Catch the blockage before the street floods.
Storm drains can become obstructed between inspections. The idea behind Storm Drain Sentinel is to bring visual evidence and local conditions into one review workflow so maintenance teams can investigate a potential blockage before heavy rain makes it more consequential.
The intended AI layer would detect persistent debris in camera footage and help prioritize preventive cleanup. The current build demonstrates the operator experience around that idea.
Connect a location to the evidence around it.
- Search for a New York City street, station, or landmark using OpenStreetMap geocoding.
- Load current weather and a three-hour precipitation forecast from Open-Meteo, alongside active National Weather Service alerts.
- Retrieve recent NYC 311 catch-basin reports and show nearby reports within roughly one mile.
- Present a simulated camera scene and route preview next to a human inspection action.
Changing the selected location refreshes weather, reports, and the Google Maps destination. The dashboard periodically refreshes its public-data feeds.
Live public data, conceptual camera analysis.
The interface uses HTML, CSS, JavaScript, and canvas animation. Rain effects follow live weather, while the camera scene, blockage annotation, event trail, and vehicle route preview are demonstrations.
No live camera or trained visual detector is connected. The displayed occlusion percentage is illustrative, not measured model confidence. Marking a drain for inspection changes the local interface; it does not dispatch a crew or contact anyone.
Measure detection and prioritization separately.
A next step would be to connect representative footage and labeled blockage events, then evaluate missed blockages, false alerts, persistence across frames, and performance under rain, shadows, and changing light. Prioritization would need separate validation against inspection outcomes and local flood conditions.
The present prototype does not establish calibrated flood probabilities or detection accuracy. Its contribution is the review workflow and integration of local public context.