How RoadSense builds the public map

The app records only the signals needed to score roads

Contributors opt into passive collection while driving. The app samples vehicle motion, location, speed, heading, and time so the backend can connect roughness to the right road.

RoadSense does not use the microphone or camera. It does not need a name, email address, or user account to contribute road-quality data.

1. 01
   Phone records

   Motion and GPS samples while a contributor is driving.

   ->
2. 02
   Phone filters

   Stopped, slow, or private-zone samples are dropped before upload.

   ->
3. 03
   Server matches

   Clean samples are matched to OpenStreetMap road segments.

   ->
4. 04
   Map publishes

   The public map shows road-level aggregates, not individual drives.

The phone drops data that should not be uploaded

Samples below 25 km/h, idle time, and samples inside a contributor's privacy zone are removed on-device. That filtered data is not sent to the server.

This is also why the public map can lag behind a drive. We prefer slower, cleaner aggregates over a map that publishes every possible bump immediately.

The server matches clean samples to roads

The backend snaps uploaded samples to OpenStreetMap road segments, then combines repeat passes from different drives. The public map is built from those segment aggregates.

Confidence depends on repeat coverage

A smooth-looking road with low confidence is still only an early signal. As more drives cover the same segment, the score becomes more stable.

What the map does not mean

• xNot a maintenance queue. The map does not decide repair priority.
• xNot emergency reporting. Dangerous road conditions should still go to the right public works or emergency contact.
• xNot surveillance. The public site does not publish individual driver traces.
• xNot complete coverage. Grey roads usually mean RoadSense does not have enough data yet.