Our goal is to improve drivers' awareness, i.e., the perception of the own driving from current events in the environment as well as anticipation into the close future. For that, we propose a methodology called careful, conscientious and contextualised drive (3C-DRIVE). The modelling of drivers' mental behaviour is obtained by a combination of local and global approaches. We obtain a twofold representation: a behaviour space and a situation space. Scenarios allow combining the two representations in a unified picture. The system thus aims to helping new drivers to contextualise their learned procedures into effective practices to avoid or correct risky behaviours by a self-training approach. This paper presents the methodology, data collection method and the implementation for the crossing of a T-crossroad.