Automatic license plate readers may present privacy concerns

A salesman puts a license plate on a pre-owned 2018 Chevrolet Colorado pickup truck at the Green Chevrolet dealership in East Moline, Illinois, U.S., on Monday, May 3, 2021. General Motors Co. is scheduled to release earnings figures on May 5. Photographer: Daniel Acker/Bloomberg
A salesman puts a license plate on a pre-owned 2018 Chevrolet Colorado pickup truck at the Green Chevrolet dealership in East Moline, Illinois on May 3, 2021.

Any technology that enables the gathering of consumer information can deliver opportunities for organizations—more effective marketing, better customer service and increased sales, for example. It can also present risks of invading privacy. Such is the case with automatic license plate readers.

Some states have begun using the technology or are considering it to find uninsured drivers, with the aim of keeping uninsured vehicles off the roads.

“In states with insurance requirements, vehicle registration often requires proof of insurance at the time of registration,” says Michael Richmond-Crum, director, personal lines & counsel, at the American Property Casualty Insurance Association.

“This means that each vehicle should be associated with an insurance policy within DMV databases,” Richmond-Crum says. “If an insured allows a policy to expire, many states require that insurance carriers notify the DMV of the policy lapse. If a vehicle is recorded being operated, and the related policy has lapsed, it may be being operated without proper insurance coverage.”

The primary involvement of insurers in automated licenses plate reader technology is related to the information that is being verified, Richmond-Crum says. “Automated readers can be used to cross check to see if a vehicle is insured when it is scanned,” he says. “Insurers provide verification data, often via an online portal, that a vehicle is insured as required by law in many states.”

This type of verification system is similar to any other insurance database, and insurers are involved by keeping DMV records databases up to date, Richmond-Crum says.

“This is often done by an online verification system that allows DMVs to check in close to real-time if a vehicle has coverage with an insurer,” he says.

Automated license plate reading technology is becoming more and more common by law enforcement at the state and local levels, Richmond-Crum says.

“The technology is typically used to enforce speeding laws, parking violations, and gather toll fees,” he says. “Currently there does not seem to be a rush to use the technology to verify insurance financial responsibility compliance, as only a few jurisdictions have adopted it or discussed the matter.”

Privacy

Any technology that gathers consumer information is liable to raise red flags about data privacy.

The privacy concerns related to automated license plate readers is typical of general privacy concerns related to data gathering of the public, and how it will be used,” Richmond-Crum says. “Because insurance financial responsibility information is—or should be—on file with the DMV, there is no significant additional imposition on individual privacy in insurance compliance verification using automated cameras.”

States and municipalities are leveraging information they already possess, Richmond-Crum says, provided by the vehicle owner and verified by insurers, to enforce existing requirements for operating a vehicle.

“Additionally, license plate readers can be used to identify vehicles that are known to be stolen or being sought in response to suspected criminal activity,” Richmond-Crum says. “The National Crime Information Center has a program that works to provide real-time information on vehicles to law enforcement and intelligence communities,” to identify individuals and vehicles suspected of criminal activities. This includes recovery of stolen vehicles, he says.

“The recovery of a stolen vehicle allows for the return of the vehicle to the rightful owner, reducing insurance losses for vehicle owner/policyholder as well as the insurer,” Richmond-Crum says.

Some say the use of automated plate readers is a good example of how technology can benefit consumers as well as the insurance industry.


“The insurtech community is rapidly innovating the insurance industry with new products and services that are meeting the evolving needs of American consumers,” says a spokesperson for the Insurance Information Institute, an online source for insurance information.



The institute “supports all advancements in technology that ultimately allow more people to obtain the insurance coverage they need, all while delivering enhanced protection against uninsured motorists for families and communities.”


Others are more concerned about how collected data potentially invades privacy. With automated license plate readers, “one of the chief concerns is that, depending on how they are used, they represent the constant threat of drag-net surveillance,” says Matthew Guariglia, a policy analyst with the Electronic Frontier Foundation (EFF), a nonprofit organization defending civil liberties in the digital world.

“If a license plate reader is set up at an intersection and is recording hundreds or thousands of cars pass by a day in the hopes that a car associated with an uninsured person passes by, there is the potential for every other person driving through that intersection to also have their movements recorded and tracked,” Guariglia says.

