Generalization of augmented-reality navigation systems to uncontrolled outdoor environments

Determine how augmented-reality navigation systems designed for people with low vision in controlled indoor laboratory settings generalize to uncontrolled real-world outdoor navigation.

Background

Existing augmented-reality systems for people with low vision have primarily been evaluated in ideal indoor laboratory environments, typically through short sessions involving predefined navigation tasks such as following a corridor or avoiding obstacles in a preset room. Such evaluations do not establish whether the systems remain effective amid real-world variability, including changing weather, lighting, shadows, surface textures, social settings, and users’ evolving needs.

The paper identifies this generalization issue as a gap motivating the development and field deployment of NavSight. Although the diary study characterizes users’ experiences with NavSight and documents several environmental and usability challenges, it does not resolve the broader question of how augmented-reality navigation systems generalize to uncontrolled outdoor use.