Breaking News
More () »

FOX CT Investigation: Is Costly Gunshot Detection System Worth the Cost?

It’s touted as a high-tech way to curb gun violence but some say the system that Connecticut cities are spending hundreds of thousands of dollars on, is a drain...

It’s touted as a high-tech way to curb gun violence but some say the system that Connecticut cities are spending hundreds of thousands of dollars on, is a drain on taxpayer and police resources. The problem is that the system has struggled to identify gunfire, picked up false positives and in some cases, sent police to the scene for nothing.

After some adjustments, the system has seen vast improvement, but the question remains, whether it’s worth the money…

For far too many Connecticut cities the sound of gunfire is a nightly ritual. For people like Jessica Figueroa who grew up in Hartford, shootings were the norm.

“It’s horrible. I saw my cousins get shot, seen my father get shot in front of me when I was a little girl,” Figueroa said.

But in 2011, the city introduced a new shooting task force and a technology to boost public safety, that was ushered in as a way to take aim at Hartford gun violence. It’s called the ShotSpotter system, built to detect gunfire and it’s also used in New Haven and Springfield, Mass.

ShotSpotter is a series of sensors set high above the most dangerous streets on buildings and light posts. It uses high-tech “acoustic triangulation” to pinpoint the location of gunfire and alert dispatchers to exactly where shots were fired.

“It gives us a very, very close location as to where the shots emanated from,” said Lt. Brian Foley of the Hartford Police Department.

But that’s only when it’s working and the problem is that so far, ShotSpotter has mostly backfired. During an analysis of ShotSpotter in spring 2012, police records show that out of 60 total alerts, only six were confirmed, meaning the system was only 10 percent accurate. Nearly a year later, an interdepartmental police memo shows the system’s accuracy on 27 alerts was even lower, at just eight percent. Two of those 27 alerts were labeled as gunfire but really weren’t, including one which was just noise from a snow plow.

“Some calls you go to you don’t know if it’s a firecracker, you don’t know if it’s a gunshot, you don’t know if it’s a dump truck swinging its tailgate,” Foley said.

The New Haven Police Department purchased ShotSpotter for $374,000 in 2009 and has made two payments of $49,000 for maintenance since then, yet police admit the system has been flawed.

“It was definitely frustrating to officers and dispatch, it was taking up a lot of our resources,” said Sgt. Max Joyner.

Despite the faulty alerts and low accuracy at times, Hartford paid $75,000 for their system and is considering another $75,000 payment to renew. New Haven Police Chief Dean Esserman, shut down ShotSpotter at one point and considered discontinuing its use, but after their nearly half a million dollar investment in ShotSpotter it’s still up and running, regardless of concerns from officers and dispatchers.

“Not knowing if we were getting gunshots or if a dump truck went by,” said New Haven dispatcher Cathy Sargent.

FOX CT asked Sgt. Joyner whether ShotSpotter is worth the cost.

“Yeah I believe so… I do believe it’s worth it. Why? I don’t know, I couldn’t tell you why,” Joyner said.

Despite concerns about ShotSpotter’s accuracy the past couple of years, in New Haven and here in Hartford, both police departments say it’s improved, to the point at which they’re willing to give it another chance.

“We’re told it’s a learning system, that it’s an intelligent system and it learns from its false readings,” Foley said.

The company’s data says that during a one-month trial of the “ShotSpotter Flex System” this winter, it displayed 100 percent accuracy. Why the sudden improvement?  After years of problems, ShotSpotter reps flew to Hartford and New Haven to change the system.

Now noises recorded by ShotSpotter are sent directly to company analysts in California who discern whether it’s really gunfire and if so, send an alert to dispatchers back in Connecticut within 60 seconds.

Dispatchers we talked to say they haven’t had any problems since the change. But one Hartford attorney sees it differently.

“I would think $150,000 for the technology is a poor investment,” said Corey Brinson.

Brinson is a criminal defense attorney who was a Hartford City Council when ShotSpotter was announced in 2011.  He disagreed with the purchase then and now, after we showed him the system’s accuracy analysis. We asked him whether the numbers are troubling.

“Yeah, the numbers are very troubling,” he said. “It’s actually a waste of resources on top of a very expensive technology.”

Hartford City Councilman Kyle Anderson disagrees.

“Is it money well spent?  I think it is,” he said. “It does concern me, but not to a point where I’m going to pull the plug.”

Anderson said he is encouraged by the system’s improvements and wants to see it through, even with the $75,000 price tag.

“Even if we got one person convicted of a shooting incident, a murder; isn’t that worth what we paid for it?”.

For Figueroa, who knows the devastation of urban gun violence first hand, the city is on the right track with ShotSpotter.

“I think it’s awesome that they’re putting that up on the buildings,” she said. “I really do. If it’s gonna help then… go ahead.”

Harford Polices say the initial results with ShotSpotter came during a testing phase from 2011 until 2013. The company behind ShotSpotter said the following in response to low accuracy numbers:

*Data collected prior to January of 2013 was before the system was approved or “live”.

*The system was used before being tested, qualified and approved for a “go-live” date.

*ShotSpotter has learned to get the system fully operational before allowing any use.

*Without ShotSpotter alerts, 96 percent of gunfire incidents in Hartford would have gone unreported, therefore meaning no police response.

Editor’s note: The company that owns ShotSpotter disputes the use of the word “accuracy” in characterizing the system’s performance. They say the 8% and 10% figures described as “accuracy” ratings are far lower than the system’s true accuracy.

Before You Leave, Check This Out