“Commute Booster” leaves problem of limited function of navigation systems underground

The new navigation app “Commute Booster” from researchers at New York University (https://www.nyu.edu) is designed to make it easier for blind and visually impaired people to find their way around subway stations. Orientation aids used today accompany people at most as far as the entrance to the stations. In the maze of corridors and platforms, however, they are no longer a help.
Closing an important gap

“Subway stations can be treacherous for people who can’t rely on vision,” says research director John-Ross Rizzo. The app’s developers include consultants from the New York Metropolitan Transit Authority (https://new.mta.info), which operates the Metro system. “Most GPS-enabled navigation apps address only the ‘first’ and ‘last’ miles, and therefore do not meet the needs of commuters who are blind or visually impaired. Commute Booster is designed to fill that gap,” Rizzo said.

Subway wayfinding signs are usually graphic or text-based. The visually impaired can see them poorly or not at all, limiting their ability to navigate unfamiliar environments independently. The Booster uses a smartphone’s camera to capture the signposts, interpret them and guide the user to their destination, such as the exit or the platform on which their train is arriving.

Success rate of 97 percent

The researchers tested the app at three New York City subway stations – Jay Street-Metrotech, Dekalb Avenue and Canal Street. In 97 percent of the cases, it guided users safely to their destination. The Commute Booster system is based on two components. The first, “General Transit Feed Specification” (GTFS), is a standardized way for public transit agencies to share their data with developers. The second is optical character recognition (OCR). The GTFS dataset contains descriptions for locations and routes within each subway station.

Commute Booster’s algorithm uses this information to create a complete list of wayfinding signs in subway stations that users will encounter during their intended trip. The OCR function reads aloud the text on the signposts presented to users in their immediate vicinity so that they can follow them. The booster omits the signposts that are irrelevant to the user’s destination.