How technology that detects gunfire is helping Pittsburgh police

The system, known as ShotSpotter, uses acoustic sensors and software to determine whether and where a gunshot has been fired. Pittsburgh police credit it with helping them arrest a suspect in a recent murder case. But the impact of predictive technologies has been debated by researchers.

Statistics are posted for PredPol, a predictive crime algorithm used to map hotspots for potential crime in the city at the at the Tech Innovation Center in the Atlanta Police Foundation in January 2015 in Atlanta. A sensor-driven technology that tracks gunshots called ShotSpotter has been hailed by police in Pittsburgh.

Ann Hermes/The Christian Science Monitor/File

January 27, 2016

Tracking technology – which uses acoustic sensors and software to listen for the sounds of gunfire and alert police about the location of each round fired – is being hailed as highly successful in Pittsburgh.

As the Pittsburgh City Council weighed extending a contract with California-based SST Inc., which makes the technology, known as ShotSpotter, on Tuesday, police said it helped them quickly learn of a recent murder case and apprehend a suspect.

“We believe that it [has] been very successful,” said Pittsburgh Councilman Ricky Burgess at the meeting, WPXI reports. "It is already helping solve crime. We believe it reduces crime and has increased public safety."

Ukraine’s Pokrovsk was about to fall to Russia 2 months ago. It’s hanging on.

Police say they were able to respond quickly to the shooting of 29-year-old Janese Jackson Talton at a bar in the city’s Homewood neighborhood because of an alert from the system.

After pursuing a silver car that was seen leaving the scene, they were able to apprehend the car’s driver, Charles McKinney, after chasing him through a local neighborhood. Documents show that Mr. McKinney allegedly shot and killed Ms. Talton because she had declined his advances, WPXI reports.

ShotSpotter works by using a series of acoustic sensors position in a particular neighborhood to listen for the sound of gunfire. Software analyzes the sounds to determine whether it is gunfire and then triangulates the location of each round fired, including determining a precise longitude and latitude.

A human technician then reviews the findings before an alert is sent to police in the neighborhood, Forbes reported in 2013. The technology can alert police of shots being fired within 30 to 45 seconds.

The following year, the company tracked an average of 105 gunfire incidents each day across 47 cities, or 4.4 incidents every hour.

Howard University hoped to make history. Now it’s ready for a different role.

“We weren’t interested in technology for technology’s sake,” Milwaukee Police Chief Edward Flynn told Forbes. “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.”

But technology-assisted policing has also proven controversial when it has been adapted to other tasks, such as predicting the risk of future violent crimes by analyzing past data. Researchers caution that while data analysis can be useful for determining patterns for some types of crimes, such as burglary or theft, for violent crime it instead yields information about who is at a high risk of violent victimization, not a list of potential offenders.

“Thinking that you do prediction around serious violent crime is empirically inaccurate, and leads to very serious justice issues. But saying, ‘This is a high risk place,’ lets you focus on offering social services,” David Kennedy, a professor at John Jay College of Criminal Justice who has led efforts to combine data analysis with face to face meetings to deter crime, told the Monitor in October.

Professor Kennedy says the risk is that so-called "predictive" analysis could yield false positives that could lead police to arrest people who triggered an alert system, but who were not actively involved in a crime.

The ShotSpotter technology has also faced questions about false positives. One issue with an earlier system was that it could incorrectly label fireworks as gunfire, a 2008 study found. The study, conducted by Tulane University criminologist Peter Scharf using data from two Virginia cities, found that the accuracy depended on where it was deployed. But that system – later purchased by SST – didn’t have a person listening in, he told Forbes, saying technology that incorporates human analysis was deserving of further study.

In Pittsburgh, Mr. Burgess, the city councilor, said ShotSpotter was helping police focus their attention on neighborhoods hit particularly hard by crime. The system, which has been in place since 2014, would cost the city $135,000 to renew for another year. The council is set to discuss the funding next week and vote on it in two weeks, according to WPXI.

“I think the system speaks for itself. It's an incredible technology. It is helping the East End solve multiple homicides,” Burgess said on Tuesday.