Federal agencies are increasingly adopting edge computing, citing its transformative power to enable field agents and military personnel to make faster decisions and respond to crisis situations, experts said.
Edge computing improves agility and enhances field operations by placing compute and tooling closer to decision-makers in the field, government and industry experts said during a recent webinar.
The Department of Defense (DoD) is using computing at the edge in areas such as object detection, while homeland security and law enforcement agencies are employing the technology in cybersecurity investigations and natural disaster response.
William Streilein, chief technology officer for DoD’s Chief Digital and AI Office, said edge computing use is growing among DoD agencies in part because of its ability to help agencies achieve decision advantage.
Decision advantage requires “democratizing access to the data that supports decisions … and that means at the edge,” Streilein said. “You can’t spend the time to move the data back to the cloud and leverage the computing there. So you really need to figure out how to have those decisions be made in an edge environment.”
Also driving the move to the edge, Streilein added, are technical advances such as in miniaturization of devices and the growth of data-collecting edge sensors. “We’re now in a position to actually do more at the edge than we were previously,” he said.
Garfield Jones, associate chief of strategic technology at the Cybersecurity and Infrastructure Security Agency, agreed that the melding of edge computing and AI is helping agencies see the possibilities of computing at the edge.
“Internet of Things devices (at the edge) generate a fair amount of data,” he said. “That’s where we’re looking at artificial intelligence (AI) really coming into play and being able to help filter that distributed data … and make sure we are getting relevant data.”
John Dvorak, chief technology officer for North America Public Sector at Red Hat, said edge computing has groundbreaking potential to help Federal leaders achieve their missions.
But he pointed to a key challenge for edge systems: data gravity, which makes today’s massive streams of fast-flowing data harder to move over time.
Dvorak used an interstellar analogy to break down the concept. “Think about data as a big planet like Jupiter. It has this strong gravitational pull that attracts other objects to it,” he said. “Data sets, as they grow larger, exert this strong gravitational pull on applications, services, and other data.”
That virtual pull is “the challenge with data gravity,” Dvorak continued. As edge computing users analyze data collected by sensors and other edge devices, he said, “you are then creating more data, which … continues to increase the amount of force.”
Dvorak said agencies are addressing the challenge by modulating how much data their systems store at the edge, using techniques such as containerization that translates applications into lightweight, portable units.
“You really have to think ahead when you are working in edge environments,” he added.
At the Department of Homeland Security (DHS), edge deployments remain a learning process, said Scott Simpson, the department’s digital transformation lead and innovation coach.
He said DHS agencies evaluate and purchase edge and related AI technology with a “show me, don’t tell me” approach. “You have to do it differently than when you’re just purchasing a commodity,” he said. “Otherwise, you’re not going to get the right solution to meet your mission.”
Despite the challenges, agencies are forging ahead with edge computing use cases, often integrating the technology with AI, speakers said.
According to Streilein, DoD is using edge computing for “object detection via computer vision” in Project Maven, which was established in 2017 to speed the integration of big data and machine learning technologies.
Defense officials are also utilizing edge computing combined with AI, he said, to leverage data to predict when maintenance will be needed for “our materiel that we use to provide national security … it’s a really great application.”
In the law enforcement and national security arena, Dvorak said agencies are sending “investigative toolkits” to the edge to help cybersecurity response teams.
The kits are “gathering data from the networks. They’re collecting digital evidence,” he said. Typically, he added, the kits – which have a finite amount of storage – need to be sent back to data centers to be wiped and then sent back to the edge, a process that can take several weeks.
But with cyberattacks growing in severity, investigators frequently want to keep the kits on the edge for additional evidence sifting. “We’re seeing more and more of these cases where teams are saying. ‘… this is more than just a basic cyber hack. This is a lot worse. I need more tools. I need something on site,’” Dvorak said.
Watch the full webinar on demand with William Streilein, chief technology officer for the Department of Defense’s Chief Digital and AI Office; Garfield Jones, associate chief of strategic technology at the Cybersecurity and Infrastructure Security Agency, and Scott Simpson, the Department of Homeland Security’s digital transformation lead and innovation coach.