by Cmdr. Bennett Christman and Lt. Cmdr. Ryan Hilger

For a good part of the Submarine Force’s history, we were the innovators, the disruptors. We broke free of the surface, sustained our depth and reach with nuclear power, and increased our stealth and mission capability through quieting, sensors, and computers.

Today, however, we are struggling to keep up with the pace of innovation in private industry. We are surrounded by examples of innovative commercial technologies changing the world before our very eyes, yet many we may not find on a submarine for several years.
One such example is the explosion of artificial intelligence and machine learning (AI/ML). Each day brings advances in image recognition, automobile autonomy, machine mastery of strategy games, new recommendation engines, and ground-breaking optimization analytics. The Navy as well as the other services have tried to follow, ingest, and replicate these innovations, but bureaucratic and cultural barriers continue to hinder the adoption of new and innovative AI solutions. Many in the Department of Defense (DoD) see AI/ML as a technology still in its infancy, while industry leaders consider a large set of AI deep learning applications “state of the practice.”
Exacerbating this problem is the explosion of “consultants” who seek to capitalize on AI hype, touting its abilities to solve any problem. Without an AI-literate class of decision makers in DoD, breaking out the signal from the noise can prove challenging. But even in this noisy environment, strong signals can emerge. It starts with a single Sailor with the right knowledge, skills, mindset, and passion to point it out to the rest of us.

This is one of those stories, a story in which one junior officer (JO) and a small group of Sailors made a big difference. It’s a story of innovation from the deckplate, the power of teams, and finding a way to own and solve the problems at hand. We hope that the lessons from this story encourage other Sailors with innovative ideas to come forward and contribute—not just in AI, but in all the many areas where the Submarine Force can and must improve.

Project Harbinger
The origins of what became known as Project Harbinger start with Lt. Austin Anderson, a JO from USS Springfield (SSN 761). In 2017, he was on his post-JO shore duty as a Secretary of the Navy Innovation Advisory Council (NIAC) fellow. Lt. Anderson was evaluating the use of AI/ML algorithms and their applicability to sonar and fire control problems like contact identification and solution development. He taught himself how to build and train AI/ML algorithms and set to work on recreating our primary fire control algorithms.

The early results were excellent, and he clearly saw the technology’s potential to substantially improve our capabilities. Despite the success, Lt. Anderson wasn’t sure how to get these algorithms deployed on submarines. While they showed great promise, his ideas needed refinement and endorsement. Like many JOs, he knew nothing about how the Navy sets requirements and acquires new capabilities. He was not going to be able to transition this technology on his own.

He did, however, start showing his results to other Submariners. He found passionate advocates in a small community of officers in the Pentagon and elsewhere. They included Capt. Scott McGinnis (Secretary of Defense Corporate Fellow at Stanford Research Institute), Cmdr. Bennett Christman (Chief of Naval Operations N00Z), Cmdr. Cameron Aljilani (Office of the CNO (OPNAV) N97), Cmdr. Dan Stock (OPNAV N97), Lt. Cmdr. Ryan Hilger (OPNAV N97), Lt. Cmdr. Joe Huck (USMC Strategic Initiatives Group), and Lt. Christian Mineur (Navy Digital Warfare Office (DWO)). Together they formed an informal team and started pitching the idea of AI algorithms to improve solution development.

Early on, they briefed several decision makers in the Pentagon. Most were initially skeptical about applying AI/ML to undersea warfare; they did not see its purpose or how it was different from what the capability development community was already doing. To better frame their argument, the team decided to create a strategy document to clearly articulate why and how AI should be applied in the Submarine Force.

With a clearer problem definition and refined strategy document, the team briefed the Submarine Transition Advisory Board, the board that prioritizes research and development for the Submarine Force, in May 2018 and later Vice Adm. Richard, then the incoming Commander Submarine Forces (COMSUBFOR), and Rear Adm. Dollaga, who had recently been relieved as Commander, Undersea Warfighting Development Center (UWDC). Rear Adm. Dollaga spent two hours with the team learning everything he could about Lt. Anderson’s prototype algorithms and the strategy the team was proposing.

Receptive to the team’s recommendations and energized by their promising ideas, he asked his outgoing chief of staff, Capt. John McGunnigle, to stay on for a year and develop an AI/ML capability pipeline for the Submarine Force. Lt. Anderson was selectively detailed to the Pentagon after his fellowship so that he could work in the DWO full time. Together, Capt. McGunnigle, Lt. Anderson, and Lt. Mineur have been hard at work on Project Harbinger ever since.

Project Harbinger is already demonstrating capability. Working hand-in-hand with partners in the submarine advanced development community, the execution team developed a data pipeline for training algorithms, fielded a prototype appliance for shipboard use, and deployed code from ashore to sea that was executed on an operational submarine—all in less than a year. The team is just getting started.

The Lessons
While the rest of the Project Harbinger story remains to be told, here are five lessons from the team’s experience that other Submariners can use to solve the problems facing the Submarine Force.

Lesson 1: Commanding Officers should be on the lookout for talent and place people with great ideas in billets where they can make a difference.

Lt. Anderson’s CO, Cmdr. Brent Spillner, encouraged him to apply for the NIAC fellowship right after the boat received the announcement message. He knew about Lt. Anderson’s coding skills and creativity. He also knew that sending Lt. Anderson to NIAC was in the best interests of the Navy, taking the time to personally engage with the detailers to endorse his selection and assignment.

Had Lt. Anderson not been in this assignment, he wouldn’t have been afforded the time to teach himself to code the algorithms, nor would he have had the opportunity to explore their applications to the Submarine Force. While NIAC no longer exists, COs can still encourage promising Sailors by assigning them to places like UWDC, DWO, our submarine program offices (e.g., IWS 5, PMS 401), and Secretary of the Navy Tours with Industry.

