Yoo, H., Tsimhoni, O., Watanabe, H., Green, P. and Shah, R.
To design automotive head up displays, the extent to which HUDs interfere with detection of elements of the road scene (other vehicles, edge markings, signs, etc.) as a function of HUD location needs to be determined. In this experiment, 24 subjects sat in a driving simulator and watched a videotape of a real road. To encourage them to scan the road as they would while actively driving, subjects pressed a key whenever they detected predetermined events in the road scene (passing cars, tail lights of a leading vehicle, road signs on the right edge of the road). In addition, subjects pressed a key when an amber-colored warning triangle appeared on the HUD at one of 15 locations.
Response times to HUD warnings, and to a lesser extent, detection probability, varied with warning location. The mean response times ranged from 842 ms to 1390 ms. The fastest response time was obtained at 5 degrees to the right of the center. The detection probability for most locations (12 of 15) was 0.97 or higher. The HUD task did not significantly interfere with road event detection. Response times to car signals increased from 1175 ms to 1260 ms when the HUD task was added, and the detection probability did not change when the HUD task was added. The corner locations were perceived by many subjects to be worse than central locations, and they disliked the center location and the lower center. If one location were to be chosen, it would be located 5 degrees right or left of the center.
|Graphical Abstract (.pdf)|||||UMTRI-99-9 Full Report (.pdf)|
Close This Window