Augmented reality heads-up display (AR-HUD) is becoming increasingly popular as a way to keep drivers focusing on roads. By overlaying visuals on the windshield, AR-HUDs improve the drivers' view of the environment outside the car, creating a stronger sense of awareness of the surroundings. However, whether AR-HUD and to what extent different AR-HUD layouts could improve drivers' driving performance are still questionable. Unfortunately, AR-HUD is still at a research stage, not yet fully commercialized. Hence, there are few actual products in the market available for testing. For this reason, this study developed a virtual reality driving simulator to tested drivers' driving performance environment under three scenarios: without AR-HUD, dispersed layout (AR-HUD1), and dense layout (AR-HUD2). Twelve subjects were invited to join the experiment. Their driving performance was measured in various aspects. This study showed that AR-HUD with interfaces that conform to human-computer interaction principles and visual design rules could improve cognitive resource allocation and promote driving safety. Conversely, a poor designed AR-HUD could negatively impact driving safety.

This paper focuses on the establishment of an ARHUD assisted driving test system based on a VR platform, which has the advantages of high security and immersion, repeatable experiments and the ability to perform eye-movement analysis. This paper first defines and designs the vehicle driving safety icons based on human-computer interaction principles and engineering psychology, supplemented by PS to define and design the AR-HUD interface while combining mental load and other factors, then uses 3Dsmax software to build the 3D model material required for driving, then builds the driving environment and designs various driving emergencies in Unity based on featured technologies such as multi-channel rendering and global illumination, and then combines HTC VIVE Pro eye display. Thirty drivers were then tested on a distraction task. Analysis of the subjects’ eye-movement data revealed that the AR-HUD system improved the cognitive efficiency of the drivers compared to the traditional driving method while allocating cognitive resources to the central driving area, speed module, navigation information, and hazard warnings in a balanced manner, thus improving the ability to react to unexpected driving events.