How it works

1. Camera data

We take a snapshot (Geofabrik US extract, October 16, 2026) of every OpenStreetMap node tagged surveillance:type=ALPR inside the 50 states + DC (144,253 nodes; this matches Geofabrik taginfo's US count minus a few hundred in territories/offshore). Much of this was mapped by volunteers using DeFlock. We don't add, verify or physically confirm any camera.

2. Your route

Place names are geocoded with Photon (OSM data), and driving routes come from the FOSSGIS OSRM server at routing.openstreetmap.de. If routing fails, we fall back to a straight line and say so. Your query goes from your browser straight to those services. We don't proxy, log or store it.

3. The estimate

We count mapped ALPRs within about 75 m (≈250 ft) of the route line. Where a camera has a mapped facing direction, we also estimate how many point toward oncoming traffic in your direction of travel (within ±60°). The 0–100 exposure index is 100 × (1 − e^(−readers per 10 mi ÷ 2)), an illustrative number and not a measurement.

Other layers

• Public/outdoor CCTV (27,965): OSM man_made=surveillance nodes with a camera/dome/guard type (or no type), excluding surveillance=indoor and ALPRs.
• Traffic-monitoring cameras (8,625): OSM surveillance nodes tagged surveillance=traffic or surveillance:zone=traffic. Many DOT cameras are not mapped in OSM.
• Speed / red-light enforcement (1,487): OSM highway=speed_camera and enforcement=maxspeed|traffic_signals nodes (enforcement relations are not yet included).

In the exposure index, ALPRs count fully, enforcement cameras ×0.5, traffic cameras ×0.3 and CCTV ×0.2. These weights are editorial choices, not measurements.

Not included (and why)

• Gunshot detection (ShotSpotter/SoundThinking): sensor locations aren't published. Only rough city coverage areas appear in press and FOIA records, without an open license, so we don't map them.
• State DOT camera feeds (e.g. CDOT/COtrip, 511 systems): these mostly require API keys and come with terms that limit redistribution, so we haven't imported them yet. Only DOT cameras mapped in OSM are shown.
• Mobile/vehicle-mounted readers (police cars, repo/DRN-type trucks), private doorbell networks, drones and phone location data: there's no public location data.

Limitations

• Mapping is incomplete and uneven; positions may be off or outdated.
• Mobile/vehicle-mounted readers (police, repo/DRN-type) and private cameras are not included.
• A camera being nearby doesn't mean it captured you, and data sharing between agencies isn't modeled.
• This can't tell anyone whether they are tracked.

Attribution & licenses

Camera and map data © OpenStreetMap contributors, licensed under the Open Database License (ODbL) 1.0. The derived camera file at /data/alpr.json is also offered under the ODbL. Basemap by OpenFreeMap / OpenMapTiles. Thanks to the DeFlock community. Further context: EFF Atlas of Surveillance.

See a missing camera? Add it to OpenStreetMap via DeFlock.