How to Identify License Plate Readers by Vendor and Hardware

What an ALPR Actually Is and What It Records

The Electronic Frontier Foundation sets the scope: "ALPRs automatically capture all license plate numbers that come into view, along with the location, date, and time." Flock Safety's camera FAQ draws the distinction that matters most for identification: unlike a general security camera, "License plate recognition cameras take still photos to focus solely on a vehicle and ensure a clear view of the license plate." A camera framed tight on one lane at bumper height is behaving like an ALPR. One framed on a whole plaza is not.

The Four Form Factors

EFF notes ALPRs are "typically mounted on street poles, streetlights, highway overpasses, mobile trailers, or attached to police squad cars." Motorola's own deployment guide, as cited in the NACDL primer, lists six types: "fixed, quick-deploy, video-based, trailer, mobile dash-cam, and app-based". Four shapes cover most of what is visible from the street.

• Fixed pole. The April 2021 NYPD policy describes stationary readers as "permanently affixed to a specific location and record the license plates of all vehicles that pass within the LPR range." A February 4, 2026 revision of that policy exists; the wording here is the 2021 version.
• Streetlight-mounted. Clamped to an existing light pole and drawing power from the streetlight's own fittings. No new pole, no solar panel.
• Trailer. A towable mast on a wheeled chassis, parked for weeks at a time. Motorola covers this with its L5M and ReaperXD lines.
• Vehicle-mounted. Axon's Fleet 3 ALPR "uses a wide 60 degree field of view covering three lanes out to 50 feet away", at closing speeds up to 140 mph. Leonardo says its Plate Hunter Mobile "can be mounted on patrol cars, unmarked vehicles, or temporary enforcement setups."

Telling the Vendors Apart

Every figure below is a vendor's own published claim, not an independent measurement.

Genetec AutoVu SharpV, the worked example

The SharpV is the best-documented fixed unit in public materials. A mirrored copy of the Genetec datasheet, printed "UPDATED: 05-FEB-20" with a 2020 copyright, gives dimensions of 2.5 by 7.6 by 8.5 inches (2.7 high with the sunshield on the black version), a weight of 5.04 lb or 2.29 kg, sealing of "IEC 60529: IP66/IP67", and "Pole and Wall Mount included". The identifying line is the power supply: "PoE+ (Power-over-Ethernet) - 802.3at Type 2 (25.5 W)". A SharpV runs off an Ethernet cable, so it implies wired infrastructure nearby and rules out solar. It is dual-lens, a monochrome plate sensor beside a colour context camera, so a two-aperture face on a compact metal box aimed down a lane is a strong cue for this family.

Flock Safety Falcon

Flock's own published spec sheet, hosted on Flock's content delivery network rather than a live product page, lists the camera at 8.75 by 5 by 2.875 inches, 3 lbs, fixed with adjustable band clamps, with footage that "Uploads via integrated LTE" and therefore no data cable. It lists dual solar panels at 18 to 20 volts weighing 25.73 lbs with hardware, on a DOT breakaway pole at 12 foot installed height. A teardown published by CEHRP on April 30, 2022 found a "5MP 16mm Fixed M12 1/2 Inch IR Filter Action Camera Lens", infrared LEDs, a motion sensor, a cellular modem, a rechargeable battery, and rear connectors for the modem antenna and the solar panel. It carries its own caveat: it examined "an older version" whose parts are "no longer being sold".

Axon, and the power contrast that actually works

The most useful discriminator in the field is where a unit gets its electricity. Axon Lightpost "installs in under an hour on existing light poles" and is powered via a "NEMA 7-pin Photocell (up to 480V) via included UbiHub device", built on Axis Q1800 dual-mount hardware; the related Axon ALPR announcement is dated April 22, 2025. Lightpost taps the socket a streetlight's dusk-to-dawn photocell normally occupies, so it carries no solar panel at all, while Flock's Falcon ships with dual solar panels and Leonardo's Street Sentry runs on solar and battery. That question, own solar panel or streetlight head, splits the vendor field faster than any housing-shape comparison.

Motorola Solutions, Rekor and Leonardo

Motorola's lineup is organised by capability rather than looks: L5F is a rugged dual-sensor unit for high-speed multi-lane work, L6Q a compact rapid-deploy model with LTE, Wi-Fi and Bluetooth, L6D pairs ALPR with HD video context, and L6A covers long-range low-speed sites. Those mappings are paraphrased from vendor material, not quoted. A Miami Lakes police procurement document calls ReaperXD a "low-profile, compact, ANPR/LPR camera for use in fixed LPR applications."

Rekor's Edge Max "Operates on direct AC or solar power with continuous runtime" and is "Die-cast aluminum, IP65", rated for up to 6 lanes, 300 ft and 120 mph, against Edge Pro's 1 to 2 lanes, 75 ft and 70 mph. Because Rekor ships both AC and solar builds, a panel neither rules it in nor out. Leonardo's ELSAG fixed line splits into Plate Hunter, "mounted along key arteries and highways, on bridges, tunnels and overpasses", and Street Sentry, a "pole-mounted LPR" with its "own power source...both solar and battery".

