Interactively Modeling and Visualizing Neighborhood Accessibility at Scale: An Initial Study of Washington DC
Anthony Li, Manaswi Saha, Anupam Gupta, Jon E. Froehlich · 2018 · Proceedings of the 20th International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS 2018) · doi:10.1145/3234695.3241000
Summary
This poster paper from the University of Maryland and University of Washington presents two prototype tools — AccessScore and AccessVisDC — for modeling and visualizing neighborhood-level physical accessibility for people with mobility impairments. The work addresses a critical gap: existing walkability indices like walkscore.com measure the proximity and density of walkable destinations (grocery stores, parks, restaurants) and have been incorporated into real estate tools, but they completely ignore accessibility-related features such as sidewalk conditions, curb ramps, and road grade — effectively excluding wheelchair users and others with mobility impairments. The prototypes leverage the Project Sidewalk API, which provides access to over 255,000 crowdsourced labels describing the accessibility and location of Washington DC sidewalks. Two qualitative studies informed the designs: one with 20 mobility impaired participants providing feedback on paper mockups (where 18 of 20 preferred the neighborhood accessibility visualization scenario), and semi-structured interviews with 12 stakeholders including government officials (4), people with mobility impairments (5), and caregivers (3). Government workers emphasized cost savings and prioritizing areas for physical examination, while mobility impaired users and caregivers focused on personal utility, enhanced independence, and travel confidence.
Key findings
AccessScore discretizes Washington DC into a grid and computes per-cell accessibility scores by finding the nearest points of interest via Google Maps Directions API, requesting pedestrian routes to each destination, and scoring those routes using Project Sidewalk data — adding weight for accessibility features (e.g., curb ramps) and subtracting weight for barriers. Scores are normalized by route length and visualized as a heatmap where darker colors indicate less accessible areas. Crucially, AccessScore allows end-users to customize parameters including their mobility level and obstacle weights, recognizing that a manual wheelchair user and an electric wheelchair user have different accessibility needs. AccessVisDC takes a complementary approach, providing a choropleth map of DC neighborhoods with drill-down functionality using semantic zoom — users start with an overview, then zoom to street-level and raw-label visualizations, with dynamic sidebar charts responding to user interactions. Both prototypes are open source and implemented with modern web technologies (deck.gl, Python, node.js, turf.js for AccessScore; mapbox-gl and d3 for AccessVisDC). The key technical insight is that despite 255,000+ data points, precomputing routes to points of interest and counting accessibility features makes the system performant and personalizable.
Relevance
This work takes the first steps toward creating an "accessibility walkscore" — a metric that could fundamentally change how people with mobility impairments evaluate neighborhoods for housing, employment, and daily life. The implications are significant: just as walkscore.com has influenced real estate markets and urban planning, an accessibility-aware equivalent could drive infrastructure investment, inform zoning decisions, and empower individuals. For accessibility practitioners, the parameterizable model is particularly important — accessibility is not binary, and different users with different mobility aids face different barriers. The Project Sidewalk crowdsourcing foundation demonstrates that large-scale accessibility data collection is feasible, though the system is currently limited to Washington DC. The multi-stakeholder design approach (government, users, caregivers) ensures the tools serve both policy and personal use cases. Long-term goals include cross-city comparisons and correlating accessibility with census, real estate, and land use data.
Tags: urban accessibility · wheelchair accessibility · walkability · data visualization · crowdsourcing · sidewalk accessibility · curb ramps · mobility impairment · geospatial