Method
How MeshSonar turns pings into a coverage map, and the limits of what it shows. The same method applies in every region.
What is measured
A ping is a short message sent from a phone-paired radio at a known place and time. A ping is heard when at least one repeater echoes it back to the radio. Only repeaters that answered directly (one hop) count. The app also records the signal-to-noise ratio of each echo. A ping nobody answered is kept: it is how a dead zone shows.
Squares
Public maps do not show pings. They show squares a few hundred metres across, each with how many pings were sent in it, how many were heard, the best and median signal and how many different repeaters answered. The hub’s totals count a square once if it has at least one public ping, however many regions or drivers measured it.
What it cannot say
- A square with no pings is unmeasured, not dead. Coverage depends on the phone’s antenna and radio, so one drive is one sample.
- GPS error and the radio’s listening window mean a ping can land in a neighbouring square.
- Counts are not weighted by time. A square measured once last year and one measured daily look the same except for their count and last-heard date.
Privacy rules applied before anything is public
- The start and end of every drive are trimmed (500 m unless the driver chooses another distance).
- Pings inside a driver’s privacy zones are removed.
- Drive times are rounded down to the hour; no raw tracks are published.
- Drives under review, hidden by their owner or from a banned driver are left out.
- A callsign is shown only for drivers who opted in to the leaderboard and are not anonymous.
Open data and code
The upload format and the public API are documented in the docs, and the app and this site are open source.