Last reviewed: September 29, 2026
The name makes the technology sound simple.
Camera sees license plate.
Computer reads license plate.
Done.
That may describe the starting point. It does not describe the platform these products have become.
Start with the ordinary license plate reader.
A modern ALPR can capture the plate plus other vehicle information. Flock publicly describes its vehicle-focused products as using visible characteristics such as make, model or body type, color and distinguishing vehicle features.
That means an investigation does not always have to begin with a complete plate number.
It can begin with a vehicle description and narrow from there.
Flock Safety: License Plate Readers
The vehicle itself becomes searchable.
Flock has publicly described searches involving visible characteristics such as vehicle type, color, accessories, stickers, trailers, toolboxes and visible damage.
That changes the mental model from:
What does the plate say?
into:
What vehicle looks like the one we are trying to find?
A missing plate does not necessarily end the search.
A witness may remember a dark SUV with rear-end damage but never see the plate.
Modern vehicle-search systems can still use those visible attributes as investigative clues.
That may be useful in an investigation.
It also demonstrates why “plate reader” no longer describes the whole capability.
Natural-language search changes the workflow again.
Flock FreeForm is marketed as allowing authorized users to search enabled vehicle and video evidence using ordinary written descriptions.
Flock’s own examples include descriptions such as a dark SUV with rear bumper damage and, on enabled video systems, a person described by visible clothing or accessories.
That means a user may be able to describe what they are looking for rather than build a rigid database query manually.
Can it search for people?
Here the product distinction matters.
Standard Flock LPR cameras are vehicle-focused. Flock says people-search features do not run on those plate-reader cameras.
Separate compatible video systems can support people searches using visible, non-biometric characteristics such as clothing or accessories.
Flock says those searches are not facial recognition.
So this statement would be wrong:
Every Flock plate camera performs facial recognition.
This statement is more accurate:
Flock’s broader platform includes AI-assisted people searches in enabled video systems using visible, non-biometric characteristics.
Privacy does not begin and end with facial recognition.
Different cameras can become one search experience.
Modern platforms increasingly bring vehicle observations, video evidence and shared sources into common investigative workflows.
The relevant unit can become not one camera, but the collection of evidence sources a user is authorized to search.
Then add sharing.
One municipality may own a relatively small number of cameras.
That does not necessarily describe the information available to its investigators.
Flock supports cross-agency sharing when participating agencies enable it.
Sharing is configurable. It should not be described as automatic in every deployment.
That is why local research needs to ask both:
How many cameras does the municipality operate?
and:
What outside information can its users actually search, and who can search its information?
Flock Safety: Law Enforcement Access
Pattern tools go beyond an individual sighting.
Flock’s Enhanced LPR materials have marketed capabilities including tools such as:
- Convoy Search;
- Multi-Geo Search;
- Visual Search;
- plate-swap detection;
- hotspot analysis;
- and other tools intended to identify patterns across vehicle observations.
This does not prove a particular Metro Detroit agency has purchased or enabled those features.
It shows where the technology is going.
“Which vehicles travel together?” is different from “What plate was that?”
Consider the progression:
- Read this plate.
- Find this vehicle.
- Find vehicles matching this description.
- Where else has this vehicle appeared?
- Which vehicles appear around the same places?
- Which vehicles appear to travel together?
That is no longer a story about reading letters and numbers from a plate.
The camera is the beginning of the system.
The physical roadside device gathers observations.
Software determines how those observations can be remembered, searched, shared, compared and analyzed.
A useful way to understand the progression is:
See it → Describe it → Store it → Search it → Connect it → Compare it → Find patterns → Build a picture
That last step brings us back to the mosaic problem.
One observation can be boring.
A gray sedan on Telegraph Road at 4:37 PM may reveal almost nothing.
A series of searchable observations around home, work, medical appointments, friends, meetings and recurring routes can reveal much more.
The collection tells the story.
Alerts make surveillance prospective as well as historical.
These systems are not only used to look backward.
ALPR platforms can also generate alerts when a vehicle meets configured criteria or appears on an authorized hot list.
That creates at least two different capabilities:
Historical search: Where has this vehicle been seen?
