Exploring the Attention Level of AR-HUD Interface Elements Based on Driving Scenarios

AR-HUD (Augmented Reality-Head-Up Display) is a driving assistance system that allows drivers to directly view navigation and warning information through the windscreen in an easily perceivable augmented reality format. However, in ARHUD, the frequency of AR graphics changes is far too frequent, which can easily lead to negative effects such as visual confusion. Using the Situation Awareness theory and the Task-Information theory, this research analysed drivers' cognitive characteristics and summed up their functional requirements for vehicle information display systems, focusing on three relatively complex driving scenarios. The purpose of this study was to examine the amount of attention paid by drivers to various AR element information in three different driving scenarios involving vehicles passing through traffic intersections with traffic lights (Scenario 1: pull-over parking; Scenario 2: waiting at a red light; Scenario 3: passing through a green light). We invited 54 experienced drivers to conduct the experiments. The experimental results indicated that in all three scenarios, the subjects generally had a higher demand for safety reminders. In Scenario 1 (pull-over parking), they had a higher demand for basic information such as vehicle speed information; in Scenario 2 (waiting at a red light), they had a relatively higher demand for signal light reminders in safety reminder information; in Scenario 3 (passing through a green light), the subjects had significantly higher attention to vehicle speed than other basic information, and due to the complexity of road conditions, the demand for safety reminder information reached the highest level in all three scenarios. In summary, driver's demand for AR element information varies in different scenarios, as does their subjective attention to different information categories. This paper developed a design method and process based on the driver's attention hierarchy, with significant implications for guiding AR-HUD design.