Concept of near future when robot dogs are used to assist police forces on assignments. The robot dog in the image has a camera mounted on top of the body.

Concept of near future when robot dogs are used to assist police forces on assignments. The robot dog in the image has a camera mounted on top of the body. mikkelwilliam / GETTY

What 5G Will Actually Do for the U.S. Military

Lots of sensors, lots of data, open up the possibilities of energy conservation, better security, robot teaming, and much more, military experiments show.

Imagine a city where self-driving electric cars anticipate when and where their passengers will need a lift long before the vehicle is called. Batteries to shore up energy shortfalls arrive before they are needed. City managers know the exact placement and condition of every city asset, and theft is immediately detected. There is no traffic, because every vehicle is in constant communication with every other car and traffic light. The residents know there will never be any delay in the services they need because the city around them functions as a perfect, seamless organism just beyond their perception. 

That internet-of-things enabled city of the future looks a lot like the experiments with ubiquitous 5G cellular networking the military is doing today. 

Marine Corps Lt. Col. Brandon Newell, the director of the Navy’s SoCal Tech Bridge, leads what he calls a 5G living lab. His experimentation, culminating in a demonstration later this month, looked at how 5G cellular connectivity across a base would unlock new uses for self-driving vehicles, greater energy efficiency on base, and even better teaming between drones and ground robots. 

It turns out that a stronger cellular connection enables a lot more than video conferencing and reliable internet. 

Newell spoke to Defense One as part of a special interview ahead of the 6th annual Defense One Tech Summit, taking place from June 21 to 25.  

In 2017, he came to California to look at how innovative technology would change life on the battlefield, and he found that nearby tech companies, such as chip-maker Qualcomm, were far ahead of the military on building connectivity into devices. Newell invited them to use the proving grounds at Marine Corps Air Station Miramar to demonstrate and test new commercial capabilities, like hardware and software for self-driving vehicles. In return, Newell got market forecast data from one of the biggest players in the emerging market. 

“We were able to learn what that self-driving unmanned future looked like, but also the key enablers to it,” he said.

Turns out that a robust cellular signal makes self-driving cars a lot more useful. So the following year, Newell worked with Verizon to build a 5G “living lab." They looked at how better cellular connectivity would enable better on-base security; connected, self-driving vehicles; drones; and energy use. 

A robust cellular signal makes all of those things more useful by interlocking them in a communication network, he said.

More cellular connectivity enabled more sensors and cameras to detect intrusions. In March, Newell and his team held a demonstration with Anduril Industries to show how more sensors made it easier to police more areas against a wider variety of threats.

“What you are seeing is that sensor suites with artificial intelligence and machine learning aggregate the data at the software level, at the user interface level," he said. "You can actually expand to multiple threat vectors. So we’re showing ground perimeter, [and] drones. We’re showing maritime counter-intrusion. We were able to show how a single police officer can have that kind of local and regional look at the threat.”  

More sensors connected to more robust cellular service also helped drones communicate. 

“We started to see the future of drones would be cellular, and not just radio frequency,” Newell said. That enabled much better teaming with on-the-ground robotics. “We had two drones and a ground vehicle. We had system-to-system communication through this network as we projected networks onto the battlefield for the future.” 

One of the big areas of interest and research was self-driving cars for resupply. Here, too, connectivity increased the realm of the possible. 

“We set up an autonomous vehicle proving grounds not just to see what self-driving cars could do, but what are the concepts of employment across the Department of Defense that they are well suited for,” he said.

They looked first at logistics as part of a two-year program. 

“The key here: don’t just do a self-driving vehicle the way you did a manned vehicle”: in large convoys requiring manned gun trucks at the front and back, he said. That form of resupply, which the U.S. military used heavily in Iraq, made convoys highly vulnerable to IEDs and other threats. Instead, said Newell, “Think FedEx: How they are going to roll into your neighborhood in the future with an unmanned truck and out of that truck, drones will fly off and deliver individual packages to your door.” 

“What does that look like on the battlefield for us? What we can see is we are going to start disaggregating logistics, making it more dynamic in how we get logistics to the point of need. It’s all built around [self-driving] electric platforms, which was a key learning for us. We’ve found that connectivity and electric is the unlock to this battlefield of the future.” 

On June 23, Marine Corps Air Station Miramar will hold a demonstration to show how that connectivity enables a new approach to logistics and resupply. 

More interconnected sensors sending data wouldn’t seem to be a good way to cut down on energy consumption, but it is, Newell said, in large part because electric vehicles are much better suited for self-driving missions than mechanical ones, which must be retrofitted to accommodate fly-by-wire control and other capabilities. With an electric vehicle that has lots of data streaming off of sensors, it’s possible to predict, from those large data flows, which locations might need energy in the near future. But  it also has a self-driving battery to meet that need. 

“This becomes exportable energy. What if we can export off of that electric vehicle energy to the fire team, to the squad, energy that is in a usable form?” he said. 

Of course, the military has a long way to go to embrace cellular communication in battlefield contexts, preferring radio and satellite communications using hardware it can better control. 

The military is missing a key opportunity and rapidly falling behind the state of the art, Newell said. “Those traditional waveforms and communications types that we use in the military, who’s funding them? It’s largely us,” meaning the government, he said. “Compare that to my cell phone. There’s been billions of R&D [funds] that’s been invested...into the functionality of cellular…What we watched is that as the 4G [long-term evolution] started to unlock new capabilities for the private sector...we in the government stiff-armed it. We said, ‘We don’t do cellular on our missions, not even on our bases.’” 

Sachin Shetty, the executive director of the Center for Secure & Intelligent Critical Systems at Old Dominion University's Virginia Modeling, Analysis, and Simulation Center, has been creating a 5G warehouse for the military at the Marine Corps logistics base in Albany, Georgia. 

He employs a zero-trust security model, he said, which ensures continuous authentication and validation of all the items on the network. It is possible, he noted, to use end-to-end encryption across the network, but that can slow down performance. 

His work showed how sensors distributed across a warehouse allowed for much better management of all the items inside that warehouse. The data also allowed managers to much better predict when items would be used, allowing robots to preposition for that moment, thus saving energy.

In a conversation Thursday, Shetty said the same data that allows for predictive inventory control would also enable better security. A lot of data can allow managers to much more quickly and easily pinpoint when something isn’t working as it’s supposed to. Anomaly detection at scale is the basic principle that many common cybersecurity software and practices are built on. 

“We are using a data-driven approach for security,” he said. For instance, if something in the warehouse changes the way it behaves, ubiquitous sensing allows for much faster identification of that fact. 

The same goes for any item that is part of a 5G network. “We go from technical analysis to behavior analysis,” he said.