Predictive Autonomy: AI Transforms Naval Operations in Contested Seas

As maritime operations become more contested, distributed, and exposed to anti-access/anti-denial (A2/AD) threats, gray zone activity, and emerging dangers to subsea infrastructure, the demand signal for autonomy is shifting beyond unmanned navigation and edge-computing. The challenge now is not just sending a platform from point A to point B and sensor-data collection, but giving operators the ability to understand mission effectiveness, actually predict problems before they escalate, and adapt in real time across long-range, long-endurance operations in degraded communications environments. This challenge is being met by Integer Technologies, ranked #95 out of 5,000 on Inc.’s 2025 list of America’s Fastest-Growing Private Companies. Breaking Defense talked about predictive intelligence with Integer co-founders Duke Hartman, CEO, and Josh Knight, COO, and how their software called DIGIT is being used by the Defense Innovation Unit on its Combat Autonomous Maritime Platform Project (CAMP) developing extra-large, extended range unmanned undersea vehicles. Breaking Defense: Describe the maritime threat scenarios where autonomy and predictive intelligence play key roles. Knight: The threat environment that we’re looking at in maritime is increasingly complex, but also contested in the ocean environment. Traditional manned operations are getting progressively more costly and risky because of threats like growing anti-access/area denial capabilities and gray zone activities. We’re also starting to see the emergence of more undersea threats to vulnerable critical infrastructure, seabed infrastructure. Those are the types of threats that are driving a lot of the work that we’re doing in long-range autonomy – long range, but also long endurance. Of course, looking at theaters like the Pacific and from where the United States has historically operated, they would require long range. But there’s also missions like continuous monitoring of threats to subsea infrastructure that require long endurance to be practical. Systems that enable persistent surveillance, scalable presence, affordable mass, these are all things that are going to pay dividends in these sorts of high-risk maritime-threat environments. Certainly, there is a contemporary reminder that mine countermeasures is still very much a real mission in the ocean or maritime domain, and we’re looking at technologies that support missions like mine countermeasures and persistent operations in A2/AD types of environments. How do you position Integer to address this need and what’s different about what you provide? Knight: Integer is an agile defense technology company that transforms raw data into decision advantage, protecting our country and ensuring a safer world. We partner with industry and academia to create purpose-built solutions that refine complex information into actionable intelligence, empowering humans and autonomous systems to predict the probabilities in uncertain environments. Our digital engineering solutions portfolio spans robotic and unmanned systems, sensors and perception, power and energy systems, advanced manufacturing and cyber-physical systems. When we talk about digital engineering and software, we’ve developed predictive intelligence and what we call ‘mission assurance’ software. We’re not positioning ourselves as a vehicle platform vendor, nor strictly an autonomy provider, nor strictly a command and control company. We’ve found what we view as an important capability gap for Navy operations where systems today, whether they’re manned or unmanned, are in many cases reactionary to information as it comes in. This mission-assurance software is bringing a new capability to predict some degree into the future so that you can have better awareness of what’s going on in your theater and better predictive tools for how to adapt to changing conditions. We’ve already started to partner broadly with the best-of-breed vehicle and autonomy providers of surface and subsurface, manned and unmanned vessels. Our mission-assurance layer is the new aspect (in maritime autonomy). How did Integer come together? Why specifically address the maritime domain? Hartman: Josh and I both come from maritime engineering backgrounds, Josh being a naval architect and myself being a mechanical engineer whose work has always been in the ocean. When we started Integer in 2021, we saw that many of the other industries were adopting technology at a much faster rate than the maritime industry, and there was a lot of stagnation in the maritime sector, which was often a technology laggard. We thought that’s a problem given the way the geopolitical situation was changing; we are a maritime nation and we decided we needed to help to change that. The basis of our company was partnering with universities on focused and applied research that had a targeted aim – to produce a prototype that would actually function in a relevant environment. Integer’s job within that partnership is to carry that innovation across the Valley of Death, so to speak, and we have found funding models to do that. That’s where the magic has happened, because we can have one foot on the academic side where the greatest scientists and engineers are doing their research in a university setting, and crave real world problems to research and solve. Then we have another foot in the closed, secure environment of the defense industry. We get to focus those innovations on operational applications for the Department of War and make sure they work. These cannot be science experiments if they are on a fleet asset. That’s where we’ve focused on helping the Navy and other services, the operators in uniform, make better decisions faster. What do you mean by predictive intelligence? What is it that you’re predicting? Hartman: This method of prediction is really our differentiator. We start with physics. We understand the physics of your own vehicle or your platform or your system of systems, whatever engineered systems you’ve got. We then take in sensor information in real time and update models that we have of those engineered systems in faster than real time, as well as the environment and estimates of other assets in theater, including Red assets. If you can fuse all of those things together and you have compute on your platform that’s powerful enough to run, in some cases, Monte Carlo scenarios that can allow you to look ahead, you can analyze the impact on your mission. Does a change in the environment or in the posture of my own assets or of the adversary’s assets change my mission effectiveness or not? In what way? You can predict what might happen if you see a given behavior in, say, an adversary system and then adapt in real time. You’re predicting the move of the adversary. The term ‘digital twin’ is thrown around a lot and has become a buzzword, so it’s worth defining how we use it. These are operational models of the battlespace that are run at the edge on vehicles or platforms. We have digital twins of the vehicles, the platforms, the environments, and the adversaries’ vehicles. Having those running faster than real time, you can develop a ‘world model’ that allows you to predict mission outcomes with a complex system of systems, which empowers operators to make better decisions more rapidly. Also, once you’re evaluating multiple paths and courses of action, you can decide which one is the best by running those scenarios out. Should I send this set of USVs west to execute on this ISR mission or not? What happens to the mission if half of them have a mechanical failure? It allows you to think ahead in the same way that warfighters do today. Describe DIGIT, how it works and what problems it solves for warfighters. Knight: DIGIT is our mission assurance software platform with which we build these physics-based digital twins that fuse real-time data and AI to give operators better awareness and decision aids for operating thei