Abstract :
Urban areas expose drivers to a lot of challenges (e.g., many kinds of distracting stimuli and road users) and critical situations, increasing the probability of driver errors. Therefore the accident rate in urban areas is quite high and driver assistance is needed. As part of the research project UR:BAN, this study investigates the effectiveness of four different types of driver warnings (+ control group), presented in a head-up display (HUD), to support drivers in a very critical urban scenario. The aim of these driver warnings was to alert drivers to a forthcoming pedestrian crossing the ego vehicle´s road, eliciting a fast and strong response from the drivers like an emergency brake. The driver warnings varied in their strategy (attention vs. reaction oriented) and specificity (generic vs. specific). In a driving simulation the drivers´ gaze and driving behavior were analyzed. A total of sixty drivers were tested in a between-subjects design (27 female, 33 male, M = 23.7 years, SD = 3.7 years). In general, all driver warnings affected the drivers´ performance positively. Even though the number of collisions was not reduced, drivers showed a faster and stronger brake reaction when being warned, which nevertheless reduced the collision severity. While all drivers gazed at the safety-critical object, only about half of the drivers showed a HUD glance. When the drivers gazed at the HUD, the positive effect of the driver warnings on the brake reaction time was reduced. Thus, driver warnings which do not have to be fixated in order to be perceived are an interesting perspective for future research. Furthermore, the warning types differed little from each other in their effects on the drivers´ performance. However, one warning, which was somewhat more unusual and difficult to understand, was least effective. This shows that a bad design of a warning can negate the potential benefits. Overall, the examined driver warnings can be recommended for very critical situations as they improved the drivers´ performance and did not distract them. However, not all driver warnings are equally well suited.
Keywords :
"Vehicles","Brakes","Visualization","Accidents","Urban areas","Roads","Analytical models"