Understanding the Edge Computing Framework

To carry out the modern mission, individuals on the front lines don’t just need access to critical information—they need it at their fingertips. This is especially true for members of the military, who are engaged in protecting the nation and whose decisions impact lives in a very real way. 

However, the organizations that carry out the nation’s defense operations face a unique set of obstacles that pose a challenge for communications:

• Service members need to be connected across locations around the world and in what could be high-risk environments.
• A complex, global mission means work can happen in unreliable connectivity environments, and across a multitude of platforms and devices.
• Any lag and latency in communications can directly harm the safety and situational awareness of warfighters on the front lines. 

In the context of those challenges, edge computing will open the door to a new communications model—allowing individuals to process data at the edge device itself. This means right where decisions are made and at the point of data collection. In this new model, there’s no more waiting for critical information from a central command. And no person’s success or safety is compromised by intermittent connectivity. 

But how does it work? In this article, we illustrate our approach to adopting edge computing—and recommend a solution designed to be open, universally-distributed, and seamlessly reconfigured to best serve individuals over time.

We know that the defense mission isn’t one mission. To synchronize across complex needs and services, the power of edge computing can’t be limited to specific systems or devices. Instead of designing a solution that, for example, operates in a vertical stack exclusive to other components, organizations need to address edge computing as something that is open and API-driven to enable integration and interoperability. This not only allows edge computing to impact operations at an enterprise level—it brings this capability within reach more easily than some might imagine. That’s because the hardware and devices you need are the ones you already have.

We make this possible through a framework built around core software that augments existing resources, by way of what we call “sensor processing units” (SPU). Developed to be installed on existing operational platforms, the SPU is hardware-agnostic and enables sensor fusion and processing using the hardware acceleration that’s already there. 

The SPU provides the ability to send data from mission systems to on-the-move and edge devices with little to no connectivity. Once the SPU is within range to the tactical or fixed cloud, it then helps to synchronize and update the central infrastructure, including mission applications and data, with the latest information available regardless of where it was collected.

Suddenly, the frontline vehicle, ship, unmanned aerial system, or device on the warfighter can process and compute data on the ground. However, the SPU is only as good as its network—which is the linchpin that turns software into an edge computing infrastructure.

With a common framework and architecture ready for deployment, edge platforms can start communicating. This is where the beauty of decentralized decision making starts to shine. Whether connected or disconnected, SPUs begin to move messages—sharing, processing, analyzing information—across a newly localized, or mesh, network. This network can transmit lifesaving data no matter the environment, so that our soldiers and field operators can stay focused on making safe, informed decisions, without worrying about connectivity.

Certainly, this local infrastructure fills a communications gap, since these individuals need trusted information to perform their jobs and ensure overmatch. But it’s also important to note that the mesh network, by design, allows its devices to leverage the compute and storage available through a distributed processing architecture. Simply put, it’s built to work hard. And above all, it’s structured to maintain resilience for all users. 

The value of this open, standard framework for edge computing is that it can adjust easily over time as the mission priorities evolve. Rather than specific hardware or devices required to power the edge network, we recommend a platform architecture that stays agile to keep up with individuals who are operating in sensitive, ever-changing environments.

This inherent interoperability opens a new set of possibilities for defense organizations.

When you start to think about the overlap between the U.S. Army, Navy, Air Force, and other military components, there are endless platforms and systems at work. And we are safer and more prepared because of them. But today, can those platforms and systems talk to one another?

Through a common language and architecture, edge processing can happen across and between any and all of them. We now have an opportunity to achieve interoperable communication that runs from the edge network to the tactical or fixed cloud. Information, software updates, and analytics can flow in new ways and reach new corners of our global mission.

At Booz Allen, we call this “smart edge.” Vendor-agnostic. Plug-and-play. Bring your own hardware. Capable of working both offline and on. When that happens, we decrease the data-to-decision chain and maintain situational awareness for warfighters and the entire operation.

Next Up: How Edge Computing Is Being Used by the Military Today