How Sentinel Collects Data
Transparency is core to what we do. If we’re mapping surveillance, we should be transparent about how we map it. Here’s exactly where Sentinel’s camera data comes from.
Source 1: OpenStreetMap (global)
OpenStreetMap contributors worldwide tag surveillance cameras using the man_made=surveillance tag. This is our largest international source, providing camera locations across 183+ countries. OSM data includes camera type, operator, and sometimes the surveillance zone.
Confidence: Medium to high. OSM contributors physically observe cameras, but data can be outdated (cameras removed or relocated).
Source 2: EFF Atlas of Surveillance (US)
The Electronic Frontier Foundation and University of Nevada, Reno catalog surveillance technology use by US law enforcement. This provides unique data on facial recognition, drones, body cameras, cell-site simulators, and ShotSpotter that no other source covers.
Confidence: High. EFF verifies through public records, news reports, and direct research.
Source 3: Government traffic camera APIs (US, Canada, UK)
State DOTs, provincial transportation agencies, and national highway authorities publish live camera feeds with GPS coordinates. We ingest from:
- Caltrans (California, 12 districts)
- 511 networks (SF Bay Area, Virginia, New York, etc.)
- DriveBC (British Columbia)
- Transport for London
- Australian state road authorities
Confidence: Very high. Government-published, regularly updated.
Source 4: Flock Safety deployments (US)
Flock Safety’s ALPR cameras are deployed in thousands of US communities. Deployment data comes from public records, community announcements, and police department disclosures.
Confidence: Medium. Coverage is partial since not all deployments are publicly documented.
Source 5: European speed camera databases
European countries publish speed camera locations for navigation systems. We ingest from national road safety agencies and standardized European databases.
Confidence: High. Government-published for the purpose of driver awareness.
Source 6: International open data portals
City and national governments worldwide publish camera location data through open data portals. Paris, Seoul, Singapore, and others provide official camera registries.
Confidence: High where available, but coverage varies widely.
Source 7: Public records and FOIA
Freedom of Information Act requests and equivalent mechanisms in other countries reveal camera deployments that aren’t published in open data.
Confidence: High. Government-verified through legal disclosure requirements.
Source 8: Research publications
Academic research on surveillance often includes deployment data. Studies from universities and research organizations provide verified camera counts and locations.
Confidence: Medium to high depending on methodology.
How we handle confidence
Every camera in Sentinel has a confidence rating:
- High: Government-published data or multiple corroborating sources
- Medium: Single reliable source, not independently verified
- Low: Inferred or estimated location (rarely used)
We’d rather have a smaller, accurate database than a larger one with unreliable data.