Achieving Decision Dominance by Empowering the Tactical Edge: Sponsored Content

“Decision dominance … is the ability for a commander to sense, understand, decide, act and assess faster and more effectively than any adversary,” Army Futures Command Commander Gen. John “Mike” Murray, USA, told the Association of the U.S. Army Global Force Next virtual conference in March.

In modern warfare, against near-peer adversaries, victory will no longer be guaranteed by strength of arms alone. Speed and accuracy of decision making will be more critical than ever, and in many circumstances, decisive.

“Decision dominance has three parts,” explains Chris Yates, Red Hat’s DoD strategy solutions architect. “It means first and foremost making the best decision, objectively—like in chess where there’s always a ‘best move’—but that means you have to have all the information.”

But in a chaotic and fast-moving combat situation, he adds, “decision dominance has a second aspect: Making the best decisions you can with the information you actually have available at the time.”

“Decision dominance requires that the right people and the right systems have access to as much of the right data as they can get at the right time,” says Yates.

“Having at our fingertips the information that we need to be able to make decisions is vitally important, because the faster that data is available, the faster we can make those decisions, and the greater the advantage we can gain … Decision dominance is thirdly about the speed of decision-making.”

Yates uses the analogy of a chess game. “I move, then my opponent moves. But if I can halve the length of my OODA [observe, orient, decide, act] loop, I can make two moves for every move my opponent makes, and I can’t lose.”

Achieving decision dominance like that in a complex, multidomain conflict means being able to access and analyze multiple data sources and streams, from multiple domains, at machine speed, to foreshorten the OODA loop. And it means being able to do so wherever needed, not just back in the rear echelon, but out with the forward deployed forces—at the point of the spear.

Yates says there are several key capabilities the Army needs to make decision dominance possible:

• Data locality—Pre-positioning the right data so the right people and systems can access it when the right time comes.
• Automation—The ability to swiftly and automatically deploy software in a standardized, repeatable way.
• Edge cloud capabilities—AI and the data it requires needs to live at the tactical edge, and that means having cloud capabilities there.

Data locality: Pre-positioning what the warfighter needs

“As a warfighter,” adds Joel Stein, business development manager for Red Hat, “I don’t need access to all the data—too much irrelevant data will actually slow me down. What I really need is the ability to have the relevant data, the data I need for my mission, as close to me as possible in the cloud environment to reduce the latency and risk.”

And that data has to be available on the move, adds Stein. “Future conflicts will be fought on the move. We know peer adversaries will exploit our use of the electromagnetic spectrum [or EMS]. We emanate, they shoot ... That means you have to keep moving. If you stop, you die.”

Stein, who is a Signal Corps officer in the Army National Guard, recalled how, for counter-insurgency operations in Afghanistan and Iraq, the Army required access to the biometric data of suspected terrorists and insurgents. “We needed to know who is coming across the border, who are these guys we apprehended. … Because we were wasting our time detaining sheep herders and we were losing hearts and minds as a result—lack of data was causing us to fail in the mission.”

Warfighters on the ground didn’t need access to all of the data in the DoD global biometric repository, points out Stein. They only needed access to the biometrics of suspects and targets in that region—only the data that was relevant to their mission.

The solution was to “take the biometric data repositories of people who were of interest in that region, and figure out how to containerize that data and pre-place it in the battlespace,” so it was immediately available to forces on the ground.

Extolling the virtues of pre-positioning for data and other materiel, Gen. Murray has joked that “The best way to get from point A to point B—is to already be at point B.”

“That’s part of the value Red Hat provides,” says Stein: “The ability to locally store that data in a hybrid infrastructure environment, and make it accessible through pre-placed containerization, so the data and the applications you’re going to require in the environments you’re going to are pre-positioned downrange.”

That preplacement is vital, Stein adds, because of the long and fragile communication links between the frontline and the rear echelon.

Even at the speed of light, delays add up, he explains. Every satellite hop the data has to make to orbit and back means a 700 millisecond latency, long enough to make some software packages using it glitchy. And it relies on a channel which will be under attack in a conflict with a peer adversary.