License plate readers should not record, and cities should not store, all of the massive data collected about peoples’ movements that the machines are capable of collecting, Guariglia says.

“One problem is that in most cities and towns in the U.S., residents have very little control over what technologies their city and police department buys and deploys,” Guariglia says. “At the very very least, towns and cities should pass surveillance ordinances that require cities to leave it to elected officials or voters to decide what technology the city is allowed to buy and use, how it's used, and how much transparency is required regarding government use of technology.”

In one early example of how plate-tracking technology can be used, Oklahoma in November 2018 launched the Uninsured Vehicle Enforcement Diversion Program, which uses readers from Rekor Systems.

With the UVED, “the customer determines which data is kept and for how long,” says Charlie Degliomini, executive vice president at Rekor. The system is often set to align with the data privacy policies of the state in which it is being used, he says. In Oklahoma, for instance, only the license plate information for uninsured vehicles is kept to compare against state data, he says.

“Law enforcement customers often store or share data only if it is related to an open investigation,” Degliomini says. “The system only captures images of the vehicle, not the driver. In Oklahoma’s case, the registered vehicle owner is the responsible party for insurance coverage.”

How it works

The goal of the statewide UVED program, which is operated by the Oklahoma District Attorneys Council, is to leverage license plate recognition technology to ensure that all drivers have at least the minimum required amount of liability insurance, according to the UVED site. This will ultimately lead to safer roadways, it says.

After multiple expansions, UVED now has 70 active cameras at locations around the state, says Amanda Arnall Couch, UVED program director.

The program involves installing a network of fixed and trailer-mounted optical sensors at key transit points across all jurisdictions in the state. While the technology produces hundreds of thousands of images, all data collected by UVED is retained solely by the program, and only for as long as it’s being used as evidence of a violation, UVED says. When the data is no longer needed, it is deleted or destroyed.

“Since inception, the UVED program has enrolled 62,540 citizens in diversion, meaning each of those citizens acquired valid insurance and agreed to maintain coverage for a period of two years,” Couch says. “Therefore, no criminal charges were filed.”

The program currently enrolls about 82 citizens per day, “preserving law enforcement and court resources for higher priority offenses,” Couch says. “We estimate at least 62% of enrollees remain in compliance throughout the two-year term. Those who do not are subject to additional notices and/or prosecution.”

Rekor’s Rekor One platform uses computer vision artificial intelligence (AI) and machine learning to identify vehicles by make, model, color, and identifying marks, as well as license plate information, says Degliomini.

“Once identified, this information is then compared to a database of vehicles that have active insurance that meets state-mandated minimum standards,” Degliomini says. For vehicles that do not meet the state-mandated minimum standards of insurance, Rekor conducts outreach through a customer service call center, through which citizens can speak to a customer service representative and address any questions, extenuating circumstances or discrepancies.

“When an uninsured vehicle is detected, the UVED program sends a notice to the citizen who registered the vehicle, at which point they can enter a diversion program by acquiring insurance and paying a $174 enrollment fee,” Degliomini says. “This allows the citizen to avoid the possibility of criminal charges, associated penalties, and a permanent mark on their driving record.”

Motorists can acquire insurance on their own or use an affordable option offered through the program. Multiple insurance company options are provided through the program, and other insurance providers may be added, Degliomini says.

Citizens who need help finding affordable insurance are referred to Rekor partner Freeway Insurance, Degliomini says.

The key components of the system include a Rekor One 'brain', which hosts computer vision and pattern recognition algorithms for identifying and securely storing and sharing information with authorized parties. Rekor sensors, often in the form of cameras, capture information and send it to the platform to be analyzed instantly and matched against the state’s insured vehicle database.

Sensors can be mounted on existing infrastructure including poles, bridges, or attached to trailers that are then positioned along roadways. Sensors can even be placed within vehicles, allowing them to capture images while in motion.

Rekor Systems has customers across the United States and is using the technology for multiple public safety applications. But Oklahoma is the only state currently using the technology to identify uninsured drivers, Degliomini says.

Other states are considering implementing UVED-like programs in the future, including Texas, Florida, and Missouri, according to Couch. As for plans in her state, “UVED is always looking for ways to evolve and enhance, including the addition of [more license plate readers], partnerships with surrounding states and other agencies,” she says.