As of now, when the Navy’s developers are building new software designed to run aboard ships, the only way to really know it will work the way it’s supposed to is to board the vessel and test it on real-world systems.
But that’s beginning to change. The Navy has started to build virtual replicas of the systems that comprise the IT networks aboard its warships, starting with the ships in the U.S.S. Theodore Roosevelt strike group. The basic idea is to let developers build and test new ideas on those “digital twins,” letting them get new technology out to the fleet more quickly, and without the risk of breaking mission-critical systems.
“We can catch all the problems before they ever hit a piece of hardware,” said Robert Parker, the technical director in the Navy’s program executive office for command, control, communications, computers and intelligence. “We’re trying to push testing and development as far to the left as possible. We want to build a model from day one that encapsulates all the interfaces, all the data exchanges, all the messages. So before you write a single line of code, you can check to see if you have the applications laid out correctly.”
But the potential benefits go well beyond developing software more efficiently.
The Navy wants its digital twins to operate in a commercial cloud environment. Under those circumstances, the universe of people who can write software that works on afloat systems could expand dramatically, including to deckplate-level sailors who are looking to solve a particular problem they’ve encountered – assuming they have some coding skills.
“One of the benefits of cloud is the ease of access – it’s ubiquitous,” said Delores Washburn, the chief engineer at Naval Information Warfare Center Pacific. “We’ve been able to architect it in a secure way where we can let our contractors, the performers, and even the government workforce, all have access to the same development environment. We have folks all across the United States accessing it to do development now.”
But getting to a point where that’s possible is harder than it sounds. Most of the modern world is accustomed to operating in an environment where computing hardware is swapped out every couple of years, and software updates happen with the tap of a button. That’s not how the Navy’s applications have historically been architected.
They’ve tended to be highly-siloed, with each monolithic system bringing its own data stores along for the ride on a naval vessel. To make them more agile, the Navy’s digital modernization project has also had to think through ways to make its systems more containerized.
To make it easier to solve mission-or-ship-specific problems, that’s mostly meant moving toward a platform-as-a-service model, where developers working on the Roosevelt Strike Group’s digital twin are dealing with the same IT environment that’s operating aboard the strike group’s vessels.
“When we go have a big ship maintenance availability and we do our software installs, we’re not done,” Washburn said. “If there’s some new game-changing technology, we can now push that update out there to the ships in the strike group just as seamlessly as you would on your smartphone, because those building blocks are all there.”
One of those key building blocks is the Consolidated Afloat Networks and Enterprise Services (CANES) program the Navy started a decade ago, which aimed to move all of its ships toward a common IT baseline. But even with CANES, every ship is different – mostly because of local software settings that have been changed over time.
With digital twins, Parker said shore-side developers can now start to work on an accurate model of a particular ship’s real-world network at a particular point in time.
“We have a model of what the as-built configuration for the ship should be – PC settings, router settings, switch settings – and as we pull the real-world settings down, we can immediately compare the two and determine where they went out of compliance,” he said. “I don’t need a perfect model, I just need to figure out where the pain points are – what are the things that are going to prevent applications from being able to talk to one another.”
Even as the Navy continues to refine the concept for the Roosevelt strike group, the eventual ambition is to create a digital twin for each of the nearly 300 vessels in its deployable battle force.
“That’s the power of the cloud,” Washburn said. “It’s so easy to spin up a copy of a ship, and we should be able to replicate any ship out there in the fleet. We’d already done a digital twin of the U.S.S. Abraham Lincoln, and now we’re just expanding that idea by doing all of the ships in the Theodore Roosevelt Strike Group. This will be our new way of doing business, but it’s been a journey, and it’s exciting to see how we’re able to make such an impactful change and deliver agility for the fleet.”