Navigating VR with Confidence: Insights from BlindWalkVR for Designing Accessible Shopping Experiences

As discussed on our earlier blogs, locomotion is one of the hardest design problems in non-visual virtual reality. Visually impaired people (VIPs) can only orient themselves through sound, haptics, and body cues, so the way they “walk” in VR must feel safe, precise, and low-effort, or the whole experience breaks down. A study titled "BlindWalkVR: Formative Insights into Blind and Visually Impaired People’s VR Locomotion using Commercially Available Approaches"[1], addressed this question. Seven adults with little or no functional vision tried four off-the-shelf solutions: two omnidirectional treadmills (Cyberith Virtualizer and Virtuix Omni), a walk-in-place setup using Vive trackers on each ankle, and a simple hand-held joystick on a Windows Mixed Reality controller. After following spatial audio beacons with each setup, participants rated safety, speed, precision, intuitiveness, and mental/physical effort on an adapted NASA-TLX scale and talked through their impressions with the researchers.

The results indicate that the joystick received the highest overall ratings among the four locomotion methods tested. Every participant gave it a perfect safety score, and it also led the pack for precision and ease of understanding. Users liked its “simplicity and practicality,” noting that finger movements demand little energy and leave the rest of the body free to explore virtual objects. The ankle-tracker method was praised as “closest to normal walking,” but the step detector was too sensitive, sometimes propelling users faster than expected. Participants suggested keeping sensors closer to the body or adding inertial units to improve accuracy.

Both treadmills felt bulky and noisy. While the Omni’s foot-mounted sensors made it “more precise” than the Virtualizer, constant sliding quickly tired users and the harnesses reduced freedom of movement. Physical workload scores for the treadmills were significantly higher than for the other two methods, and several testers doubted whether such large devices would ever be practical at home. Mental workload, by contrast, was similar for all four options, suggesting that once a control scheme is learned, cognitive effort levels out.

In summary, BlindWalkVR offers three actionable insights:

1. Prioritise perceived safety. Users feel safest when they can stay in one spot while the virtual world moves, exactly what a hand-held controller enables.

2. Match real effort to virtual reward. Treadmills that require heavy sliding deliver little added benefit over lighter inputs yet impose fatigue; designers should minimise unnecessary physical strain.

3. Provide calibratable, consistent mappings. Whether users step in place with foot-trackers or push a joystick, their actions should always trigger movement sounds in a consistent, predictable way so they can easily understand how their gestures translate into motion.

Applying the findings to VR shopping for visually impaired users

A virtual store must let shoppers browse aisles, compare products, and checkout without worrying about collisions or exhaustion. BlindWalkVR shows that a head-relative joystick is the safest, most precise, and least tiring locomotion method for blind users. Defaulting to this scheme, with optional well-tuned walk-in-place for those who prefer body-driven motion, will free designers to focus on richer audio descriptions, haptic product cues, and accessible cart interactions instead of solving basic movement. In short, making navigation effortless lays the groundwork for an inclusive VR shopping experience that truly empowers visually impaired customers.

Reference

1. Kreimeier, J., Karg, P. and Götzelmann, T., 2020, June. BlindWalkVR: formative insights into blind and visually impaired people's VR locomotion using commercially available approaches. In Proceedings of the 13th ACM International Conference on PErvasive Technologies Related to Assistive Environments (pp. 1-8).