How Call Centers Are Using AI and Cloud-Based Tech in 911 Emergencies

Estimated read time 6 min read
  • Emergency-response centers are using AI to improve and streamline services like live translation.
  • Better tech could help workers prioritize calls, avoid logistical issues, and catch crucial details.
  • This article is part of “Build IT,” a series about digital tech and innovation trends that are disrupting industries.

Thanks for signing up!

Access your favorite topics in a personalized feed while you’re on the go.

download the app

Bull

Raquel Lewandowski has been working as a call taker at a call center in Delaware County, Pennsylvania, for nearly 30 years.

In that time, she’s seen technology modernize emergency-response services to better serve people who need assistance. She recalled one instance where she used an app to direct a woman who was lost at the 1,000-acre John Heinz National Wildlife Refuge at Tinicum.

“Lights are going out, and she doesn’t have that much left on her phone, and she’s scared, and she doesn’t know how to get out of here,” Lewandowski said. “I was like, ‘Hey, you know what? I can send you a link. Open it up.'”

With the click of a link, Lewandowski could see where the woman was through the camera on her phone. “I can see the Philadelphia skyline — I can see that in the background, so I can kind of tell her which direction to start walking,” she said.

Lewandowski was using Prepared Assist, a platform launched by the emergency-response technology company Prepared. The software uses location data along with text and video capabilities to help with 911 calls.

It also uses artificial-intelligence tools to provide real-time translations and speech processing to dispatchers in high-stress situations.

Michael Chime, Prepared’s CEO, said he launched the company in 2018 in response to several high-profile mass shootings that prompted a new interest in 911 technology. He questioned why smartphones had better features to help in emergencies — like location tracking and video calling — than many call centers did. “Why was I, an everyday citizen, better equipped from a data perspective to communicate to my friend in any odd moment than I would be to 911?” he said.

That was 18 years after the US and Canada started planning the Next Generation 911 initiative, meant to ensure that technology in public-safety answering points, also known as PSAPS or call centers, remained updated.

While many people in the industry have acknowledged a need for better emergency tech, its implementation has varied across states. “There’s a lot of work that’s needed to move our nation to the 21st century,” said Brian Fontes, the CEO of the National Emergency Number Association.

One of the most widespread innovations so far is text-to-911. Now artificial intelligence is introducing more possibilities in 911 operations.

The AI tools solving a big emergency-response problem

Alex Dizengof, the cofounder and chief technology officer of the emergency-communication platform Carbyne, said a big challenge in 911 operations is language.

Operators taking a call from someone speaking a different language typically have to figure out on their own which language it is and then reach out to a third-party live-translation service.

“You’re losing people because they don’t understand what’s going on,” Lewandowski said.

Prepared and Carbyne are both tackling this problem. Prepared’s live-translation feature is designed to identify a caller’s language and translate for the call taker. The feature can be added to existing call-taking infrastructure or location-service software like RapidSOS.

Carbyne’s system uses AI and natural-language processing engines to automatically translate a call. The company says the software can shorten a call by nearly five minutes.

Alleviating stress in understaffed call centers

AI-powered systems could help reduce call takers’ workload. In a 2023 survey from Carbyne and the National Emergency Number Association, 82% of respondents said their call centers were understaffed. Additionally, 74% of those surveyed suggested that they or their coworkers felt burned out.

Prepared’s transcription service is designed to assign keywords to trigger certain dispatches or alert management, helping call takers catch details they might miss over the phone.

Anthony Mignogna, the chief of communications for Delaware County Emergency Services, recalled using the transcription service while taking a call from someone in danger. “There was a customer dispute at a gas station service shop, and he whispered, ‘The guy pointed a gun in my head,'” Mignogna told BI. It was hard to hear the caller’s warning because of the background noise, Mignogna added, but “through the transcription, we caught that.”

Carbyne also recently rolled out an AI-powered triage system designed to help centers prioritize calls during high-volume periods or nonemergency situations. Heather Hilliard, the deputy executive director of the Orleans Parish Communication District in Louisiana, said that over six weeks, they used Carbyne’s triage system 48 times on about 200 calls.

Barriers to modernizing emergency tech

There are some concerns about using AI in the 911 workflow. Emergencies need to be handled with great care, as they’re often highly sensitive and complex. AI is still a fledgling technology susceptible to making mistakes and reproducing bias from the data it’s trained on.

“Bias is definitely a concern that we’re always thinking about, and also the accuracy of the data,” Dizengof said. Carbyne says it checks a selection of translations with a person to verify its system’s accuracy. The company says it’s also developing a language-processing engine trained on 911 calls to provide results that are more precise.

Modernizing emergency tech would also require updated PSAP infrastructure, which isn’t cheap. PSAPs are run at the state and local levels, and individual municipalities often decide whether to migrate to a new system.

The Emergency Communications Center in Lincoln, Nebraska, for example, recently started using Prepared Live, a cloud-based platform the city said would cost about $50,000 a year. The goal is to reduce the costs of faulty legacy technology and save money on hardware in the long run.

“I can’t tell you how many times I had problems in my server room,” said Karima Holmes, a former 911 call taker who now serves as Carbyne’s vice president of public safety. “I had technology go down because there was moisture in the air.” She added that with cloud-based systems, that’s not an issue.

The future of 911 technology also relies on systems innovating past verbal-only communication. AI-powered transcription services, location-data gathering, and expanded video capabilities could help give responders a better picture of a situation before help arrives.

But Fontes said progress can be incremental when many centers lack the resources and budget to upgrade their systems. He called on the federal government to step in to help centers access new services. “Congress has to fund this now to enable all the benefits of technology that the private sector has out there,” he said.