In this study, we propose a new model to analyze the data sensors evolution during different situation encountered by the driver.

This modelling add to the classic Semi Hidden Markov Model (HMM) framework a weight feature. We then used our modelling to identify the driver's aim and the driving situation he's in.

To assess the capacity of our modelling, we conduct an experiment which able us to record 718 driving sequences.

On these sequences, our modelling choice allows us to predict the driver's situation with a 85% success rate.

Moreover, this modeling gives some interesting results on the organisation of the driving activity. These results show the HMM effectiveness to model and predict drivers behavior.