Jiayan Zhao, Muxu Wang, Rui Li
Introduction The proliferation of autonomous driving raises concerns about the psychological and behavioral adaptations of future drivers. One emerging challenge is the potential degradation of spatial knowledge due to reduced navigation and control over autonomous vehicles; however, limited research has examined this issue or potential solutions. The present study investigates how visualizing landmarks on an augmented reality head-up display (AR-HUD) influences incidental spatial knowledge acquisition during autonomous driving. Methods We developed an AR-HUD that visualizes both local (on-route) and distant (off-route) landmarks on the windshield and evaluated it using an immersive virtual reality (VR) simulation. Sixty-eight participants experienced one of four conditions in a 2 × 2 factorial design with windshield display (AR-HUD versus normal windshield) and road type (local road versus highway), while their eye movements were recorded. After the autonomous driving experience, participants completed multiple-choice questions assessing route, directional, and configurational spatial knowledge. Results Compared to a normal windshield, participants using the AR-HUD exhibited better route- and direction-related knowledge, while its benefit for configurational knowledge depended on road type. Eye-tracking data further indicated that the highway environment, characterized by fewer local and more distant landmarks, imposed greater difficulty for spatial information processing than local roads, and that the AR-HUD promoted more effective visual and cognitive engagement with task-relevant environmental features. Discussion These findings suggest that AR landmark visualization can support multiple levels of incidental spatial learning during autonomous driving and highlight the value of VR and eye tracking for evaluating future in-vehicle spatial interfaces.