Future alerting: Tell me when this vehicle or qualifying match appears.
Both matter.
Audit logs are safeguards, not erasers.
Flock emphasizes individual accounts, search reasons, audit logs, permission controls and agency-controlled sharing.
Those safeguards are relevant.
They answer questions such as:
Can use of the database be reviewed?
They do not answer:
Should all these records have been collected in the first place?
Software can change after the camera is installed.
Some newer capabilities are software additions rather than entirely new roadside hardware.
That means a public debate held when cameras were first installed may not describe everything the same infrastructure can support years later.
A community should ask:
- What products and subscriptions are currently active?
- What capabilities can be enabled later?
- What outside networks can be searched?
- What integrations exist?
- Who decides when the software changes?
Capability is not proof of local use.
This is one of our most important evidence rules.
Vendor capability is not local configuration.
Local configuration is not written policy.
Written policy is not actual use.
Finding a Flock product page does not prove Redford uses every feature described on it.
The local claim should match the local evidence.
That evidence rule does not make technical capability unimportant. Installed hardware and platform access can support later software, integration, subscription or permission changes, and those changes may not be obvious from the street—or publicly visible at all without entitlement records, configuration-change logs and audits.
Ask about both: what is verified now, and what could be enabled later without a new public vote.
Other vendors matter too.
Flock is a major focus because of its networked platform and rapid adoption.
Metro Detroit has also used other LPR technologies and vendors, including Motorola/Vigilant and Genetec.
Leonardo/ELSAG markets SignalTrace, formerly called EOC Plus, as a separate signal-intelligence layer. The company says strategically placed sensors can detect publicly broadcast Wi-Fi, Bluetooth and supported RFID signals, group devices that repeatedly travel together and correlate those observations with LPR data when both systems observe the same place and time.
Leonardo says SignalTrace does not decrypt messages or read device contents. That does not eliminate the privacy issue: time, location, a recurring device group and an associated plate can still become a searchable record of movement and association.
We have not verified a SignalTrace deployment in Metro Detroit. Detroit records show Selex ES—Leonardo’s legal subsidiary—submitted a losing bid for the city’s freeway-ALPR project, but that is not evidence that Detroit purchased or deployed SignalTrace.
Read our sourced SignalTrace investigation and Metro Detroit status →
Do not assume every product works exactly like Flock.
Research the actual system.
Our campaign is about the surveillance practice, not one corporate logo.
The simple version
If someone says:
It just reads license plates.
we do not need to answer with an exaggerated claim.
A better response is:
The plate is only one searchable characteristic in modern vehicle-surveillance systems. Flock itself advertises vehicle-description searches, natural-language searches, cross-jurisdictional sharing and newer tools for finding patterns between vehicles. Separate video products can also search people by visible characteristics. The local question is which capabilities our own department actually has.
That is accurate.
And it opens the right conversation.
The public record now includes concrete examples involving people-search demonstrations, sensitive private-camera feeds, exposed Condor video and protest-related queries. Those incidents do not prove every feature is used locally, but they help test broad assurances against documented events.
Compare the claims with the documented public record →
Can Condor automatically follow and zoom toward a person?
Yes. Flock’s Guardian Mode announcement says its Condor video cameras can automatically detect and track people or vehicles, follow their movement and zoom in for identifying details. That is different from facial recognition: the system can keep a person in view and produce a more detailed image without necessarily comparing the face to a biometric database.
Reporting published December 22, 2025 documented exposed Condor interfaces where cameras appeared to zoom toward or follow people. The exposed footage included children on a playground and adults on paths and in parking areas. It also demonstrated how high-resolution imagery can be combined with information from outside the camera system to identify somebody.
That incident does not prove every Condor camera is exposed or every local agency enables Guardian Mode. It does show why “no facial recognition” is not a complete answer to questions about people tracking, zoom, placement, retention and access.
Read the full sourced Condor investigation →
The platform also extends beyond plate readers.
Flock sells acoustic detection and live-video products that can connect with its vehicle-observation system. Those tools raise separate questions about audio analysis, event clips, people detection, pan/tilt/zoom video and integrated real-time monitoring.