MCP was out in force again yesterday at this year’s National Emergency Number Association (NENA) conference. Here’s a snippet of what we learned:
Amazon Web Services talked about how it is trying to discern what technologies currently used in its call centers might be applied to the public-safety sector, particularly the 911 community. The company is exploring whether AI could “listen” to a 911 call and determine the stress levels of telecommunicators by analyzing tone, cadence, breathing, and other voice attributes. If the AI solution decides that a telecommunicator is on the verge of a stress-related breakdown, an alert automatically would be sent via text, email, or the computer-aided dispatch (CAD) system to a supervisor.
Though it would need a considerable amount of vetting, Kate Elkins, emergency medical services/911 specialist for the National 911 Program, said in an interview that the idea is worth exploring. “Anybody who’s trying to leverage technology to help the mental health of their people, we need to partner (with them), do the research, gather the data, and see if what we’re doing is helpful and confirm that it’s not harmful.”
Elkins, who also is a doctoral student at Johns Hopkins University and a Bloomberg American Health Initiative fellow, added that the 911 community has a long list of issues that need to be resolved – including CAD system interoperability and the staffing shortage that plagues 911 centers from coast to coast – but the mental health and well-being of telecommunicators tops the list.
“As we improve the (telecommunicator) experience, as we improve prevention, intervention, and the mental health and well-being of our people, we’ll improve recruitment and retention,” Elkins said.
She added that anything related to mental health and well-being is a complex problem that needs a public-health approach to resolve.
“If we better define the problem with data, to investigate how best to prevent the consequences we want to prevent, but also as we professionalize, as we improve access to resources and education, and opportunities for evidence-based intervention, to protect our people, to treat our people, and to improve their resiliency, I think that’s going to be the important thing,” Elkins said.
However, in an educational session, one speaker cautioned that the public-safety sector needs to proceed with great caution regarding relying on AI to support emergency response. While AI seems to show promise, it has to be proven that the technology will stand up to the rigors of the mission-critical environment in which public safety agencies operate, because if something goes wrong and AI is the culprit, lives could be lost, and the affected agency could face serious liability exposure. He urged that agencies avoid being AI “beta testers,” and that the technology be thoroughly vetted before it is used to support emergency response. This process could take five years or more – and never may be realized.
Often, telecommunicators are promoted because they have long tenure at an emergency communications center (ECC) or have performed their role admirably. Just as often, they don’t receive the training needed to perform their new duties effectively – consequently, they flounder. A corollary aspect is that telecommunicators and supervisors typically don’t receive leadership training – and they should because leadership is difficult to achieve. It’s also essential to moving the 911 community forward. On a related note, ECC officials need to spend more time and effort determining whether a candidate for promotion has the attributes needed to perform effectively in a leadership role, which at its essence, requires the ability to think strategically.
There was a time, not long ago, when public-safety agencies didn’t seem to have cybersecurity on their radar screen, perhaps believing that cyberattackers would target the private sector, where large companies have the financial wherewithal to pay hefty ransoms to decrypt their files. But over the last 24 months, numerous high-profile incidents – the attack on the city of Dallas in May being the most recent – have gotten the attention of public-safety agencies.
The Dallas incident involved a ransomware attack that affected numerous city servers and caused several noteworthy service interruptions, including knocking the police and fire department websites offline. The attack also affected the police department’s computer-aided dispatch (CAD) system, but a backup system quickly was turned up, so emergency response was unaffected, according to news reports.
Consequently, there has been a lot of dialogue at NENA concerning cybersecurity. The consensus thinking is that the best thing that any public-safety agency can do, including emergency communications centers (ECCs), is to develop a strategic plan that covers prevention, mitigation, and communication with the media, stakeholders, and government officials if a cyberattack occurs. The plan should provide for conducting a baseline network assessment to identify potential vulnerabilities; central to this is crafting a network diagram that identifies all components, especially endpoints, because they essentially are breach points.
It was noted that the ECC’s internal information technology (IT) and cybersecurity teams should not conduct this assessment. Rather, it should be conducted by a third party that is not only well-versed in cybersecurity but also understands public safety’s mission-critical environment. This is something that the National Institute of Science and Technology (NIST) cybersecurity framework recommends.
One educational session featured a panel discussion that explored simple things that public-safety agencies can do to avoid a cyberattack. The place to start is to assume that, sooner or later, a cyberattack will occur. Then, create a list of experts across multiple disciplines that can be called upon to assist in the recovery effort. This is especially true in the case of a ransomware attack. In that case, the agency will need specialized help fast – time will be of the essence.
It also was noted that agencies often don’t log when and where cyberattack attacks are attempted, and even when they do, they don’t keep the logs for very long. According to the experts, this is a mistake because the logs will provide insights regarding vulnerabilities that might be exploited in future attempts. Another tip was to ensure that the center can go back to the basics if an attack occurs, e.g., if an attack cripples the computer-aided dispatch (CAD) system, then pull out the run cards that were used before CAD was implemented – and if those old run cards no longer exist, recreate them. It is essential to be able to operate the ECC if all of its technology is compromised for a significant amount of time.
Five years ago, RTT was introduced as a better way for the hearing-and-speech-impaired community to access the 911 system. It was a fantastic idea that proved more difficult to deliver than anyone anticipated. However, during the NENA conference, the North Central Texas 911 District (NCT911) provided details on the nation’s first end-to-end RTT deployment, which went live in August 2022.
NCT911 worked closely with its call-handling equipment (CHE) vendor (Solacom), wireless service provider (T-Mobile), and emergency services Internet Protocol network (ESInet) vendor (Comtech). T-Mobile needed to adjust its network to transmit calls via the Session Initiation Protocol (SIP); Comtech had to adjust its ESInet to route RTT calls to NCT911’s emergency communications center (ECC); and Solacom had to program its CHE to accept RTT calls.
All of this took much longer than expected. Initially, the idea was that RTT calls would be similar to Short Message Service (SMS) calls, i.e., text calls, whereby the recipient can see the entire message. However, RTT requires the telecommunicator to see the message as it is being typed, keystroke by keystroke. This is important because it enables faster communication, reducing response times. Further, the telecommunicator, in many cases, can anticipate where the conversation is going, and even if the transmission is stopped, armed with that information plus the caller’s X, and Y coordinates, the telecommunicator can dispatch a response in most cases.
During the conference, MCP subject-matter experts met with a next-generation core services (NGCS) provider that shared its experience with a state 911 agency. The assumption is that location-based routing will result in a 911 call arriving at the correct emergency communications center (ECC) faster and more often. After examining its call-transfer data from last year to this year, the agency found that it has cut the number of call transfers nearly in half. This saves as much as 60 seconds per call, which is fantastic given the Federal Communications Commission (FCC)’s estimate that 10,000 lives can be saved annually if the 911 system could dispatch emergency response just one minute faster.
The state 911 agency that was cited was generating 13 hours of delay each year due to the thousands of call transfers it was executing – all of which were eliminated after the agency implemented location-based routing. Cutting call transfers by nearly 50 percent means fewer calls dropped, less confusion experienced by emergency callers, fewer errors, and fewer instances of callers needing to repeat themselves – in other words, better service, faster response, and more lives saved. It’s a great outcome.