For Dustin Jones to navigate a new neighborhood in New York City, it takes some research.
Jones, a disability rights advocate, uses a wheelchair, and he’s learned not to expect that the nearest subway station is going to have a working elevator — or even have an elevator at all. He consults Google before even leaving his apartment, plotting out the route that will cause the least difficulty.
Google has slowly added accessibility information to its mapping feature, including wheelchair accessible transit route planning in 2018, which Jones said makes him feel "so much more comfortable." But the ultimate dream, he said, would be if an app could tell him about the sidewalk along his route so he could more easily find usable curb ramps or stay away from broken concrete.
"Man, I’m laughing thinking about that. That’s next level," Jones told Smart Cities Dive. "That would be awesome to have that. I mean, there’s a lot of creative people out there, so maybe it can happen."
Jones’ dreams are closer to reality, thanks to data-gathering efforts focused on city sidewalks. There have been apps and smaller projects that have mapped campuses and cities for disabled users, but a new focus on getting robust sidewalk data from sources like Google Street View are accelerating those projects and making it easier to give detailed information to the estimated 85.3 million Americans with disabilities.
One such program, Project Sidewalk, uses Google Street View and machine learning to replace map-a-thons. The University of Washington venture has online crowdworkers carry out "missions" by reviewing Google images of city sidewalks, flagging obstacles like overgrown trees or broken pavement and assessing the state of curb ramps.
Jon Froehlich, an assistant professor at the University of Washington School of Computer Science & Engineering, first had the idea nearly a decade ago as he saw how adding GPS to phones had improved driving directions. Why not, he said, bring "big data to accessibility?"
Since then, Project Sidewalk has cataloged Washington, DC, with users putting more than 205,000 labels on sidewalk images with roughly 72% success (a common problem included mixing up driveways and curb ramps), and is now working on doing the same for Seattle and Newberg, OR.
Eventually, Froehlich also hopes to use an artificial intelligence (AI) network to do image evaluation, speeding up the process and making it possible for Project Sidewalk to take advantage of rolling updates to Street View and Mapillary, keeping accessibility up to date.
That data could be used for all sorts of tools, like a super-detailed directions app that would help wheelchair users avoid uneven sidewalks. Froehlich even imagines an accessibility version of Walk Score, the website that ranks neighborhoods and buildings by how pedestrian-friendly they are.
"Imagine loading up a map of a city and seeing a heat map with an accessibility score, that includes features like topography and sidewalk condition," he said to Smart Cities Dive. "You could start to compare cities and neighborhoods by how accessible they are. People are asking for this, and we’re working on it."
New data sources
If successful, Project Sidewalk would bring a macro-level approach to what’s largely been a micro-scale and labor-intensive effort to assess accessibility conditions.
Mark Bookman, a doctoral student at University of Pennsylvania, created the Accessibility Mapping Project (AMP) in order to "illustrate physical and social barriers to access on Penn’s campus and in the surrounding community.” The project relies on volunteers to assess buildings for accessibility issues — not just physical, but also information on features like gender-neutral bathrooms, prayer rooms and lactation spaces — to an app.
Bookman says he wants data to be updated in real time, but that relies on users constantly reworking the map. Instead, Bookman hosts regular events to go back over campus, usually once a month. That, he said, also helps bring a more individual experience to the app; rather than meeting a certain "code," he said AMP can "illustrate the breadth of experiences of accessibility so that those codes and standards can be revised to accord with users' needs."
Still, there is basic information that meets the "accessibility" framework, like curb cuts and paved sidewalks, that cities should seemingly have on file. But data on curb assets may be kept by different city agencies, or sidewalks may be controlled by homeowners or businesses, making it difficult to get a full sense of the state of sidewalks.
"The activist in me wants pedestrians to be as important as cars for cities and mapping software."
Jon Froehlich
Assistant Professor, University of Washington School of Computer Science & Engineering
That’s forced lawsuits against several cities to do a better job keeping an infrastructure census. Jones, the New York advocate, was part of one such lawsuit that recently ended in a settlement, in which the city agreed to survey and repair curb cuts to ensure safety for disabled and visually impaired residents. During the years-long lawsuit, a court-ordered report found that 80% of curb cuts were not compliant with the Americans with Disabilities Act.
Jess Powers, communications director for Center for Independence of the Disabled, NY (CIDNY) said the settlement was significant. Her group has also produced maps of subway station accessibility.
Sidewalk Labs-backed data mobility company Coord has also been gathering data on curb rules and assets — like fire hydrants or bollards — and releasing them in various open API and visual formats. The intent is to help cities get a better handle on parking and curb congestion issues, but co-founder and CTO Jacob Baskin said in an interview with Smart Cities Dive that clients have also embraced the potential for the disabled community.
"There’s a big need to manage and understand accessibility, whether it’s accessible parking or availability of curb ramps," he said. The startup’s Surveyor tool allows people to upload photos of curb ramps and eventually place them on a map. "That’s helpful for route planning if you have a disability, but it’s also a big help to cities to make sure they’ve got accessibility improvements in the appropriate places."
Brute force
Still, images and digital logs can’t give cities the full scope, which is why some have resorted to old-school brute force. In 2017, Seattle had 14 interns walk 34,000 blocks of sidewalk as part of a citywide survey to assess just how much would be needed to restore them to normal repair.
Emily Burns, the city’s asset management strategic advisor, said the survey turned the sidewalk overhaul from "a mostly reactive strategy to a proactive one." Instead of only responding to citizen complaints, the city could find the most troublesome spots and go there first.
"With a goal of improving walking conditions and living conditions for people with disabilities, this was a big bang for the buck," Burns told Smart Cities Dive. Plus, Seattle could then use that data for its own hackathon, and to build out its own accessible route planning map that shows sidewalk conditions, curb cuts and accessible transit stations.
As more tools come online, Project Sidewalk’s Froehlich said his hope is that the data doesn’t just transform life for the disabled community, but can reorient how cities think about sidewalks in relation to other city assets.
"The activist in me wants pedestrians to be as important as cars for cities and mapping software," Froehlich said. "We should have the same level of information. There’s a tremendous opportunity for routing and safety and accessibility, but getting that information is the great, grand challenge."