Séminaire : Bernard THIRION

In this presentation, I will review the key concepts associated with AD, take-over and driver monitoring. I will then present different studies that I completed, which are dedicated to the improvement of attention during automated driving and driver monitoring with and without the presence of distraction. I will detail the specific methodology used in experiments in a driving simulator that relies on distraction induction, eye-tracking and physiological data (i.e., ECG and EDA). In these studies, I mainly focused on visual processing, depicting how it can be possible to build adaptive HMI relying on bottom-up and top-down processes, in order to keep drivers’ attention in the driving loop and also help them reconstruct their representations of the ongoing situation during take-over requests. Ultimately, a better representation of the situation improves proper drivers’ decision-making, such as performing an evasive maneuver rather than an emergency braking in the middle of a highway with incoming traffic. I will also discuss the use of augmented reality and how this feature can be incorporated on-board with the current technology.