City Implements AI Monitoring For Sidewalk Sheds
Cornell Tech researchers led by Prof. Wendy Ju trained an AI model on 29 million Nexar dashcam images to map New York City’s sidewalk sheds, confirming more than 5,000 active permitted structures and about 500 unpermitted or expired ones. The team proposes a public "Shedfolio" digital twin and continuous monitoring to prioritize inspections and accelerate takedowns. They urge the DOB to partner with researchers and computer-vision firms.
Key Points
- 1Detects scaffolds using AI on 29 million dashcam images, confirming >5,000 permitted and ~500 unpermitted.
- 2Reveals gaps in DOB data and slow complaint-driven inspections, exposing static oversight of dynamic infrastructure.
- 3Enables continuous, prioritized inspections and public 'Shedfolio' transparency to accelerate takedowns and enforcement.
Scoring Rationale
Strong Cornell Tech methodology and direct municipal applicability, limited by single-city demonstration and reliance on private dashcam data partnerships.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems