X-Road: Virtual Reality Glasses for Orientation and Mobility Training of People with Visual Impairments
Lauren Thevin, Carine Briant, Anke M. Brock · 2020 · ACM Transactions on Accessible Computing · doi:10.1145/3377879
Summary
This paper presents X-Road, an affordable and accessible virtual reality system designed for Orientation and Mobility (O&M) training of students with visual impairments. O&M classes teach essential navigation skills like crossing streets, but real-world training poses risks from traffic, weather, and unpredictable conditions. X-Road addresses these challenges by creating a safe, controllable virtual environment where students can practice road-crossing skills through trial and error—including virtually "getting hit" by cars to understand consequences without physical danger. The system uses mainstream hardware—a smartphone (OnePlus 5), audio headset, and 3D-printed holder—making it far more affordable than dedicated VR headsets like HTC Vive. Built with Unity and Google's ARCore framework, X-Road allows users to physically walk in real space at a 1:1 scale, with movement tracked via the smartphone camera rather than external sensors. This enables use in large spaces like school hallways or outdoor courtyards without special installation. The system provides multimodal feedback: 3D spatialized audio for vehicle sounds (direction, distance, Doppler effect) and optional visual feedback for low-vision users, while physical tactile paving can mark pedestrian crossing positions. Developed through participatory design with three O&M instructors over six months, the team iterated through multiple prototypes—from HTC Vive (too expensive and complex) to Google Cardboard (inadequate movement tracking) to the final ARCore solution. Two virtual environments were created: VR#1 (a street crossing with randomized traffic) and VR#2 (a traffic light crossing), each representing typical scenarios used in O&M instruction.
Key findings
A user study with 13 students with visual impairments (7 low-vision, 6 blind, ages 12-20) and 3 O&M instructors demonstrated X-Road's effectiveness. Students achieved 82-100% success rates across seven O&M tasks including identifying whether a location is a street or crossroad, positioning themselves parallel to traffic, and initiating safe crossings. Critically, the errors students made in VR mirrored errors commonly seen in real O&M training—such as crossing at the wrong moment at traffic lights—validating the simulation's authenticity. Students rated the system as easy to use (6.15/7) and comfortable (5.08/7), while instructors rated it as usable (92%), autonomously manipulable (100%), and enjoyed by students (96%). The ability to experience consequences—like virtual collisions—was highlighted as a unique advantage, helping students understand crossing timing in ways impossible to safely demonstrate in reality. One surprising finding was that low-fidelity audio (basic Unity 3D sound) proved sufficient for spatial learning; high realism was less important than spatialization accuracy. Students with low vision used the visual display as a magnifying glass for distant elements, while blind students relied entirely on audio cues. Nine students who tested both X-Road and traditional VR goggles preferred X-Road (5 of 9) for O&M training, appreciating the ability to physically move.
Relevance
X-Road demonstrates that accessible VR training systems can be built with mainstream smartphone hardware, dramatically reducing cost barriers compared to specialized equipment. For O&M instructors and schools serving students with visual impairments, this offers a practical tool to supplement real-world training—allowing students to practice dangerous scenarios safely, train regardless of weather, and repeat exercises without traveling to different locations. The design recommendations are directly applicable to VR accessibility: basic audio features suffice for spatial learning; users should move physically rather than using teleportation; visual feedback should be toggleable for different vision levels; and systems should be "captivating but not too immersive" to maintain instructor communication. However, the study notes that skill transfer to real-world crossings remains unvalidated—an important gap for future research. The participatory design process with O&M professionals offers a model for developing accessible training technologies that meet actual pedagogical needs rather than assumed requirements.
Tags: virtual reality · orientation and mobility · visual impairment · spatial cognition · accessible VR · participatory design · O&M training · 3D audio