project No. 25589 supported by the ESPRIT program, conducted by Matranortel communication and including Fiat, Renault, AKG, Metravib RDS, INRETS and Lisbon University. The part of the system developed in this paper is the DSRU (Driving Situations Recognitton Unit) which evaluates the availability of the driver to receive vocal messages during the driving task. To design this part of the system, some experiments have been carried out in France and in Portugal and the collected data have been coded in terms of driver availability. Then, this coding has been estimated by a neural network using the dynamic parameters of the vehicle. This paper is composed of two parts: the first one describes the experiments, the second one describes the DSRU itself and the first prototype evaluation results. (A*)
Samenvatting