ShotSpotter Acoustic Detection System Another Example of a Forensic Tool Shrouded in Secrecy and Prone to Questionable Results
by Casey J. Bastian
ShotSpotter markets technology that uses microphone sensors to detect the specific acoustic signature of gunshots and then record the time and location of the signature it detects. The technology can also detect firecrackers, car backfire, and other similar sounds—not just gunshots. It is up to a human analyst—who ShotSpotter hires as an “acoustic expert”—to accurately interpret the data for law enforcement purposes.
Currently, more than 100 cities in the U.S., South Africa, and the Bahamas use the ShotSpotter technology. According to the Electronic Frontier Foundation, this makes ShotSpotter the largest vendor of acoustic gunshot detection systems. ShotSpotter earned around $34.75 million in 2018 and went public in March 2019. The technological equipment consists of microphones and sensors that are constantly “on,” persistently surveilling neighborhoods and residents. Though typically mounted on light poles or other elevated structures, the systems can be fixed or mobile and both indoors or outside. A main selling point is that gathering such information can help police respond quicker and provide evidence to solve crimes. Now, examples are emerging of just how faulty the ShotSpotter technology really is and how it has endangered people, led to false arrests, and can be used to spy on private conversations between law-abiding citizens. Perhaps the most frightening aspect of this issue is that ShotSpotter competitors like Raytheon use technology to detect small arms fire and have been used in warzones like Afghanistan, both casting further doubt on the integrity of operations abroad and foreshadowing the prospect of increasingly militarized neighborhoods here in the U.S.
On May 31, 2020, Safarain Herring was shot in the head in Chicago. Herring was taken to St. Bernard Hospital by Michael Williams. Herring died two days later. Williams insisted that Herring was a victim of a drive-by shooting. That did not stop the Chicago police from arresting the 64-year-old Williams for the murder. The primary evidence against Williams was video surveillance footage that showed Williams’ car on the 6300 Block of South Stony Island Avenue at the time Herring was shot.
Police claimed this precise data was collected by ShotSpotter devices. But this is not true, according to court filings. On the night Herring was shot, 19 ShotSpotter devices identified a percussive sound at 11:46 p.m. at 5700 South Lake Shore Drive. A motion filed by Williams’ public defender shows that this location is over one mile from where prosecutors say Williams shot Herring. Never mind that the ShotSpotter system’s algorithms originally identified the sounds as fireworks—something dozens of protesters were discharging during the middle of the George Floyd protests in Chicago that weekend. Worse, a ShotSpotter analyst manually “reclassified” the sound as a gunshot after the initial detection alert came in. Months later, a different analyst changed the alert’s coordinates from South Lake Shore Drive to South Stony Island Avenue—where Williams’ car was seen— after “post-processing” of the data.
“Through this human-involved method, the ShotSpotter output in this case was dramatically transformed from data that did not support criminal charges of any kind to data that now forms the centerpiece of the prosecution’s murder case against Mr. Williams,” wrote Williams’ public defender in a Frye motion. A Frye motion is a state court proceeding where it is requested for the judge to determine “whether a particular forensic method is scientifically valid enough to be used as evidence,” according to a Vice article on police interference with ShotSpotter evidence.
After the Frye motion was filed, the prosecutor withdrew all of the ShotSpotter evidence before the judge could even examine it. Afterwards, Williams’ public defender said, “Rather than defend the evidence, [prosecutors] just ran away from it. Right now, nobody outside of ShotSpotter has ever been able to look under the hood and audit this technology.”
Had the Frye hearing actually taken place, it would have been the first formal examination of the ShotSpotter technology or source code by any Illinois court, according to Jonathan Manes, an attorney at the MacArthur Justice Center.
ShotSpotter representatives continue to claim the system is “up to 97 percent accurate.” Paul Greene is a ShotSpotter employee and has testified as an expert witness in dozens of cases in the U.S. In 2017, Greene testified in a San Francisco court that, “Our guarantee [of the technology’s accuracy] was put together by our sales and marketing department, not our engineers. We need to give them [customers] a number.... We have to tell them something.... It’s not perfect. The dot on the map is simply a starting point.” And it is. It’s a starting point for police to aggressively enter neighborhoods on the offensive, justify bogus arrests, and falsely strengthen prosecutors’ cases on the basis of false or misleading evidence.
