Designing Future Pedestrian Navigation Interfaces for Future Augmented Reality Smartglasses

Abstract

Head-worn augmented-reality (AR) technologies overlay 3D digital content directly onto physical environment, freeing users from traditional 2D interaction paradigms and positioning AR as a promising next-generation, always-on computing platform. However, designing effective AR interfaces remains a challenging problem. AR devices must operate in a variety of lighting and background conditions, which presents challenges for hardware and graphics. They offer novel interaction affordances and must work effectively for a wide range of individuals to be true consumer-level devices. This dissertation investigates the design space of head-worn AR interfaces, focusing on the display of spatial navigational cues using empirically grounded methods in simulated mixed reality environments. It provides practical recommendations for AR navigation on smart glasses, derived from comprehensive evaluations of gaze behavior, cognitive load, and spatial learning. Furthermore, the research finds distinct engagement patterns with AR cues based on users' spatial abilities, informing personalized interface design. Notably, we demonstrate the utility of content-independent gaze metrics as indicators of spatial navigation ability and learning, suggesting their potential in adaptive extended reality applications. This research contributes to the advancement of human-centered AR navigation systems capable of adapting to individual characteristics and enhancing the process of spatial learning.