AI system teaches cars to report road potholes, and then forwards this information to road repair services

The gist of the news

In recent years, automakers and tech companies have been increasingly using sensors to report road problems. Waymo and Waze launched a pilot project for sharing pothole data with municipalities. In response, Samsara announced its own solution, “Ground Intelligence,” which will use data from sensors already installed in millions of trucks.

Key facts

- Data collection is implemented: Cameras and sensors mounted on more than 1 million trucks capture road surface conditions.
- AI training: Based on these images, the company trained a model capable of recognizing defects (potholes, graffiti, damaged barriers, etc.) and predicting their future deterioration.
- Scale advantage: The truck fleet far exceeds Waymo’s fleet (≈ 3,000 vehicles), so Samsara receives more frequent and repetitive data about the same road segments.
- Ground Intelligence features:
1. Map alerts for new potholes and other issues.
2. Ability to view vehicle video recordings to verify citizen reports.
3. Turning reactive work into proactive: planning repairs in advance.

Partners

Contracts have already been signed with several cities, including Chicago.

Additional Samsara projects

- Waste Intelligence – monitoring trash collection.
- Passenger flow management service – notifying drivers of “unexpected pickups” and creating a digital list of school buses.

The main idea

Samsara uses the vast network of sensors in trucks to turn ordinary road monitoring into an early warning and repair planning system. This allows municipalities to respond faster and save money on road maintenance.