Abstract

To develop automated systems that can execute acceptable and comfortable lane changes on the motorway, it is useful to understand how lane changes are performed by human drivers, what factors influence the lane changes, and how they can be described parametrically. More specifically, it is important to understand the acceptable thresholds for parameters which are associated with the manoeuvre execution from a driver’s perspective. Furthermore, the cognitive process of the intention and decision for a lane change needs to be addressed to provide a well-accepted automated lane change decision and timing. This work aims at gaining first qualitative insights into these aspects. Two small-scale on-road studies were conducted to initially collect data on the process of human lane change behaviour in different driving scenarios. Objective data has been collected by equipping a measurement vehicle with a broad sensor set. Qualitative subjective data was gathered through different survey methods. In the first study, both the driver and a co-driver were included by means of questionnaires. In the second study, the ‘thinking aloud’ method was utilised to gain a deeper insight into the driver’s cognitive process. Based on the analysis of the objective data a parametric description of the lane change profile has been established. They can be used to generate automated lane change trajectories based on parameters from real data. For the evaluation of such automated lane change profiles, the implementation of driving studies in a highdynamic driving simulator is useful and has been initially tested in an expert pilot study