In some cities, second thoughts about gunshot detection sensors

This article was originally published by Undark.

More than seven years ago, when the city of Chicago began its broad deployment of acoustic technology to identify and locate gunfire in high-crime neighborhoods, supporters promoted the system — which uses acoustic sensors, GPS software, and machine learning algorithms to alert the police in real-time — as an effective way to reduce handgun violence.

At that time, more than 90 cities had adopted the technology developed more than 25 years ago by a San Francisco Bay area firm called ShotSpotter. Today, more than 160 cities are using the ShotSpotter technology, according to its parent company, now called SoundThinking.

But from the beginning, critics have questioned ShotSpotter’s actual impact on handgun violence, its accuracy, and the reliability of SoundThinking’s closely guarded, proprietary data. Now a growing number of cities — including Chicago, believed to be one of ShotSpotter’s largest markets — are having second thoughts or abandoning their commitment to the strategy.

Brandon Johnson, who was elected mayor of Chicago in 2023, vowed during his campaign to curb some mass surveillance tools such as speed cameras and gunshot detection technologies. Johnson has described the ShotSpotter system as expensive and ineffective.

In late May, the City Council voted to take control of the contracting process for the city’s ShotSpotter system from the mayor’s office, effectively keeping the technology in place until at least November, when the contract is set to expire. But the council also ordered the Chicago Police Department to collect more data on ShotSpotter’s effectiveness before the contract is renewed. “We haven’t seen gun violence go down as a result of this system,” Jessie Fuentes, an alderperson, said during the meeting, according to the Chicago Sun-Times. “We only get reports of the sound detected.” 

“The mayor ran on a platform and he believes that he is doing what he should,” Tom Chittum, SoundThinking’s senior vice president of forensic services, said in an interview with Undark. “But I think our view is removing ShotSpotter from Chicago would be a tragic mistake that would just result in more lost lives in a city that's already suffered too much gun violence.”

In 2017, when Chicago began expanding ShotSpotter into about half of the city’s neighborhoods, mostly in Black and Hispanic communities with higher rates of firearm violence, there were very few independent studies on gunshot detection technology in general and none on Chicago’s s system.

Since then, there has been a steady stream of research into the technology’s effectiveness. According to a 2024 analysis by an interdisciplinary research team at multiple institutions including Northeastern University and the University of Tampa, the system has had “no effect” on both fatal and non-fatal shootings.

Researchers have reached similar conclusions in several other cities, including Kansas City, Missouri; Portland, Oregon; and Durham, North Carolina, where officials recently voted to end the city’s gunshot detection pilot project.

Meanwhile, an audit conducted by the New York City Comptroller in June found that ShotSpotter “is overwhelmingly inaccurate and leads officers to spend hundreds of hours each month investigating nonexistent shots,” according to The New York Times.

“Unsolved homicides and non-fatal shootings are exceptionally damaging to public safety and the community's trust of police,” Ajenai Clemmons, a University of Denver public policy professor whose research focuses on police-community relations, wrote in an email to Undark. “If there are better strategies and tools for solving these cases, then money should be shifted there rather than to a technology that is expensive to install and maintain, whose benefits remain unclear, and whose effects would always need to be paired with changed officer behavior in order to work.”

The ShotSpotter system uses a battery of microphones and sensors installed high above street level to record, identify, and locate the source of potential gunshots. The sensors are usually on rooftops, the sides of buildings, or on top of streetlights or utility poles. They are trained to recognize the two distinctive audio signatures of gunfire. The first is the muzzle blast from the bullet exiting the chamber. The second audio signature is the distinctive waveform.

The technology determines the location of the gunfire — usually within several feet — using readings from three or more sensors. Algorithms use the speed of sound to determine the difference in “arrival” times.

A statement SoundThinking public relations manager Jerome Filip provided to Undark, which he asked be attributed to the company, described ShotSpotter as a “vital tool in modern policing” that’s designed to supplement the 911 reporting system. “In more than 80% of all shooting incidents, 911 calls do not come in, leaving victims without necessary aid from first responders. ShotSpotter picks up where 911 leaves off, enhancing police response to shooting incidents, aiding in evidence collection, and ultimately saving lives.”

