Tea For Streets offers an infrastructure intelligence platform that turns a smartphone into a sensor to capture the condition of streets, blocks, and properties without additional hardware. The system scores each issue, incorporates human QA for verification, and then routes repair tasks while tracking them to completion, enabling cities to identify and fix physical defects efficiently. Their three-layer vision platform—capture, understand, act—supports end‑to‑end street maintenance workflows.
Funding
Funding not disclosed
Founders
Product
Problem
Municipalities and property managers lack a low-cost, scalable way to detect and quantify street‑level defects such as potholes, graffiti, and illegal dumping, leading to hidden maintenance expenses and delayed repairs.
Solution
Tea For Streets provides a vision platform that turns standard smartphones into field‑grade sensors using combined RGB camera and LiDAR depth data. The mobile capture layer records defects, the analytics layer classifies and scores each condition with confidence, and a dispatch layer routes repair tasks and tracks them to completion. Human quality‑assurance verifies critical detections before they are sent to crews, ensuring reliable data for decision‑making. The system delivers geolocated, millimeter‑accurate measurements for each defect, enabling block‑level reporting and city‑wide aggregation without additional hardware installations.
Target Audience
Primary customers are city public works departments, municipal infrastructure agencies, and property management firms that need systematic street‑level condition monitoring and repair coordination.
Features
- Smartphone‑based capture that fuses RGB imagery with on‑device LiDAR to produce precise depth measurements (e.g., pothole depth, graffiti area)
- Automated AI classification and confidence scoring of street conditions directly on the device
- Human QA workflow that validates detections before dispatch, reducing false positives
- Integrated dispatch module that assigns repair tasks and monitors progress to closure
- Geospatial mapping of defects at the quarter‑mile “ped‑shed” level around properties for targeted reporting
- No extra hardware required; works with existing consumer phones