Community safety technology startup Flock Group Inc., better known as Flock Safety, says it’s pioneering the use of artificial intelligence to detect illegal sideshows – those dangerous, impromptu events, usually at night, where drivers of muscle cars take turns doing donuts in intersections for cheering crowds.

The startup says it has integrated AI and machine learning with its existing Raven audio detection system to provide law enforcement agencies with real-time alerts so they can respond to sideshows as soon as they begin.

Sideshows are a growing phenomenon. Although they have been going on for years, their popularity began to increase dramatically during the COVID-19 pandemic. Also called “exhibition driving,” these events can draw crowds of hundreds of cheering fans, but as fun as they might be for the drivers and attendees, they can also be extremely disruptive.

Not only do they block traffic intersections, but they disturb residents nearby, and they can also endanger lives. Daring spectators like to make a game of getting as close as possible to the spinning cars, taking close up videos to post on social media, and the collisions that sometimes result are popular viewing online.

 

View this post on Instagram

 

A post shared by Bay Area, CA (@sideshowshiestyy_)

Although cities have spent millions of dollars on public safety initiatives in an effort to deter sideshows, Flock Safety believes that such events are becoming more popular and widespread.

That’s why Flock Safety founder and Chief Executive Garret Langley believes there’s an opportunity for AI to help play spoilsport. He argues that sideshows create serious public safety issues that put entire communities at risk.

“Flock Safety’s mission is to equip law enforcement with innovative solutions to keep neighborhoods safe,” he said. “Our sideshow detection technology is a groundbreaking tool that enables agencies to act faster and prevent these dangerous events before they escalate.”

The startup is no stranger to helping law enforcement, selling automated license plate recognition, video surveillance and gunfire locator systems. It was founded in 2017, and operates these systems under license to police forces, neighborhood associations and private businesses, with customers in more than 4,000 U.S. cities. It claims that its systems help to solve about 2,000 crimes per day on average.

Its sideshow detection technology represents an innovative new use case for AI, helping law enforcement agencies to be more proactive in shutting down such events before too many people converge on the area. One of the challenges for police is that, when the crowds are too big, they can often become extremely hostile. By alerting police as soon as the events begin, Flock Safety says they’ll be able to arrive on the scene faster and take action before large crowds can gather.

The technology can be installed at intersections where sideshows are common and it works by detecting the distinct tire screeches made as drivers start performing donuts and other stunts. According to Flock, the AI system can easily distinguish between donuts and normal traffic sounds, such as when a driver brakes a little too hard, to avoid false alarms. So police will only be alerted when there’s a genuine meet taking place.

Using the system, law enforcement officers can listen in on the audio recordings as soon as they’re generated to confirm that a sideshow is taking place. It will then pinpoint the exact location on a map, enabling them to plan their response. Moreover, the technology can integrate with Flock’s existing products, such as its Drone-as-a-First-Responder system, which can be on the scene within an average of just 86 seconds to record video evidence of the events.

Holger Mueller of Constellation Research Inc. said Flock’s new offering might be a tad controversial if it involves placing microphones all over our cities, but believes it’s an interesting vertical use case nonetheless. “It’s a novel idea, using AI to alert police to illegal sideshows the moment they begin,” he said. “It’s an innovative way to increase public safety, but it remains to be seen how well it can be implemented in practice.”

Flock’s license plate recognition system can further assist law enforcement by identifying the vehicles present at sideshows, so officers can track down their drivers and take action against them, if necessary.

Image: SiliconANGLE