Allocation of visual attention while driving with simulated augmented reality

Augmented reality (AR) technologies aim to optimize the visual attention of the driver by increasing the salience of high value elements. In such systems, ‘value’ is typically seen as linked to the general activity of driving, but not manoeuvres. However, several studies have shown that during activity, eye movements are specific to the immediate goal. In our experiment, 48 participants watched videos of automobile driving situations, during which they had to make decisions. In these videos, some situational cues were graphically highlighted. Depending on the experimental group, highlighted cues related to either the general driving task (e.g. road signs, pedestrians) or to a specific manoeuvre (e.g. look for overtaking cars before changing the lane). The results show that AR impacts the allocation of visual attention more strongly during the decision-making phase. In all AR conditions, the ability to extract information is less efficient. In particular, highlighting (by AR) general cues does not affect the detection of cues related to a manoeuvre, but it does change the allocation of visual attention: fixations are more numerous and less task-specific.