Citizen Sensing for Collaborative Construction of Accessibility Maps
Kelly Shigeno, Sergio Borger, Diego Gallo, Ricardo Herrmann, Mateus Molinaro, Carlos Cardonha, Fernando Koch, Priscilla Avegliano · 2013 · Proceedings of the 10th International Cross-Disciplinary Conference on Web Accessibility (W4A) · doi:10.1145/2461121.2461153
Summary
This demo paper presents IBM Sidewalks, a mobile application developed by IBM Research Brazil that enables citizens to collaboratively report physical accessibility issues in urban environments through crowdsourcing. The application is part of IBM's broader Citizen Sensing platform, adapted specifically for the accessibility context. The system has three components: an iOS mobile app for data collection, a data server with a REST API for persistence, and a geo-statistics dashboard for city administrators. Citizens use the mobile app to report sidewalk conditions by selecting from predefined attributes — issue type (step, pothole, or other), condition severity (regular, bad, or terrible), and size (small, medium, or large) — along with a geolocated photograph. Optional text or voice comments can provide additional detail. The interface was deliberately designed for minimal interaction time to maximise the likelihood of citizen participation. Reports are automatically geolocated via the device's GPS. The authors note that existing street-level imagery services like Google Maps cannot reliably detect sidewalk accessibility issues because parked cars frequently block the camera's view in high-traffic pedestrian areas, making crowdsourced human reporting essential.
Key findings
The dashboard component provides city administrators with a real-time, full-screen overhead map displaying all reported accessibility issues with colour-coded and shape-coded icons indicating issue types and locations. The interface updates automatically as new reports arrive and includes tables and charts showing aggregated statistics for the currently visible map region. This enables visual analytics to identify problematic areas and trends, allowing city management to prioritise infrastructure repairs where they matter most to people with disabilities. The system was designed to function as a peripheral display in "war room" settings, requiring minimal active interaction. The authors propose that with sufficient citizen engagement, the collected data could support the construction of comprehensive accessibility maps that augment existing mapping services with an accessibility information layer. Future work directions include incorporating mathematical and statistical analysis tools for prediction and recommendation, and developing intelligent context-aware citizen sensor agents.
Relevance
This work represents an early example of applying crowdsourcing and mobile technology to physical accessibility mapping — a concept that has since been adopted by platforms like Wheelmap and AccessNow. For accessibility practitioners, the key insight is that physical accessibility data is fundamentally different from digital accessibility data: it requires on-the-ground human observation and cannot be fully captured through automated sensing or satellite imagery. The system demonstrates how relatively simple mobile reporting interfaces, combined with analytics dashboards, can empower both citizens and city administrators to identify and prioritise accessibility barriers in the built environment. The approach of minimising reporting friction — just a few taps and a photo — is a design lesson applicable to any crowdsourced accessibility data collection effort. While the paper describes a research prototype, the underlying model of citizen-reported accessibility data feeding into municipal decision-making tools remains highly relevant to smart city initiatives worldwide.
Tags: crowdsourcing · accessibility mapping · mobile application · physical accessibility · urban accessibility · citizen sensing · data visualization · smart cities