AI and volunteers map and repair U.S. sidewalks

The online platform Project Sidewalk is leveraging artificial intelligence (AI) and volunteer efforts to address the deteriorating condition of U.S. sidewalks. The platform allows users to navigate virtual streets using Google Street View imagery, cataloging damaged sidewalks, missing curb ramps, and faded crosswalks. This crowdsourced data can then feed AI tools for automated issue detection.

Founded in 2012, Project Sidewalk was created to solve a fundamental problem: many cities lack comprehensive data on the condition of their sidewalks. This deficiency poses significant risks to pedestrians, particularly those with disabilities. While AI has advanced considerably, it cannot yet independently evaluate sidewalk conditions or fully account for user experience, underscoring the continued need for human input.

The platform has proven effective as a tool for community engagement and local political action. For instance, in Oradell, New Jersey, Girl Scout troops and community members mapped nearly 36 miles of sidewalks, identifying over 11,000 potential issues. This data was presented to the city council, leading to efforts to install missing curb ramps.

While data collection is challenging, AI shows promise in specific areas like ramp detection, with one recent tool proving more accurate than human evaluators. However, AI struggles with damage caused by natural elements, such as tree roots cracking pavement.

Obstacles to sidewalk improvement often include funding shortages and fragmented responsibilities. Many cities lack updated accessibility transition plans. Without robust data, improvements are frequently driven by individual complaints. Initiatives like Project Sidewalk provide the data necessary for cities to justify funding requests and promote equitable accessibility.