VStroll: An Audio-based Virtual Exploration to Encourage Walking among People with Vision Impairments
Gesu India, Mohit Jain, Pallav Karya, Nirmalendu Diwakar, Manohar Swaminathan · 2021 · ASSETS '21: The 23rd International ACM SIGACCESS Conference on Computers and Accessibility · doi:10.1145/3441852.3471206
Summary
People with visual impairments (PVIs) engage in significantly less physical exercise than sighted adults, with 1.5 times higher odds of obesity. Barriers include inaccessible infrastructure, parental discouragement due to safety concerns, lack of walking partners, and the monotonous nature of walking. While accessible exergames exist (VI-Tennis, Blind Hero, Eyes-Free Yoga), they require expensive hardware that limits adoption. India et al. developed VStroll, an Android smartphone app that motivates walking by enabling users to virtually explore real-world locations while physically walking safely indoors. Users select any start location via Google Maps, then walk while hearing spatial audio announcements about nearby points of interest (POIs), road names, and intersections sourced from OpenStreetMap. At each intersection, users choose their route via voice input, creating a sense of autonomous exploration. The app provides fitness tracking (steps, calories, distance) and uses binary spatial audio (left/right ear) to indicate which side of the road a POI is located. A five-day study with 16 blind participants in India (11 blind from birth, 5 who lost sight later) resulted in 253 trips totaling 50.8 hours of walking and 121.6 kilometers covered. The study was conducted during COVID-19 lockdown in January-February 2021, with participants recruited through NGOs working with PVIs across eleven Indian cities.
Key findings
VStroll successfully motivated sustained walking: participants took an average of 15.8 trips each, and six participants continued using the app for 2-8 additional days after the study ended without being asked. Key motivators included real-time fitness updates (10 participants set daily goals like "120-150 calories everyday"), discovering new POIs (which participants compared to a treasure hunt game), and the ability to virtually revisit meaningful locations — seven participants became nostalgic exploring childhood neighborhoods, universities, or places they visited before losing their sight. The spatial knowledge acquisition was striking: 12 participants serendipitously discovered unfamiliar POIs near their own homes, including one who found a tailor shop next door he had not known about in two years. Route selection at intersections was particularly valued, as it gave participants a sense of autonomy and independence they rarely experience in physical navigation where they typically depend on sighted guides. Participants described the experience as being a "virtual avatar" exploring an accessible virtual reality. The app also served as a more engaging alternative to podcasts or music during walking, with the continuous POI announcements providing interactive feedback that "nudged" users to keep going. Limitations included OpenStreetMap data quality issues (unnamed roads, sparse POIs in some Indian locations), the assumption of straight roads between intersections, and all participants being TalkBack users, making findings less generalizable to low vision users.
Relevance
VStroll demonstrates a compelling approach to accessible exercise technology that is low-cost (requiring only a smartphone), safe (users walk indoors), and intrinsically motivating through virtual exploration. For accessibility practitioners, the key insight is that combining exercise with spatial learning creates a dual benefit — participants were motivated to walk more while simultaneously gaining knowledge about their physical environment that reduced their dependence on sighted assistance. The design decisions offer practical lessons: binary spatial audio was preferred over full 3D audio due to reduced cognitive load; voice input kept the experience hands-free; and filler information (step counts, distance updates) prevented long silences that would break immersion. The finding that participants wanted to use VStroll for real-world navigation planning — not just exercise — suggests opportunities for bridging virtual and physical navigation tools. The study also highlights an underserved population: India has one-third of the world's blind population, and participants' enthusiasm for "virtual travel" to international destinations they could not physically visit points to the broader potential of accessible virtual exploration technologies.
Tags: visual impairment · blindness · physical activity · walking · spatial audio · navigation · virtual exploration · exergame · smartphone · assistive technology