Lesson 2: If you see something, say something.

Leadership isn’t just for COs. Even deep within a staff, you as a JO or petty officer can advocate for changes that affect your boat or even the entire Force. This type of staff-level leadership is about owning an idea from inception through to its implementation. The first step is to tell someone about your idea, typically your department head or CO. Lt. Anderson didn’t know how to implement his idea, but he knew that his former Engineer, Lt. Cmdr. Hilger at OPNAV N97, might. That got the ball rolling and ultimately led to the formation of a diverse team of mid-grade officers who rallied around Lt. Anderson’s project and generated the access to senior leadership that the project needed to succeed.

If you’re at a loss for whom to contact, consider UWDC. Its role is to serve as an entry point for any good idea to improve capability and concepts of operation in the Submarine Force. Sailors who want to make a change should brief their chain of command and then contact the UWDC Capability Development Division.

Lesson 3: It takes a (small and diverse) village to make a change.

If you’ve got an idea, find a group of supporters who can help you. Recognize what you don’t know and reach out to find those who can contribute what you need to move forward. There is plenty of evidence that small, diverse teams produce the best results, so be selective. While all of your friends may be great people, they may be too much like you to bring about success. Instead, find people who know things you don’t, who challenge your thinking, and who bring different perspectives. If you can, find people who have connections to resources and access to those with influence. The connections may not only afford you the resources and leadership endorsement you need, but they can provide an outside perspective to evaluate your team’s ideas.

Be careful, however, not to let the team’s size continue to grow unbounded. There is also evidence that groups become less effective as they expand beyond what is absolutely necessary. This is the idea encapsulated by Jeff Bezos’ famous “Two Pizza Rule,” which says that a team’s size should be no greater than the number of people that you can feed with two pizzas.

The Harbinger team met this test. It was small and spanned multiple perspectives, ranks, and organizations. Lt. Mineur had just joined the Navy’s DWO after a Secretary of the Navy Tour with Industry at General Electric Digital. He had numerous commercial industry practices to inform the team’s thinking. Similarly, Cmdr. Stock worked closely with the acquisition community on what capabilities to bring to Submarine Warfare Federated Tactical Systems. His knowledge was critical to understanding how Lt. Anderson’s algorithm could get fielded. Finally, Capt. McGinnis and Cmdr. Christman had a wealth of contacts across the Pentagon and beyond. These were key to getting an audience with the right decision makers.

Lesson 4: Write it down.

Adm. Rickover famously said, “Nothing so sharpens the thought process as writing down one’s argument.” The team initially met skepticism from Submarine Force decision makers because their argument wasn’t yet fully formed. Merely presenting a demonstration of Lt. Anderson’s algorithms failed to clearly answer the frequent question of “What problem are you trying to solve?” Recognizing the need to strengthen their argument, the team turned to writing their ideas in the format of an operational design. They laid out an argument for why the Submarine Force needs AI algorithms and what investments and decisions senior leaders needed to make to solve the problem. As the team developed this product, it sharpened their own thinking and gave them a clear point of departure for discussions with senior leaders.

This experience shows the value of going through the exercise of defining the problem that you are trying to solve. If you have an idea, write down an explanation of how it will solve that problem, and then identify the steps that you and others must take to bring your idea to fruition. Don’t be satisfied with your first draft. Keep sharpening your arguments by presenting your paper to different audiences for feedback. Listen and understand the criticism, and try to address valid concerns through revision. In the end, you will not only have a document that provides clarity on what you are trying to achieve, but you will be able to explain and defend your ideas in any brief or discussion.

Lesson 5: The seeds of ideas need senior leaders to provide the soil in which they can take root and grow.

We are fortunate to have many young Sailors with sharp digital skills. Some were acquired in college, while others might have been learned in less traditional ways. This is in sharp contrast to most of our key Submarine Force decision makers, who attended college before the advent of smartphones, tablets, the cloud, and even the Internet. This description of reality is in no way meant to be pejorative; it just describes differences in the backgrounds of those who might be most likely to see and have ideas about areas of digital potential and those who are going to decide on the adoption of those ideas. Knowing this, it is critical that the Submarine Force leaders keep generally up to date on the latest technology trends and listen with an open mind to Sailors presenting a different approach to solving a particular problem.

In this case, we were fortunate that Vice Adm. Richard knew a great deal about the technology the team presented. He even wanted to broaden the scope to more than just the applications and algorithms the team presented, ultimately telling us, “Full speed ahead!”

The Submarine Force is maintaining its undersea superiority through rapid innovation and deployment of capabilities to the Fleet. This new paradigm is slowly transforming the acquisition community and demonstrates that Sailors with technical expertise have the potential to directly contribute to the fight, keeping the Submarine Force battle ready.

The team was able to innovate from the deckplate through the combination of technical expertise, critical and visionary thinkers, networks, and the commitment to owning the problem and the solution. The lessons from this effort can help other hidden teams around the Submarine Force get buy-in for their ideas with leadership and find like-minded Sailors to help develop and implement the solutions. We need to foster this kind of innovation and promote ideas that have the potential to improve our warfighting capability. If you have them, COMSUBFOR stands ready to back you in trying to implement them. Send them a note at

Cmdr. Christman is the Navy’s Federal Executive Fellow at the Johns Hopkins University Applied Physics Lab and Prospective CO, USS New Hampshire (SSN 778).

Lt. Cmdr. Hilger is an Engineering Duty Officer (EDO) assigned to Strategic Systems Program. Prior to becoming an EDO, he served in OPNAV N97 and as Engineer on USS Springfield (SSN 761).