According to Vice, the alteration of ShotSpotter data is not isolated to the Williams case. In 2016, police shot Silvon Simmons three times in the back in Rochester, New York. Simmons had been a passenger in a car that was pulled over because police were looking for a suspicious vehicle. Not only was the wrong car pulled over, but Simmons was charged with attempted murder and firearms possession when police alleged that he fired at them first. The prosecution’s case relied solely on ShotSpotter evidence. The ShotSpotter sensor did not detect any of the identified sounds to be gunshots and the algorithms ruled that the sounds it did detect had come from helicopter rotors. After the Rochester police contacted ShotSpotter, an analyst claimed that there were in fact four gunshots identified—the exact number of times the police fired at Simmons.
Greene would later testify at Simmons’ trial that he personally found a fifth shot in the data after “he was asked by the Rochester Police Department to essentially search and see if there were more shots fired than the ShotSpotter picked up.” There was zero evidence that Simmons had fired a gun at the scene, and police had refused multiple requests by Simmons to conduct gunshot residue testing on his hands and clothing. The ShotSpotter audio files that purportedly contained the “fifth shot” have since mysteriously disappeared. ShotSpotter and the Rochester Police Department “lost, deleted and/or destroyed the spool and/or other information containing sounds pertaining to the officer-involved shooting,” according to court documents in a separate civil suit filed by Simmons. Vice reported that Simmons was ultimately acquitted of the attempted murder charge, and a judge overturned his weapons conviction during the criminal proceedings.
Another report cited in the Vice article, also from Chicago, occurred in 2018. Ernesto Godinez was charged with shooting a federal officer. According to court records, video surveillance did not show any muzzle flashes from the doorway in which Godinez had been standing, and the casings that were found in the area did not match the bullets that hit the federal officer. Seven shots had allegedly been fired per a ShotSpotter report, but Greene would testify under cross-examination that ShotSpotter only recorded two shots. Upon request from Chicago police, Greene would again “re-analyze” the audio files. Chicago police had asked Greene to “essentially, search for additional audio clips,” according to Greene’s testimony. After being re-analyzed, Green would later rule that the audio files contained five additional sounds that were determined to be gunshots that the system’s algorithms had missed upon initial review.
A statement from a ShotSpotter spokesperson read: “The detailed forensic report is never altered because it is a completely separate process from the alerts. Forensic analysis may uncover additional information relative to a real-time alert, such as more rounds fired or an updated timing or location upon more thorough investigation by forensic analysis. We respond to requests to further investigate an incident for a forensics report only to provide the facts that we can determine and not to fit a predetermined narrative. If we find an error, we provide a more accurate location to the customer to assist the investigation.”
The problem, according to forensicmag.com, is that researchers have proven that forensic bias can occur from the examiner having been given a “narrative” prior to secondary analyses.
Opposition to the ShotSpotter systems arises from the lack of empirical evidence as to whether or not the system produces accurate results and its overall impact on reducing gun crime. “The reliability of their technology has never been challenged in [an Illinois] court and nobody is doing anything about it,” said Gal Pissetzky, Godinez’s attorney. Dennis Mares is a professor of criminal justice studies at Southern Illinois University.
Along with co-author Emily Blackburn, manager of the crime analysis unit with the St. Louis Metro Police Department, Mares published a report in the Journal of Experimental Criminology concerning an evaluation of the ShotSpotter systems and its effect on reducing gun violence in St. Louis: “Despite rapid growth and substantial investments in acoustic gunshot detection systems (AGDS), remarkably few academic evaluation studies on gunshot detection systems have been published,” wrote Mares. In fact, the evaluation revealed that citizen-initiated calls were much more effective, “seven times more efficient in uncovering and responding to criminal behavior,” than the ShotSpotter systems.
Chicago’s current contract with ShotSpotter expires next August, and the city must decide whether to renew it. The current contract costs the city $33 million, making Chicago ShotSpotter’s second biggest customer behind New York City. Citizens want the contract ended. Gun violence is skyrocketing in the city, and worse, the ShotSpotter system caused the police to respond to an alert in the early morning hours of March 29, 2021, where police would end up shooting unarmed 13-year-old Adam Toledo. This incident sparked more outrage about police brutality and heavy-handedness towards the communities the police are supposed to protect.
“Every ShotSpotter alert is putting Black and Latinx people at risk of interactions with police. That’s what happened to Adam Toledo,” said Alyx Goodwin, a Chicago organizer with the Action Center on Race and the Economy.
Asiaha Butler is president of the Resident Association of Greater Englewood, and she believes that when shots ring out in the neighborhood, there is an “over-militarized police presence. You see a lot of them. It’s not a friendly interaction.”
Critics argue that the ShotSpotter systems should stop being used until the real public safety value can be determined. No city in America needs more tragic cases like that of Adam Toledo.
Source: forensicmag.com; vice.com; eff.org
As a digital subscriber to Criminal Legal News, you can access full text and downloads for this and other premium content.
Already a subscriber? Login