“A peer competitor will be trying to jam those channels, and even if we succeed in keeping them open, bandwidth will be limited,” says Yates. “We can’t afford the delays and risks involved in collecting all the data, important and trivial, piping it all back to the continental U.S., processing it there and then pushing the decision back out to the edge.”

“We need the ability to triage and process data and make decisions based on it at the edge,” says Yates. “Red Hat’s open hybrid cloud provides the software infrastructure that makes that possible. … It’s the plumbing behind the scenes.”

Vendor agnostic automation: A scalable, flexible infrastructure

When it comes to automation, open standards-based architectures, like those in Red Hat’s Ansible Automation Platform, allow generalizable automation, which is vendor agnostic, explains Yates. “You may have proprietary automation for one [network equipment] vendor, but you are absolutely for sure going to be operating in a vendor-heterogeneous environment. ... There are no more monocultures. You need a solution that can deal with that.

“Ansible allows you to automate anything that you can manage with mandates, so you can automate many things with this same automation. And it can leverage proprietary [network equipment] vendor automation, so even if you already made an investment in that, it can still be integrated through the general automation environment,” Yates concludes.

Data at the tactical edge: The lifeblood of artificial intelligence

Data at the tactical edge is also the key enabler for artificial intelligence, or AI. Data is used to train AI, and once the algorithms are deployed, data is the raw material on which they work.
“Data localization plus artificial intelligence equals decision advantage at the edge,” is the way Yates puts it.

Mike Zizza, Red Hat’s DoD strategic accounts manager, traces this thinking to the military decision-making process (MDMP.) “Data at the edge will come from multiple sources—intelligence assessments, signal intercepts, satellite imagery, weather, open source, maps etc. It all must be accessed and arranged quickly to really enable ‘edge MDMP.’ So distilling actionable and consumable information for a brigade staff, even down to platoon level, will mean dealing with constant streams of information necessitating continuous IPB [intelligence preparation of the battlefield]  and therefore continuous, living, COA [course of action] analysis,” says Zizza.

At a recent Technical Exchange Meeting for the Army’s Capability Set 25—the third and most advanced stage of the service’s ongoing network modernization program—the Army announced it was seeking AI capabilities to help with logistics assessments.

Currently, according to a draft procurement document, sustainment staff do much of their work “manually or with basic office applications that require time-intensive preparation and data entry,” including re-entering data from one system into another. “The expected growth in the scale and complexity of the technology footprint in all warfighting functional domains will only serve to increase the difficulty of information management tasks without automation.”

By automating these processes, Zizza says, better information will be available more quickly to the right decision makers, enabling decision advantage, and eventually decision dominance.

A proven solution

Red Hat software infrastructure provides edge cloud capabilities flexible enough to achieve decision dominance in multidomain operations, where units may be required to shift mission at a moment’s notice. “Pre-positioning data and apps is all very well,” says Stein, “but if the mission changes, it may require access to different data. I need repeatable, automatable processes that can place that downrange where the warfighter is.”

Red Hat’s OpenShift is a proven solution for just that kind of scalable, flexible, deployable infrastructure. It provides the foundation for Verizon’s 5G core, helping to enable edge computing capability for the telco.

“Think about the Super Bowl, or any event where you suddenly have hundreds of thousands of additional customers needing bandwidth, needing data, using apps,” says Yates. “The way Verizon handles that is they have ‘burst capability.’ They bring in the mass-produced and affordable 5G hardware components, and they use automation around that to be able to rapidly and repeatably deploy configurable systems. … They use content delivery networks to pre-position data and computing power nearer to the customer.”

Those are exactly the kind of flexible, scalable capabilities that the Army will need to deploy to ensure decision dominance in complex multidomain operations, notes Yates: “The same kind of burst capability is needed on the modern battlefield. That’s what Red Hat delivers.”

For more information, visit redhat.com/dod.