The map is based on GPS telemetry data collected from ground-vehicle fleets deployed across Israel - thousands of vehicles. Raw data is processed on Tracer servers and is not transferred to any third party. We publish detected interference only - no source attribution, no identification of perpetrators, and no information that enables individual identification.
All data is aggregated by H3 cell at resolution 6 (approximately 36 km² per cell). No precise locations, license plates, equipment identifiers, or vehicle counts are published. A minimum group-size threshold (k-anonymity) is enforced before any area is shown - areas that do not meet the threshold are not displayed. A publication delay of a few minutes is applied to prevent real-time movement inference.
A minimum k-anonymity threshold (minimum group size before showing an area) is enforced before every publication.
Each cell receives an independent severity level for jamming and spoofing:
The levels do not represent exact published numerical thresholds - they are based on a qualitative stage within the internal scoring model.
GPS Jamming causes signal loss - the receiver cannot compute a position. GPS Spoofing causes the receiver to compute an incorrect position without knowing it. The map detects both types of interference, but does not attribute them to a source, device, or intent. The map shows only what was detected on the ground, not what caused it.
The filter excludes tunnels, underground parking, dense urban canyons, and parked vehicles. Correlation across multiple vehicles reporting the same anomaly is required before a cell is shown as affected. Single, random anomalies are filtered out.
Publication is based on three principles: (1) full cell-level aggregation - no individual vehicle can be identified; (2) minimum k-anonymity threshold before publication; (3) a delay of a few minutes between the event and publication. The combination ensures that even a correlation attack on external sources cannot expose a single vehicle from the published data.
After a service restart, there is a short window (a few minutes) during which data is accumulating and not yet sufficient for publication. During this window, the map may show all-green or reflect an older state until enough data accumulates. This is documented in full disclosure.
All of Israel is painted with hexes, and green is the default. But green is not a guarantee: it reflects 'we didn't detect interference here,' and how much you can trust it depends on how many vehicles passed through that cell. Coverage follows the roads where vehicles travel, so the picture is sharp in busy areas and fuzzier in remote, low-traffic ones such as distant nature reserves. We do not cover every kilometer of Israel at every moment.
For additional context and external corroboration, the following sources are available - they are based on aviation data (ADS-B) and therefore present a different angle (high altitude above ground):