Where's My Bus Stop? Supporting Independence of Blind Transit Riders with StopInfo
Megan Campbell, Cynthia Bennett, Caitlin Bonnar, Alan Borning · 2014 · Proceedings of the 16th International ACM SIGACCESS Conference on Computers & Accessibility (ASSETS) · doi:10.1145/2661334.2661378
Summary
This paper presents StopInfo, a system integrated into the OneBusAway transit app that provides detailed information about bus stops to help blind and low vision riders locate and verify stops, particularly in unfamiliar areas. Locating a bus stop is a significant challenge for blind transit users: stops vary widely in design, may lack distinctive physical landmarks like shelters or benches, and their position relative to intersections is often unclear. StopInfo combines internal data from King County Metro (Seattle's transit agency), which provided its database of 8,000 stops, with community-contributed information from riders who verify or add details while waiting at stops. The information categories were prioritised based on prior interviews with blind transit users and include: stop position relative to intersection (e.g., "near side, 290 feet from intersection"), bus sign type, sign position relative to curb, schedule holder presence, number of shelters, number of benches, and presence of trash cans — physical landmarks that help blind riders confirm they are at the correct location. StopInfo uses a voting system for verification: information is considered verified when it has at least three votes with 75% supermajority agreement. The system is built on OneBusAway, a widely-adopted open-source transit app used by over 100,000 riders weekly in the Puget Sound region, accessible through VoiceOver on iOS. The research employed value sensitive design methodology, focusing on human values including independence, safety, equity, participation, and respect.
Key findings
A five-week field study with six blind and low vision participants (ages 31-62, median 45.5; three totally blind) demonstrated that StopInfo was usable and supported independence. Participants submitted 76 web forms reporting on transit experiences, using StopInfo on 55.26% of trips with a median usage of 67.5% of trips. Critically, 38.16% of web forms indicated trips participants would not normally have attempted on public transit — and StopInfo was consulted on 89.66% of those trips. Participants were twice as likely (2.04x) to consult StopInfo on unfamiliar trips (familiarity levels 1-2) compared to very familiar trips (levels 4-5), though even familiar trips saw 80-100% usage, suggesting value for verification. The most helpful information categories (rated 5/5) were stop position relative to intersection, bus stop name, and sign position relative to curb. All participants rated independence as very important (5/5), and one participant used StopInfo after exiting the bus to help with wayfinding to a destination. In the three months of public deployment, 467 users contributed 870 information submissions covering 576 of 8,481 stops (7%), concentrated in high-traffic areas. An accuracy audit of 38 stops in three Seattle neighbourhoods found aggregate information accuracy of 97.3% for stop position, 94.7% for number of shelters, and 93.3% for number of benches, though sign type (65.8%) and schedule holder (57.9%) were less accurate due to transit agency sign replacement and ambiguous field definitions. Verified information was 100% accurate in nearly every category. Participants defined independence broadly — not just doing things alone, but "having a balance" and knowing how to use available resources effectively.
Relevance
This paper addresses a practical transportation accessibility gap that persists today: the "last 100 feet" problem of actually locating a bus stop after navigating to the general area. While turn-by-turn GPS navigation can get a blind traveller to a street intersection, finding the precise location of a stop sign or shelter requires detailed, localised information that standard mapping apps do not provide. The community-sourcing approach — where transit riders contribute stop details while waiting — is an elegant solution because it leverages natural idle time at stops and produces information from the perspective of people who actually use the stops. For accessibility practitioners and transit agencies, StopInfo demonstrates that combining institutional data (transit agency databases) with community knowledge produces more useful and accurate results than either source alone. The value sensitive design framework's emphasis on avoiding paternalism is noteworthy: the system provides information that supports riders' own decision-making rather than making decisions for them. The finding that participants attempted trips they otherwise would not have taken — and that even experienced, confident blind travellers found value in the system — suggests that information gaps, not just navigation skill, are a primary barrier to independent transit use by blind people.
Tags: public transit · blind navigation · crowdsourcing · wayfinding · mobile accessibility · independent living · value sensitive design · community-sourcing