ShotSpotter Lets Police Pinpoint Exactly Where A Gun Was Fired

In the criminally underrated Schwarzenegger film Last Action Hero, there’s a scene where the villain (played brilliantly by Charles Dance, a.k.a. Tywin Lannister) from a cheesy action movie enters the “real” world. After witnessing a mugging for which there was no consequence, he proceeds to shoot someone in cold blood and laughs as he hears nothing: no sirens, no police telling him to drop his gun. Nothing. That’s when he embarks on his plan to run loose in a world where “the bad guys can win.”

Ironically, in many cities around the country in the actual real world, this scene would play out quite differently. In Milwaukee, for example, if you fire a gun in some neighborhoods you might find yourself surrounded by police in just a few minutes.

That’s thanks to a technology deployed in Milwaukee and 74 other cities both in the United States and other countries created by ShotSpotter. The company, which was founded in 1995, uses an array of acoustic sensors deployed in a neighborhood that listen for one thing: the sound of gunfire. The sensors are paired with software that can analyze the sounds in real time, triangulating and pinpointing the location of each round fired down to the latitude and longitude. The software also filters out non-gunfire sounds that might have triggered the sensors, and the sound is also sent to trained officers who act to confirm that it’s gunfire before an alert is sent to officers on patrol.

In the first quarter of 2013, the company told me, their system detected 7,584 separate gunfire incidents in 30 cities they pulled data from. In those incidents, 26,617 rounds were fired – about 3.5 rounds per incident. One of those incidents saw 63 rounds fired. 63% of gunfire, they found, takes place between 8pm and 4am (local time) with 54% of that gunfire happening on Friday, Saturday and Sunday.

That’s data that law enforcement finds useful – especially for predictive policing.  In the City of Milwaukee, police use the system as part of a bigger strategy for fighting crime.

“We weren’t interested in technology for technology’s sake,” Milwaukee Police Chief Edward Flynn told me. “We wanted to ensure that it assisted us in accomplishing our mission. Which is to help people live in a safe neighborhood so they can raise their children and pursue the American dream.”

“Our strategy,” he continued. “is community based and problem oriented. We’re looking for the root issues that crime more likely. We spend a lot of time analyzing data to find problems and measure success.”

One measure of that success is just catching bad guys in the act. The Milwaukee police have used ShotSpotter to catch several criminals shortly after they fired shots. ShotSpotter data has been used to obtain convictions in several cases, including a murder conviction. The latitude and longitude locations also help police recover shell casings, which are used to provide forensic evidence in court. But Flynn doesn’t only see ShotSpotter as a tool for a fast response.

“It does that, and it does it well,” he said. “But we also want to use it to build up neighborhoods. People in the community are very supportive, because it enables us to catch people before a 911 call is made. It also lets us catch people without neighbors being accused of ‘snitching.’”

That latter point is important when you consider that in the neighborhoods where ShotSpotter is deployed, Milwaukee police determined that only about 14% of them were reported to 911. The company told me that similar numbers have been seen in other neighborhoods where the system is deployed.

Chief Flynn is also seeing measurable results. The number of shots fired in the area where ShotSpotter is deployed has seen a year to date reduction of 18%. They’re also using the patterns of the gunfire to identify drug markets and gang battles so that the police can effectively intervene in the area.

“Gunfire is a symptom,” he said. “It often means something else. We use it in conjunction with other data to pinpoint the intelligence we need.”

That support from law enforcement has enabled the company to successfully sell its services to more and more communities over the years, as well as attract investors. The company has raised $62 million from its investors (but would not disclose its revenues).

Despite the endorsement of law enforcement, though, not everyone is convinced about the usefulness of the technology. That’s largely because independent data about ShotSpotter is pretty thin on the ground.

“It’s a neat technology,” University of California Berkeley Criminologist Frank Zmiring told me. “But the real question is the cost and benefit for police response. There are very effective ways of evaluating this, but it’s not clear that this has been done.”

“In general, though, policing is so goddamn labor intensive that any technologies that displace human labor would be terrific,” he continued. “Whether this is worth it is an answerable question. The problem is that there is no department rigorously studying this kind of urban policing.”

One study that has been performed was in 2008 of the SECURES system, which was purchased by ShotSpotter in 2009. That study was led by Peter Scharf, a criminologist at Tulane University. That study found some high number of false positives in one area where it was deployed, especially when fireworks were set off. However, in another area of deployment, they found a higher level of accuracy.  That system, however, didn’t have a human listener to analyze shots the way the current one does.

“They’re trying to avoid the false positives by having a human person listen in,” Scharf told me. “But there needs to be a study done to see how that improves the accuracy.”

For Scharf, one of the biggest questions is whether the return on investment is worth it. It may be, he said, for bigger cities. “But you might have to seriously ask if it’s rational for smaller markets.”

That said, he does think that the technology is worth developing. “I think that this is like the airplane in 1912,” he told me. “It’s a promising note on a series on integrated technologies that are coming down the pike.”

ShotSpotter itself, however, is eager to provide as much data as it can. “We are passionate about data at the local level to help cities address gun violence issues,” ShotSpotter President and CEO Ralph Clark told me. “Now we’re excited about a second chapter for our company to aggregate data and help inform policy decisions and discussions happening at a Federal level.”

“The data that local law enforcement agencies have is limited, incomplete and noisy,” he continued. “We want to tell the ground truth as to where they were deployed.”

Data aside, though, anyone reading this article so far has probably wondered about what Scharf calls the “1984 issues.”

“It’s a little creepy that you can police without touching the community,” Scharf told me. “So there’s a sci-fi part of the technology that bothers me a little bit.”

In light of the debate over the NSA’s surveillance operations, I had to admit that the privacy issues bothered me, too. So I asked Clark about how the system impacts the privacy of people in the neighborhoods where ShotSpotter is deployed.

“That’s a timely question,” he said. “But first of all, this is a passive technology. We refer to them as sensors rather than microphones because it’s important distinction. The system is programmed to detect impulsive noises. In order to get what we need and not have them going off all the time, we have to deploy them about 30 feet in the air.”

“We don’t record everything,” he continued. “We’re just listening to the ‘booms and bangs.’”

For the company, though, the bottom line is about assisting local law enforcement in making safer neighborhoods.

“We’ve become a source of data that police don’t otherwise have,” Clark said. “We allow them to analyze which of their neighborhoods have the most gun violence so that police can mobilize the right resources.”

And maybe, in the end, make sure we live in a world where the bad guys can’t win.

Labels: Literature, Technology