When a man went on a shooting spree two weeks ago in Fresno, Calif., police there credited his quick arrest to the gunshot detection system it installed two years earlier.
Using audio sensors placed around the city, the ShotSpotter detected 16 gunshots fired in two minutes during the April 18 shooting, alerting police to the exact location the shots came from, police there said. Within minutes, the gunman was arrested and charged with killing three men that day and another days earlier.
Allentown looked into that technology a decade ago as ShotSpotter, a Silicon Valley company, was just taking off. After watching an impressive demonstration, Allentown officials shelved the plan amid concerns it might not work with the city’s network of surveillance cameras.
“I wasn’t really sold on the fact that it worked as advertised,” Mayor Ed Pawlowski said recently.
York was the first city in the state to use the technology, in 2008, but not for long. After about a year, the city determined the cost exceeded the system’s usefulness, said Henry Nixon, City Council vice president.
Trenton, N.J., and Atlantic City also found problems in their first years using ShotSpotter, one getting too many false alarms and the other missing shots-fired calls.
Allentown is reconsidering now that the system has improved and the cost has dropped.
“I am much more confident today than I was a few years ago that they can provide what they say and deliver what they say,” Pawlowski said. “The technology is evolving quickly.”
Allentown interim police Chief Glen Dorney will look at the system by visiting the New York City and Boston police departments, which use ShotSpotter.
“I want to see how they operate and look at some of the pros and cons of the system, to see whether or not the money spent is going to be something that we are truly going to move forward with,” Dorney said.
Eric L. Piza, a researcher and assistant professor at John Jay College of Criminal Justice in New York City, is evaluating ShotSpotter systems for the Newark, N.J., and Kansas City, Mo., police departments.
In reviewing ShotSpotter studies done by other researchers, Piza said he found that the technology hasn’t reduced gun violence. But, he said, that shouldn’t be the only reason a department would purchase it.
“The question then becomes what are the other benefits besides crime reduction,” he said. “It may decrease response times, it may pinpoint hot spots of gun violence otherwise not detected.”
Piza, who also specializes in surveillance camera studies, said if a community is only looking at the gunshot detection technology to reduce crime, “your money is better spent elsewhere.”
In March 2007, a training officer fired a 9 mm submachine gun from a hidden location at the Allentown Police Academy’s outdoor shooting range and in less than five seconds, a camera linked to the ShotSpotter system swiveled to the sound of the gunshot, quickly finding the gunman. The demonstration had police officials buzzing about how the futuristic technology might reduce gun violence.
Police officials were looking for solutions, having just endured two of the city’s most violent years. In response, the city started installing the camera system that summer that operates in several sections today.
With the gunshot detection system linked to the cameras, Pawlowski said he could envision police swooping in on criminals and making quick arrests.
But reality never matched the demonstration, he said, and the cameras didn’t work with the system the way he hoped.
Then, the city was told it would have cost about $400,000 to cover two square miles. A recent quote from the company covering almost twice that area had the price at $272,500, which includes a one-time setup fee, Dorney said.
More than 90 police departments, including many in the nation’s largest metropolitan areas and some in England, use ShotSpotter. Pittsburgh is the only city in Pennsylvania using it.
About 15 to 20 audio sensors, which are actually small microphones, are placed in every square mile to triangulate gunshot locations and alert police. The sensors capture a snippet of audio that may be gunfire, which is then filtered through the system’s algorithms and verified by a trained gunfire expert at ShotSpotter’s incident review center in California. The alerts, which include the sound’s time and location, can let police know if the gunfire came from an automatic weapon, according to the company. It takes 20 to 45 seconds from the gun being fired to the alert reaching police.
New York City added the system in 2015 with a pilot program covering 10 square miles. It is expanding to cover an additional 40 square miles, according to published reports.
In the first year, ShotSpotter alerted New York City police to 1,672 gunshots. Three out of four were never reported by 911, the department tweeted last year. Those alerts led to 21 arrests and the recovery of 32 guns, the company reported.
In Fresno, Chief Jerry Dyer told reporters after the April 18 shooting spree that without ShotSpotter helping to quickly track alleged gunman Kori Ali Muhammad, more people may have died.
But while there are success stories, other departments reported mixed results.
York stopped paying an annual maintenance fee for the system after finding it was not always accurate, Nixon said. That fee can range from $65,000 to $95,000 per square mile, according to ShotSpotter’s website.
The system worked about half the time, Nixon said, and was useful when it worked.
“I can think of several cases that clearly identified a shooting,” he said.
The year after Trenton began using the system in 2009, officials there determined that two out of every three gunfire detections were false alarms, according to published reports. The city later expanded the system, working out the kinks by using ShotSpotter’s experts, instead of dispatchers, to verify that a bang is actually gunfire, according to the reports.
The Press of Atlantic City reported that when Atlantic City added ShotSpotter in 2013, the system missed several shootings in one section of town, so it corrected the problem by adding audio sensors to that dead spot.
While ShotSpotter isn’t the only gunfire detection company, it has a strong market share and is about to go public with an initial stock offering. Calls to the company were not returned last week, but on its website, ShotSpotter addresses the “few isolated” failures experienced by some police departments:
“The very few cities that did not have a positive experience had poor practices around the following: not responding to gunfire alerts when they come into PD from the ShotSpotter alert and not responding to the location on the map; not including ShotSpotter data as part of an overall gunfire intelligence and crime reduction program; and lack of community engagement.”
In Allentown, Dorney said he’s aware of the issues some departments had with ShotSpotter, which is why his department has taken so long to pull the trigger on getting the system.
He said some cities had problems because they were not using the system with cameras or were using an earlier version that didn’t include verification experts, leading to officers’ getting dispatched to calls of backfiring cars or doors being slammed.
Piza said for the system to be effective, it requires an investment in human resources. That includes having someone in the department analyzing and addressing the increased gunfire numbers, adding additional patrols to areas identified as hot spots and having more interaction with residents in gunfire-plagued neighborhoods.
In Allentown, ShotSpotter would cover the downtown area, from the northern city line to Martin Luther King Jr. Drive and from the Lehigh River to possibly as far as 15th Street, he said.
Allentown has seen an increase in shootings this year compared with last year, with 19 people shot and three killed through mid-April. However, the number of gunshot complaints was considerably lower this year, from 123 in 2016 to 90 through mid-April, Dorney said.
That could mean there have been fewer shootings or that people are reporting fewer of them, he said.
There have been cases, he said, when officers find shell casings on a street, but never get a report of gunfire. Other times, a gunshot victim will get dropped off at an emergency room but not say where the shooting happened, leading to a critical delay in the investigation.
ShotSpotter, Dorney said, could give a more accurate total of gun use and increase the recovery rate of evidence.
“Now, oftentimes, community members will not call or say something because they will rely on someone else to call and say something,” he said. “They figure it’s not their place.”
With ShotSpotter, shooting reports would likely increase, which has been the norm in cities just starting out with the system, he said.
Dorney said that is why educating residents and city leaders about ShotSpotter and being transparent with the numbers is important. More reports of shots fired wouldn’t mean the city is becoming less safe, he said. It could mean it is becoming more aware.