AUSTIN, Texas — As more Central Texas communities cut ties with companies that operate automated license plate readers, new research is raising questions about where the cameras are placed, what they collect and whether they reduce crime.
Data provided by the nonprofit SciLine examined whether automated license plate readers, also known as ALPRs or LPRs, help curb crime and how the technology is being used as policing increasingly relies on camera-based surveillance.
Researchers said some communities are more likely to see license plate reader technology than others. Steven Keener, an assistant professor of criminology and director of the Center for Crime Equity and Justice at Christopher Newport University in Virginia, said ALPRs are more likely to be placed in communities of color and in areas that have historically experienced overpolicing.
Keener’s work follows a case in Virginia in which a federal judge ordered the release of all publicly funded locations of automated license plate readers.
“Your default is surveillance, like you cannot move about your daily life in a vehicle without coming across an ALPR camera. And there’s other patterns you see, for example, they largely surround public housing complexes in the region,” said Keener.”
Separate research from Texas Christian University professor Johnny Nhan, a criminology and criminal justice expert, found mixed results when looking at the relationship between solving crimes, reducing crime rates and the spread of Flock cameras, leaving it unclear whether the technology lowers crime.
He said he was involved in a study connected to Flock and was asked for advice, but said he could not draw that link because, “I don’t know how this data is collected by the different police departments. There’s no standardization in terms of how it’s collected.”
Nhan’s research also suggests the future of policing could involve fewer officers in the field.
“A lot of it is in the name of officer safety. You could see this scenario where, hey, maybe we don’t do high speed pursuits anymore,” said Nhan. “We’re gonna just do automatically drone deployments from a real-time crime center detected crime, and then you deploy the drones, and then they follow the person home, and then that’s where you deploy a SWAT team or something like that.”
SciLine, an editorially independent nonprofit based at the American Association for the Advancement of Science, connects journalists with scientific expertise and research.
During the briefing, Quinnipiac University assistant professor of criminal justice Kim Prozowski described license plate readers as cameras paired with software that photographs vehicles as they pass, allowing information about the vehicle to be quantified. She said police agencies generally use the systems in real time or near real time for alerts, and in retrospective investigations to determine where a vehicle passed a reader after an incident.
Prozowski said agencies may receive alerts tied to the National Crime Information Center, including cases involving wanted people, missing people or Amber Alerts. She also said the technology differs from older surveillance tools because of its scale and searchability, with the ability to scan hundreds or thousands of plates in real time.
Prozowski said license plate readers can be part of real-time crime centers, which she described as centralized hubs that can house multiple investigative technologies, depending on what an agency has acquired. She said those tools can include license plate readers, facial recognition software, databases and gunshot detection systems, allowing information to be gathered and sent back to officers more quickly.
Beyond capturing a license plate, Prozowski said some systems can collect what Flock calls a “vehicle fingerprint,” including the make and model, camera location, date and timestamp, vehicle color and body type. Depending on settings, she said the systems may also capture details such as damage to a bumper or a bumper sticker.
She also cautioned that license plate readers generally do not identify the driver and that a camera “hit” should not be treated as proof of criminal involvement on its own.
“I think a common misconception is that it can prove criminal involvement. And I think that’s where this conversation is really important because given the span of them and the scale of them, obviously one hit is innocuous meaning, you know, it might not be overly concerning of it when you have,” she said.”
Prozowski said an image alone should not generate probable cause unless tied to an NCIC case, and she noted that NCIC updates information that feeds into LPR systems about twice a day, requiring officers to verify whether a warrant or entry is still active.
Keener also discussed efforts to map where license plate readers are deployed and said some departments have released their locations, but “not many.” Keener said the most consistent response he has heard when asking why locations are not released is concern about vandalism.
“The response I’ve most consistently gotten is they’re afraid of vandalism,” Keener said.
Keener said that if a locality has not officially released all locations, “DeFlock is really the best tool you got,” while cautioning that accuracy can vary by place. He said in some areas people have worked to verify locations and the information is “highly accurate,” while in other places there can be missing entries.
He also warned that people sometimes upload images of other types of cameras and mistakenly identify them as license plate readers, adding that license plate readers are “fundamentally different than other surveillance tools.”
Nhan said one area where police find the technology particularly effective is locating stolen vehicles.
“One thing that is particularly pretty effective that police use it for are stolen vehicles,” he said.
He also said the technology can produce false positives, including cases where a vehicle is incorrectly listed or information comes from another state and can be dismissed by an officer. Nhan said alerts can come while an officer is driving or through an RRTCC real-time crime center that distributes the information.
Texas Gov. Greg Abbott has also expressed concerns over the systems after one officer was indicted for 100 felonies connected to abuse on the cameras.
Nhan also said there is no universal standard across police departments for tracking whether Flock cameras were used in a case, calling the process “ad hoc for now.”
He said some departments use an automatic check-off, while others note it in narrative reports, making it difficult for researchers to extract consistent data at scale. Nhan said his research has focused on what policies are in place for data collection, data retention and deletion, and he said there are “lots of questions” about those practices.
Abbott has said on the campaign trail that he views the technology as something worth stepping back from and pausing.
“This is an issue that we’re going to need to step back, put a pause on it, gather information to figure out what is the best pathway forward,” said Abbott.
The governor also added that fundamentally he does not want to see people’s Fourth Amendment rights, which establishes protections against illegal search and seizures, violated.
Source: cbsaustin.com
