By Erin Beresini
A computer screen lights up with red dots showing where shots were just fired in East Palo Alto, Calif. A police dispatcher sees where the incident occurred and listens to the sound of the shots to rule out false alarms, like backfiring motorcycles or firecrackers. Then the dispatcher calls squad cars to the scene.
This is ShotSpotter’s acoustic gunshot detection and location technology in action. ShotSpotter’s main competitor, Safety Dynamics, invented a system that works differently, but with the same goal: to alert public safety officials to the location of a crime involving firearms within seconds of the discharge.
The systems themselves, however, have created political and cultural controversy among law enforcement officials and members of the public, who believe the cost of the systems outweighs the benefit, that the systems’ existence raises privacy concerns, and that the systems’ accuracy is questionable.
Founded in 1995 in Mountain View, Calif., ShotSpotter bases its technology on acoustic data similar to that used by geologists to locate earthquakes. A minimum of three acoustic sensors is placed on poles and rooftops in a high-crime area. When a shot goes off, the sensors send data to a computer that determines through triangulation, or how loud the shot sounded to each sensor, where the shot came from. The ShotSpotter system costs $300,000 per square mile covered, plus a flat annual maintenance, update, and retraining fee of 15 percent of the purchase price.
“That kind of money could be better spent on hiring more police officers,” said Dr. Tom Nolan, associate professor of criminal justice at Boston University, and 27-year veteran of the Boston Police Department. “What might be a more valuable or appropriate use of scarce resources is to put police where the problem is instead of putting technology where the problem is.”
Phil Amicone, mayor of Yonkers, N.Y., announced in March 2009 that his city would team up with neighboring Mount Vernon to buy a $3 million ShotSpotter system, without having first tested the technology in either city. This announcement came two months after the layoff and subsequent rehiring of 11 Yonkers police officers.
“The county supplied the money for the gunshot technology. The money that would have paid for the cops, or prevented them from getting laid off, came from the city—that was the city’s argument,” said Detective Keith Olson, president of the Yonkers Police Benevolent Association.
The ShotSpotter system was installed in late 2009.
“It has not been a success so far,” said Olson. “If I had a choice, I wouldn’t have had it. I would’ve hired many more cops.”
Because gunshot location technology has often been installed in cities as part of larger programs to reduce overall crime rates, judging a system’s effectiveness is difficult.
The City of Boston began experimenting with gunshot detection technology in 2007 in response to increasing homicide rates, ultimately spending $1.5 million to install the system. In 2005, there were 73 reported murders in Boston, according to FBI crime statistics; in 2009, there were 29.
“I believe that this system has the potential to greatly enhance crime-fighting in Boston by providing critical intelligence about specific crimes, as well as crime patterns,” Boston Mayor Thomas M. Menino said in a 2007 press release.
“Any resource that gives us the opportunity to more effectively zero in on criminal activity is a major benefit,” said Police Commissioner Ed Davis in the same press release.
Other factors may have played a part in the decline in Boston’s reported homicides. In 2005, Boston employed 2,075 police officers. In 2008, that number had increased to 2,213.
The City of Chicago began experimenting with the technology in 2003 as part of the Chicago Police Department’s ongoing “Operation Disruption,” aimed at disrupting “drug sales and other criminal activities through highly visible rotating cameras installed on light poles in high-crime areas,” according to a press release. That year, the city had a reported 598 homicides, according to FBI statistics.
The city tested both ShotSpotter and Safety Dynamics systems.
Tucson, Ariz.-based Safety Dynamics, founded in 2003, bases its technology on neural recognition; each box-like sensor is designed to act independently by distinguishing the sound of a gunshot from other sounds like human brain does.
“In the midst of all kinds of noise, the system has been told to listen specifically for gunshots,” said Sally Fernandez, Safety Dynamics’ head of press and publicity. “It will ignore everything else in the background,” similar to how the human brain would recognize the sound of a television as non-threatening background noise.
The Safety Dynamics gunshot location system, dubbed SENTRI for Sensor Enabled Neural Threat Recognition and Identification, played a part in Operation Disruption.
Each SENTRI unit currently costs between $5,000 and $7,000. The units can be operated with or without surveillance cameras, and can be used with whatever communications systems a police department already has in place.
When a shot is fired, the system sends text messages or emails with coordinates of the discharges directly to officers on duty and to the police dispatch center. Because SENTRI units operate independently, triangulating the location of a gunshot within the unit itself, only a single unit is needed to detect a gunshot. The range of a single unit, however, would be limited to about 600 feet.
Chicago Mayor Richard M. Daley announced at the end of 2004 that the city would install 50 surveillance cameras coupled with gunshot location technology costing $2.8 million in high-crime areas. Money confiscated from drug dealers paid for the system. In 2005, there were 448 reported homicides in Chicago—140 fewer than in 2003.