What It Is Not

• Red-light and speed cameras. Verra Mobility's own explainer says "Radar or LIDAR is used to measure a vehicle's speed" and that "High-resolution cameras are triggered when a violation occurs." The difference is triggering: enforcement cameras fire on a measured violation, an ALPR reads every plate. The same page also undercuts the strobe shortcut, noting that "Many modern cameras use infrared or no-flash imaging, especially in residential or school zones."
• Transportation-department dome cameras. A New Jersey DOT specification dated August 2017, one state's spec rather than a national standard, calls for an "Anti-impact, vandal resistant, dual wall aluminum enclosure with a min, 0.090-inch thick, clear polycarbonate lower dome", rated "NEMA Type 4X". The giveaway is shape and motion: a rotating dome, versus a fixed-aim rectangular box locked on one lane.
• Radar traffic counters. Radar units, not cameras: no lens, no infrared illuminator, and aimed across the roadway to count vehicles rather than down a lane to read a plate.
• Weather stations. The Federal Highway Administration defines an Environmental Sensor Station as "a fixed roadway location with one or more sensors measuring atmospheric, surface (i.e., pavement and soil), and/or hydrologic (i.e., water level) conditions". Visually, a mast of assorted instruments, not a camera housing.
• Toll gantries. Overhead structures spanning the roadway with transponder antenna arrays alongside plate cameras, per the FHWA toll terminology glossary. The distinguishing feature is scale: a structure over every lane, not one pole at the kerb.

Looking Up What Is Deployed Near You

OpenStreetMap holds the crowdsourced data. The tag is surveillance:type=ALPR, defined on its wiki page (last modified April 20, 2026, "de facto" status) as "A camera designed to record license plate identifiers of passing vehicular traffic." Required keys are man_made=surveillance, surveillance=* and surveillance:zone=*. The optional keys make a record useful later: camera:mount (typically pole), direction, electricity (solar or grid), and manufacturer. The wiki notes that "Flock Safety, Axis Communications, and Motorola Solutions are common manufacturers."

DeFlock renders that data at maps.deflock.org and takes reports through either "a simple mobile interface" (its app) or "a more advanced interface for detailed reporting in your browser" (the "OSM Web Editor"). For Flock-specific spotting cues and the civic-action routes available once you know what is deployed, see our guide to finding and avoiding Flock cameras.

EFF's Atlas of Surveillance maps agency-level deployments rather than individual poles, built by "More than 1,500 students, teachers, volunteers, journalists, and other researchers". It states its own limitation: with more than 15,000 datapoints across 6,000-plus jurisdictions, "our research only reveals the tip of the iceberg". Its data was last updated on July 27, 2026.

Local law can also create a paper trail. Community Control Over Police Surveillance ordinances require public process before an agency acquires surveillance technology, and the ACLU states that "CCOPS laws have been adopted in 26 jurisdictions from coast to coast". Two have citable text: Oakland Municipal Code Chapter 9.64, with annual reporting to the Privacy Advisory Commission; and Seattle, where Ordinance 125679, "codified as Seattle Municipal Code Chapter 14.18", took effect on November 4, 2018 and requires a Surveillance Impact Report before qualifying technology is acquired or expanded.

How Long the Data Is Kept

There is no single answer, and a retention figure quoted without its jurisdiction is meaningless. The spread runs from minutes to years.

• New Hampshire, the shortest in the country. Under N.H. Rev. Stat. Ann. 261:75-b, non-hit plate reads "shall be purged from the system within 3 minutes of their capture", with exceptions for reads tied to an arrest, a citation, protective custody, or a missing or wanted person.
• Maine, 21 days. Under 29-A M.R.S.A. 2117-A, non-investigative ALPR data "may not be stored for more than 21 days."
• Arkansas, 150 days. A.C.A. 12-12-1801 to 1808 provides that captured plate data "shall not be preserved for more than one hundred fifty (150) days", as quoted in the NACDL ALPR primer.
• Massachusetts, one year by policy. In Commonwealth v. McCarthy, 484 Mass. 493, 506 (2020), the court characterised the state policy this way, as quoted in the NACDL primer: "The one-year retention period indicated in the [Executive Office of Public Safety and Security] retention policy is certainly long enough to warrant constitutional protection."
• New York City, five years as of 2021. The 2021 NYPD policy states that "Data collected through NYPD's LPRs is retained for five (5) years." A February 2026 revision exists and could not be checked here, so treat five years as the 2021 position, not a confirmed current one.
• California, no statewide ceiling. Cal. Civil Code 1798.90.51 sets no maximum, requiring instead that every operator publish a policy stating "The length of time ALPR information will be retained" and how it decides when to destroy it.
• Vendor default. Flock's published position is that it "defaults to 30-day data retention, adjustable based on local needs". A vendor default is a starting configuration, not a legal limit.

Which Cues Are Reliable and Which Mislead

Three signals hold up reasonably well: a small rectangular housing with a multi-lens array and an infrared illuminator, on a pole with no traffic-signal head; a discrete solar panel paired with a small camera box and nothing else on the pole; and, strongest of all, an OpenStreetMap or DeFlock record where a contributor has already tagged manufacturer=* after identifying the unit in person.

Four cues mislead often enough to name.

• Assuming any pole-mounted box with a lens is enforcement equipment. Radar traffic counters have no lens at all, and environmental sensor masts carry no camera housing.
• The flash heuristic. Looking for a separate strobe is undercut by Verra Mobility's own statement that many modern cameras use infrared or no-flash imaging.
• Treating a rotating dome as a plate reader. Domes are the standard form for general traffic-management cameras. ALPRs are fixed-aim, because the point is to hold one lane in frame.
• Shape-matching against a single reference photo. The CEHRP teardown notes the Flock parts it examined are "no longer being sold", and housings change between hardware generations.

Work from function to vendor: confirm the unit is a camera, then that it is fixed-aim on one lane rather than surveying an area, then look at how it is powered, and only then reach for a vendor name.