How Flock Safety's License Plate Cameras Work
Drive through a growing number of American subdivisions, small towns, and city streets and you may notice a slim, matte-gray box mounted on a pole, quietly blinking a small LED ring. It carries no obvious brand name, needs no wall outlet, and looks more like a birdhouse than a surveillance device. It is a Flock Safety camera, and over the past several years these unassuming units have quietly become one of the most widely deployed pieces of civic surveillance technology in the United States.
What Flock Safety actually builds
Flock Safety, founded in 2017 and headquartered in Atlanta, Georgia, produces a family of solar-powered, cellular-connected cameras designed specifically for outdoor property crime and vehicle investigations. Its flagship product is the Falcon, a license plate reader (LPR) camera that captures images of every vehicle that passes it — day or night, at speeds up to roughly 100 miles per hour — and extracts the plate number, state, vehicle make, color, and body type.
Unlike older license plate readers used by toll agencies and some police departments, which often require a vehicle, a trained operator, and hardwired power, a Falcon can be bolted to a utility pole by a homeowner association, strapped to a streetlight by a small-town police department, or clustered at the entrances of a subdivision. The unit is entirely self-contained: a solar panel charges an internal battery, and a built-in LTE modem pushes data to Flock's cloud without any local network configuration. Once installed, it just runs.
The company also makes the Sparrow, a more general property crime camera, and the Raven, a device aimed at gunfire detection and broader audio monitoring. But the Falcon is the product most people encounter, and the one driving Flock's rapid expansion into thousands of municipalities.
How the technology works
A Falcon camera uses a high-resolution sensor paired with an infrared illuminator so it can read reflective plates in total darkness. As a vehicle enters the camera's field of view, the system triggers a burst of captures. Onboard or cloud-based computer vision then performs two tasks: optical character recognition to read the alphanumeric string on the plate, and vehicle attribute classification to guess the make, model, and color.
Each detection is packaged into a record that includes the plate number, state of issue if identifiable, vehicle attributes, a timestamp, GPS coordinates of the camera, and a cropped photo of the vehicle. That record is uploaded to Flock's cloud platform, where law enforcement customers can search by plate, partial plate, vehicle color, or time window across every camera they have access to — and, increasingly, across cameras owned by neighboring agencies that have opted into data sharing.
Flock retains plate data by default for 30 days, though customers can configure shorter retention windows. The company says it does not sell data to third parties and that access is restricted to verified law enforcement and authorized private users. Searches are logged, and the platform offers an audit trail so administrators can see who queried what and when.
Why police and HOAs adopted it so fast
The appeal is straightforward. Property crime — stolen cars, burglaries, package thefts, catalytic converter thefts — is notoriously hard to investigate after the fact, and most local departments lack the staff to canvass for witnesses or chase down every lead. A network of plate readers effectively turns every covered roadway into a continuous, searchable log of vehicle movement. Detectives can take a plate from a burglary report and, within seconds, see every camera that vehicle passed in the preceding days.
For homeowner associations and private communities, the pitch is even simpler: a few thousand dollars a year buys a virtual gate. Flock's pricing is subscription-based, bundling hardware, cellular service, cloud storage, and software into a single per-camera annual fee. That removes the capital expenditure and IT burden that historically kept small agencies out of the LPR market.
Flock also lowered the friction of inter-agency collaboration. A feature called Flock Total lets participating departments search across shared cameras, which means a suspect driving through three neighboring towns can be tracked across jurisdictional lines without formal data-sharing agreements.
The privacy debate
That same frictionless sharing is exactly what worries civil liberties groups. The American Civil Liberties Union and the Electronic Frontier Foundation have argued that networks of always-on plate readers effectively create a mass location history for every driver, regardless of whether they are suspected of any crime. A vehicle's plate, tied to a registered owner, is enough to reconstruct commutes, medical visits, political meetings, and religious attendance with disturbing precision — especially when data from dozens of agencies is pooled.
Critics point out that 30-day retention, while shorter than some legacy systems, is still long enough to build a detailed movement profile, and that opt-in sharing arrangements expand the dragnet far beyond the community that originally bought the camera. Some cities, including San Francisco and Oakland, have passed ordinances sharply restricting LPR use or requiring explicit city council approval before cameras are deployed.
Flock and its law enforcement customers counter that the technology is narrowly investigative, that plates are read in public view where there is no reasonable expectation of privacy, and that the retention window and audit logs provide meaningful guardrails. They point to solved homicides, recovered stolen vehicles, and successful Amber Alerts as evidence the systems work as intended.
Where it goes from here
Flock is now valued in the billions and reportedly has tens of thousands of cameras deployed across all 50 states. The company is expanding into audio detection, drones, and broader evidence management software, positioning itself as a full operating system for small and mid-sized police departments. That growth is forcing a broader conversation that the United States has mostly avoided: what rules should govern a private company quietly building one of the largest vehicle-movement databases in the country, and who gets to decide when the cameras watch, what they remember, and who can look.
For now, most of those decisions are being made one homeowners association and one town council at a time — which is precisely how Flock designed its go-to-market. The technology will keep spreading. The debate over its limits is only beginning.