“Installing high-tech surveillance equipment in communities where we are trying to build trust creates a conflict of interest,” said Nolan. “We’re asking the community to trust us and give us information about what’s going on in the community, but we’re installing high-tech gadgetry to listen to what’s going on in the neighborhood.” It’s not too far-fetched, he said, to believe the government could “tweak” the system to pick up people’s voices and conversations.
“Everybody’s going to be worried about the invasion of privacy” if cameras are linked to the gunshot location system, said Kevin Williams, who works in security at PAL Market in East Palo Alto, Calif.
East Palo Alto is the first city to install a gunshot location system covering the entire municipality. The city first tested the ShotSpotter system in early 2008, then expanded coverage to the entire city one year later. The system is only acoustic and is not connected to surveillance cameras.
George Cardenas, alderman of Chicago’s 12th Ward, a working-class section of the city about six miles south of where Chicago tested the SENTRI system in 2004, advocates installing the system—with cameras—throughout the city.
“Technology can have an impact on a community feeling safe,” Cardenas said. “I think having the technology will allow us to deter people from committing crimes, because they’ll know we have accurate information showing where the crime is coming from.”
Adeeb Silmi, 18, works in East Palo Alto. When he found out the gunshot location system was used there, he said, “This is a city with a lot of crime, and to know that there’s a system like this, it does make me feel safer.”
Both ShotSpotter and Safety Dynamics advocate advertising a city’s use of their systems to residents.
“Disclosure is part of the bargain with the city,” said James G. Beldock, president of ShotSpotter. “There’s a combined effort both by the media and by police to get the word out. The result is not only are fewer guns fired, but those who still make the mistake of firing a gun are caught.”
ShotSpotter installed 13 sensors in Redwood City, Calif. in 1996 as both a test of the system, and a response to that city’s problems with celebratory gunfire. The city then created “Operation Silent Night,” an effort to increase awareness of the risks of celebratory gunfire on New Year’s Eve. It was also an effort to advertise the system in place to detect firearm discharges.
“We send fliers home with kids at schools, to neighborhood associations, and put up posters in stores,” said Redwood City Police Sergeant Steve Blanc.
ShotSpotter, however, does not publicly disclose the look and locations of its sensors to keep the sensors from becoming targets. Only 10 of the 2,000 sensors currently deployed in the US have been tampered with, said Beldock.
Safety Dynamics openly displays its sensors, both on the company’s website, and in public to let people know the sensors are there, and they are watching.
ShotSpotter currently claims on their website their technology will locate a gunshot to within 25 meters of where the shot was fired. Blast Proof and Fire Solutions, official distributors of the Safety Dynamics SENTRI system, claims this system will detect a gunshot to within a “one-degree differential” of where the shot was fired. For instance, if the shot was fired 600 feet from the sensor, the location of the shot as determined by the sensor could be off by a maximum of about 21 feet.
“We’re not at the iPod touch phase” of the technology yet, said Fernandez, “but we’re getting closer and closer.”
Just as it happened with cell phones, gunshot location technology is becoming more dependable and cost-effective over time, said Fernandez.
One concern police officers have with the technology is the frequency of false alarms that set off the system. Because the system reacts to loud noises, fireworks and backfiring vehicles can set it off.
Detective Olson said Yonkers’ system has “so far been a very good fireworks detector. It’s still kind of calibrating, but so far there have been a number of shootings that it’s missed, and hundreds of calls where it said there was a shooting and there wasn’t.”
Nolan said these false alarms are cause for concern. “Operationally, the police are going to become inured to this.” After responding to an excessive amount of false alarms, he said, “police are not going to take the system seriously, and they’re going to take their time getting” to the location of the alleged firearm discharge.
Including June and July, Redwood City’s system goes off an average of 187 times a month said Sergeant Blanc. Not including June and July, the system goes off an average of 95 times per month—43 times for gunshots, 38 times for fireworks and 14 times for other loud, impulsive noises like backfiring motorcycles. The total number of activations in 2009 was 2247, or an average of just over six times per day.
“Often times the only information we receive about a gunshot or fireworks call comes from ShotSpotter,” said Sergeant Blanc, highlighting an argument advocates of the system make: not every shooting is accompanied by a 9-1-1 call.
“Our dispatcher sends us out to investigate possible shootings, where at the same time nobody called 9-1-1” for that incident, said East Palo Alto Police Captain Carl Estelle. “The citizens can get complacent, and just might not call. I’ve had them tell me they’re not going to call for whatever reason.”
However, Estelle said, it is rare that notification from the ShotSpotter system alone has led to any arrests or shooting victims.
In 2007, there were 7 homicides in East Palo Alto, a city that covers 2.6 square miles and has a population of about 33,000 people according to FBI crime statistics. In 2008, the most recent year statistics are available, there were 5 homicides in East Palo Alto.
“It’s just another tool for law enforcement,” said Estelle. “It’s the system with many other factors, such as community participation in cracking down on crime,” that are responsible for the reduction in crime.