AUSTIN (KXAN) — Gunfire erupted. Screams followed. A woman died.

Innocent bystander Teqnika Marie Moultrie, 30, died after she was shot while walking down a crowded sidewalk on Sixth Street with her wife and friends last summer. In that case, cellphone video captured the sounds of gunfire. Even though police were nearby, they weren’t able to catch the suspected shooter, Endicott McCray until several days later.

On New Year’s, a bullet struck a woman in the leg while she was downtown celebrating.  “I looked down and I could see the blood running down my leg,” said shooting victim Victoria Rocha. She went to the hospital where she learned her injury was from a .45 caliber bullet.

A HALO camera was on that corner. Despite the cameras and a street full of revelers, Rocha says police have not found her shooter.

While taking a closer look at the HALO system, KXAN found a May 2016 document mentioning APD’s current systems could potentially integrate gunshot detection technology. So, we started looking into what the technology does and if it could help police in cases like Rocha’s.

Gunshot Detection

In some cities, it takes just seconds after a shooting for police to pinpoint the location of the gunfire. Crews place sensors—basically microphones—across an area, typically on rooftops or poles about 30 feet in the air. The gear can triangulate the sound. Trained staff listen to the noise to try to filter out noises such as fireworks and cars backfiring. Then, police get a location and can head to the scene.

About 90 cities across the country are using technology from the company ShotSpotter. Some cities have tried out other systems such as SENTRI, which pairs surveillance cameras to turn the lens in the direction of shots and, possibly, catch a shooter on camera. However, SENTRI has not reached the wide-spread adoption of ShotSpotter, based on information from the company’s websites.

SureSpotter software (via SureSpotter)

San Antonio is the only city in Texas using ShotSpotter. The city council is still evaluating the effectiveness of ShotSpotter and if it will continue to use the system. Data from the San Antonio Police Department show since the city implemented the technology on April 30, 2016 through Oct. 31, 2016, ShotSpotter detected 1,428 gunshots in 359 shootings. Police made 11 arrests related to the alerts. However, police cannot say the arrests are solely because of the technology.

“I grew up in these neighborhoods. [ShotSpotter] covers the house that I lived my first 10 years in,” said Alan Warrick, II, San Antonio city council member.

Warrick is a supporter of the technology and says he traveled to Washington D.C. to see how it worked before the system came to San Antonio. In San Antonio, the system covers roughly two square miles. Information from ShotSpotter shows it can cost $65,000 to $95,000 per square mile each year.

“We have specific ShotSpotter officers that only address these gunshots in the areas that they occur on the east side of San Antonio, in our area. We have ShotSpotter response times as low as 3 minutes 30 seconds, on average,” said Warrick. “The citywide emergency response time is about 9 minutes. So, you’re cutting it almost to a third of what the response time is.”

The city council in San Antonio and other cities will have to decide whether those kinds of results are the best use of taxpayer dollars.

Remaining Questions

Researchers are still evaluating the systems’ effectiveness. The company claims the ShotSpotter, when partnered with practices from police, can reduce gunfire by up to 80 percent and see a drop in related violent crime and homicides by up to 40 percent. The company has also found that fewer than 1 in 5 shootings are called into 911.

A 2011 study commissioned by ShotSpotter and endorsed by the National Organization of Black Law Enforcement Executives found the system helps police respond more quickly to gunfire and officers trust ShotSpotter over 911 calls. However, it also found the most common complaint is that the system can produce false positives–an alert for something that was not a gunshot. Dispatchers surveyed for the study believed that about 67 percent of activations were due to gunfire, while the rest were attributable to fireworks, construction, cars or other noises. The report also notes that the lack of an alert when there is a confirmed gunshot is a rarity.

Since that study in 2011, ShotSpotter has changed to a subscription-based service where the company’s employees analyze the recordings.

“There is some research out there. Right now, it is still not as well evaluated as it may have should have been in the past,” said Dr. Daniel Lawrence, a research criminologist at RTI International, who is working with the Urban Institute on a study of gunfire detection technology funded by the National Institute of Justice.

Lawrence says researchers will explore a number of topics including the costs of the system. So far, he’s noticed police respond to ShotSpotter calls more quickly than 911 calls. However, the research has not reached that phase.

“It’s not an end-all for this type of crime,” said Lawrence. “It’s just one tool of many that police officers use to address this crime.”

Critics of the system also have privacy concerns about ShotSpotter. They worry the system can record voices.

The company told the American Civil Liberties Union that audio from its microphones gets stored locally for hours or days. But it’s only the gunshot and a few seconds before or after that gets downloaded and sent to police. The ACLU notes that voices have been used in court cases, but those voices were recorded in those short clips alongside the gunshots.

Would It Work in Austin?

An Austin Police Department spokesperson says the department does not believe there is enough gunfire in the city to warrant the costs of the system.

“My gut reaction is, that it’s just not worth it,” said John Hunt, a retired Austin police detective. “It’s a fine tool to have, it’s a great idea, but I’d rather see all the cops with bullet proof vests, more things that protect the officers.”

Hunt believes people in Austin will typically call in gunshots. He questions how much ShotSpotter would help investigators, especially if there are witnesses or if police have already pinpointed where someone was shot.

The crime stats alone show a difference between Austin and San Antonio. San Antonio ended 2016 with 149 homicides. Austin had 39.

Still, Rocha hopes technology like this could help police in finding suspects, like the one who shot her.

“It could prevent it from happening again,” said Rocha. “It could prevent someone else being hurt, dead, and not being able to talk about it.”