You Can Help Map the Accessibility of the World

Josh Cohen in Next City: “…using a web app called Project Sidewalk….The app, from a team at the University of Maryland’s Human-Computer Interaction Lab, crowdsources audit data in order to map urban accessibility. After taking a brief tutorial on what to look for and a how-to, participants “walk” the D.C. streets using Google Street View. The app provides a set of tools to mark curb ramps (or a lack of them), broken sidewalks, and obstacles in the sidewalk, and rank them on a scale of 1 to 5 for level of accessibility.

Project Sidewalk’s public beta launched on August 30. As of this writing, 212 people have participated and audited 377.5 miles of sidewalk in D.C.

“We’re starting in D.C. as a launch point because we know D.C., we live here, we can do physical audits to validate the data we’re getting,” says Jon Froehlich, a University of Maryland professor who is leading the project. “But we want to expand to 10 more cities in the next year or two.”

Project Sidewalk tutorial

Project Sidewalk wants to produce a few end products with their data too. The first is an accessibility-mapping tool that offers end-to-end route directions that takes into account a person’s particular mobility challenges. Froehlich points out that barriers for someone in an electric wheelchair might be different than someone in a manual wheelchair or someone with vision impairment. The other product is an “access score” map that ranks a neighborhood’s accessibility and highlights problem areas.

Froehlich hopes departments of transportation might adopt the tool as well. “People tasked with improving infrastructure can start to use it to triage their work or verify their own data. A lot of cities don’t have money or time to go out and map the accessibility of their streets,” he says.

Crowdsourcing and using Street View to reduce the amount of labor required to conduct audits is an important first step for Project Sidewalk, but in order to expand to cities throughout the country, they need to automate the review process as much as possible. To do that, the team is experimenting with computer learning….(More)”.