One notable study of ShotSpotter’s effectiveness in Chicago was conducted by the MacArthur Justice Center at Northwestern Pritzker School of Law, and assessed ShotSpotter dispatches in the city between July 2019 and April 2021. Eighty-nine percent of the notifications did not lead police to evidence of any gun crime. And about 86 percent of notifications led to no criminal charges at all. There were more than 40,000 total unfounded ShotSpotter deployments during that period.

That report was not peer reviewed, but two others that were have appeared this year focusing on the ShotSpotter programs in Chicago and Kansas City, both of which launched their programs in 2012.

The Chicago study, which analyzed about 85,500 gunfire events between January 2008 and December 2019, was led by Eric Piza, a professor of criminology and criminal justice at Northeastern University. It is believed to be part of the largest research project on gunfire detection technology to date. The researchers wanted to determine the impact of gunfire locators on officers’ responses and searches, ballistic, and evidence collection, and on gun violence and crime overall.

The Kansas City research, which was also led by Piza and published in Criminology & Public Policy, examined about 11,500 gunfire events located by sensors in Kansas City through the end of 2019.

“What we found in both cities was that the police response to reports of gunfire seemed to have improved following the installation of ShotSpotter,” Piza told Undark.

He added that in both cities, districts that used ShotSpotter saw increases in the collection of ballistic evidence and the recovery of illegal firearms. And while the police in both cities also spent more response time at ShotSpotter locations as opposed to 911 calls, none of this activity led to a reduction in crime.

“Gun violence victimization did not reduce in either Kansas City or Chicago following the installation of ShotSpotter,” said Piza. “We also found that shootings were no more likely to be solved in either Kansas City or Chicago following the installation of ShotSpotter.”

Piza said he was surprised by some of the results. He assumed that there would be a higher clearance rate for shootings. “And we just didn’t find that.”

Clemmons, the University of Denver public policy professor, wrote that “despite its rigorous method, this study echoed the results of many previous studies of gunshot detection technology (GDT) — that it seems to have no effect on serious gun crime.”

“Without a positive impact on gun violence prevention, the next best benefit we could hope for would be improvements in clearance rates of violent gun crime cases,” wrote Clemmons, who was not involved in either study. “What we still need to know is whether the increased number of guns recovered by police translates into more cases solved and more shooters charged.”

Many cities are struggling to assess the costs and benefits of the technology.

Cleveland recently expanded the technology across the entire city, but other cities like Seattle, Washington and Dayton, Ohio have abandoned their programs.

City officials in Durham voted in March to reject a three-year contract extension for ShotSpotter after an evaluation from the Wilson Center for Science and Justice at Duke University School of Law. The study raised questions about ShotSpotter’s accuracy and impact on crime, even as it found the technology more than doubled the number of gunfire notifications, reduced police response times by more than a minute, and resulted in increased arrests. The study, which was not peer reviewed, concluded that it was unable to determine if the technology had prevented or reduced crime.

“At the end of the day we can’t say with great confidence that there were shootings that were prevented as a result of the ShotSpotter installation,” said Philip Cook, a Duke University economist and criminologist, and author of the study. Even so, he added, “The possibility of taking a minute or two off the response time would mean the difference between life and death for a victim who's bleeding out on the scene.”

Other cities, including Houston — the nation’s fourth-largest city — are rethinking the merits of gunfire detection. John Whitmire, the city’s mayor, recently described the technology as a “gimmick” with no impact on the city’s serious gun violence epidemic, according to ABC 13 Houston. The news station found that only 5 percent of ShotSpotter notifications over nearly two years led to an arrest.

Researchers and policy analysts like Clemmons emphasize that policing technologies cannot work in a vacuum. “Whether it's body-worn cameras, CCTV, or an Apple Watch tracking your heart rate, changes in policies, procedures, and — most importantly — practices must accompany any tool in order for it to make a meaningful difference,” she wrote in an email to Undark. “Then the question becomes: if you're making all of these other changes to the way you're doing things, what is the value-add of the technology?”