The effects of viewing modality and positioning of a car head-up display (HUD) on response times

Purpose: Car HUDs which, unlike those in aircraft, do not usually overlay the outside world, have been positioned in a variety of locations. With the likely increase of HUDs in cars, it is important to establish the optimum location to place the virtual HUD image. The effects of presenting the virtual HUD image to just one eye, or both eyes, while the outside world (traffic lights) is viewed binocularly, were also assessed.\ud \ud Methods: Six subjects responded to changes in simulated traffic lights and a HUD image. Response times and the number of occasions a change was missed were recorded. The HUD was presented at a range of eccentricities of up to 10° horizontally and vertically, relative to the traffic lights. The outside world was always seen in both eyes, with the HUD image being viewed monocularly and binocularly for each of the eccentricities tested.\ud \ud Results: Response times to changes in the outside world view (traffic lights) are significantly faster (P < 0.01) than to changes in the HUD image. Response times to changes in the HUD virtual image and the amount of information missed increased with increasing eccentricity. More information was missed and response times were slower with a monocular HUD than a binocular HUD.\ud \ud Conclusion: The optimum location based on response times and information gathering was found to be approximately 5° diagonally down from the driver's line of sight. A virtual HUD image seen in both eyes is preferable. It is of concern that gross outside world changes are missed when drivers attend to the virtual HUD image and the outside world