加拿大 YouTuber 自制 Flock 式车牌识别摄像头反拍警察，Peel 警方上门

As he watched the U.S. debate over law enforcement’s use of Flock surveillance cameras, an idea came to Canadian YouTuber Anthony Sistilli. He would build his own license plate reader — but it would keep an eye on the police officers who use real Flock devices to track others.

The police were not amused.

Download the Straight Arrow app today to get the stories that matter free from manipulation, bias or agenda.™

Point phone camera here

After Sistilli produced two short-form videos about his project for his 140,000 subscribers, officers from the Peel Regional Police in Ontario showed up at his home on Sept. 20. Sistilli said the visit lasted roughly 30 minutes, with officers focusing on his intentions and whether he planned to build a map tracking police movements.

“Their concern was that mapping police movements could put officers at risk and help people committing crimes,” Sistilli said. “They also made the case that their own cameras and ALPR technology weren’t as concerning as I was suggesting. My impression was that they weren’t familiar with many of the controversies or examples of misuse in the United States.”

The episode underscores the tension between police agencies that say Flock cameras help them catch criminals and citizens who worry the devices have created an unprecedented surveillance state. As Sistilli discovered, the middle ground between the two sides is narrow.

Turning the tables

Flock is believed to operate more than 120,000 devices in 49 states. Those cameras photograph the license plates of every vehicle that passes, adding the tag number and other details to a massive database that can often be accessed by law enforcement agencies in different cities and states than where the images were originally collected.

Sistilli thought that creating a device of his own — one that turned the tables on law enforcement — would provide the public with a new perspective on the technology.

“I never intended to create a giant network mapping every police officer, and I still don’t,” Sistilli told Straight Arrow. “I wanted to see whether it was feasible and use the irony to illustrate a point. Sometimes it’s easier to show people what this kind of surveillance means by turning it around than by simply explaining what companies like Flock are doing.”

Sistilli did not expect his first short clip, uploaded on Sept. 4, to garner much attention. As of October, however, it has been viewed on YouTube more than 1.6 million times. Over on Instagram, the clip has racked up 523,000 views.

Build-a-Flock

The device he built, Sistilli said, is quite basic and relies on a relatively cheap camera purchased from Amazon, his laptop and machine learning software. The software, which Sistilli says he built with artificial intelligence in about 20 minutes, scans the footage in real time, looking for anything that resembles a police vehicle.

Unlike license plate reader cameras made by Flock and other companies, which photograph every plate that passes by, Sistilli’s system is designed only to attempt to read a license plate when the vehicle is believed to belong to law enforcement.

“If the software is confident that a vehicle is a police car, it tries to read its licence plate,” he said. “My phone supplies GPS information to the laptop, and everything gets stored in a database so I can look at it on a map later: the picture, the time, the location and the licence plate.”

While testing his setup, Sistilli first aimed his camera at images of police vehicles and regular cars on his computer. He then visited a nearby police building, where he says he stood in an intersection for an hour.

“I had a giant backpack with a portable power supply inside, my laptop open and the camera connected to it,” Sistilli said. “It picked up plenty of regular cars, but I didn’t actually see any police vehicles. It turned out I’d chosen a headquarters rather than an operational station.”

After publishing his first short video on the project, a viewer left a comment suggesting he visit a police station in Brampton, Ontario. There, he said, he recorded two of the five police cars that passed him in an hour. He uploaded his second short-form video uploaded to YouTube on Sept. 9.

Because his camera was not affixed high up on a pole like a normal license plate reader, and was much less sophisticated, he said he failed to capture several police vehicles. In some cases, the camera’s view was blocked or the cars were too far away for his camera to accurately read the plates.

Sistilli said many commenters on his second video noted that Canadians are slowly recognizing the proliferation of license plate reader cameras, albeit not operated by Flock, across their country.

Others, Sistilli said, had more concerning remarks.

“A lot of people supported it, and some wanted to build a shared map,” he said. “But there was also the darker side of social media, with people talking about violence or tracking down individual officers. I don’t condone any of that.”

A knock at the door

Sistilli had tested his rough prototype for only a few days when the Peel Regional Police officers came to his door.

“I had collected very little data and hadn’t publicly released actual licence plates or identifying information about individual officers,” he said. “It was essentially just, ‘I saw a police car at this location.’ I explained that what I’m really advocating for is more transparency about how the data is being used.”

The visit, he said, was “incredibly ironic.”

The officers, Sistilli said, “were concerned about how transparency could work without compromising ongoing investigations.”

“I brought up approaches used in the U.S., such as clear data-retention policies and audit logs that don’t expose personally identifying information or details of specific cases,” he said. “I don’t think I managed to get that distinction across.”

Round out your reading

• Lawyer says 57 Cornell frat members were in the group chat. Could their silence be a crime?
• Florida county discovers mysterious Flock cameras with no obvious owner.
• Off Script: They won their primaries on one message. Now it’s disappearing.
• Some question whether a popular affordable housing policy is making the nation’s home shortage worse.
• Kelvin wave incoming: Why the US should care.