Naval Applications of Machine Learning Workshop Advances AI Adoption
Technologies grouped under big data, artificial intelligence and machine learning are impacting virtually every aspect of life today. More importantly for the U.S. military and for the companies in the defense industrial base, these technologies, generally called AI, have the potential to change warfare in profound ways.
Numerous articles in the media have discussed various aspects of how AI will impact national security agencies. This constant drumbeat is important in highlighting how essential AI is to the long-term security and prosperity of the United States.
At the highest levels of national, military and naval doctrinal and policy publications, AI is called out as a critical technology. A sampling of policies makes this point:
The National Security Strategy notes: “New advances are already transforming the way we fight and the field of AI in particular is progressing rapidly.”
The National Defense Strategy put it this way: “The security environment is also affected by rapid technological advancement including artificial intelligence and autonomy.”
The Navy’s publication, A Design For Maintaining Maritime Superiority, says: “We should focus Navy efforts for fielding AI/ML algorithms on areas that most enhance warfighting, training and corporate decisions.”
The nation’s leadership established the National Security Commission on Artificial Intelligence to address this crucial technology. Co-chaired by Google CEO Eric Schmidt and former Deputy Secretary of Defense Robert Work, the commission issued their Final Report as a call to action to accelerate the insertion of these technologies into U.S. military platforms, systems, sensors and weapons.
Their report put their call to action in direct terms, stating: “We can still defend America and our allies without widespread AI adoption today, but in the future we will almost certainly lose without it.”
In her testimony before the U.S. Senate during her confirmation hearing for the position of deputy secretary of defense, Dr. Kathleen Hicks emphasized the importance of AI to the U.S. military. More recently, Secretary of Defense Lloyd Austin, speaking at the Reagan National Defense Forum in December 2021, addressed this issue directly: “We want to successfully lead the AI revolution.”
Earlier this month, the Department of Defense’s (DoD's) second iteration of a joint artificial intelligence organization opened for business as the Chief Data and Artificial Intelligence Office. The CDAO, as it is known, evolved from the Joint Artificial Intelligence Center and merged with the Defense Digital Service and the Office of Advancing Analytics.
Craig Martell, the new chief digital and artificial intelligence officer, emphasized that the DoD can achieve tactical wins in AI, but it needs to have scale and consistency over the next several years. “We can win against our pacing adversaries today, but we need to win 10 years from now, too. We need to win five years from now. We need to win three years from now. So, we need to be spending a lot of time thinking about how do we balance those two? How do we make sure that the AI that we build today is using a foundation that's going to make us even better five years from now?”
While the aspiration of the U.S. national security leadership to leverage AI technologies to provide U.S. and allied warfighters with a decisive edge is clear, what is less well-defined are the specific steps that are being taken to accelerate this process.
To answer that question, it is worth looking at what is being done at the grassroots level to find ways to make these technologies part of the DoD’s DNA.
One bright spot in this journey is the annual Naval Applications of Machine Learning Workshop (NAML), co-hosted in March 2022 by the AFCEA San Diego Chapter and the Naval Information Warfare Center Pacific.
The goals of NAML are to spread awareness of current machine learning research relevant to naval applications, to connect machine learning researchers with experts in military requirements and to build and strengthen collaborations within the defense research community.
The sixth annual NAML Workshop brought together keynote speakers and subject matter experts from across the nation. The latter group collaborates to do the hard work of determining where AI can be most useful in the military sphere, and then embarks on writing the code to make this a reality.
In his keynote address, the Naval Information Warfare Center's commander, Rear Adm. Douglas Small, USN, captured the importance of signature events like NAML, noting “The collective expertise from artificial intelligence communities ensures a running start in the race to information dominance. We have a few ideas on where we’re going with the machine learning operations paradigm, and delivering containerized applications through our platform is something we’ve demonstrated already through Advanced Naval Technology Exercises. We’ll continue to do more of that.”
Adm. Small’s keynote remarks put an exclamation mark on the importance of events like NAML. Emerging technologies like big data, artificial intelligence and machine learning are being worked by countless individuals in the military, industry and academia. No one organization has all the answers as to how to advance the adoption of these technologies by the U.S. military. That is why it is important to bring a diverse group of participants together—in person and virtually—to wrestle with these technology challenges.
NAML did just that this year. Over one thousand people registered for this event, representing the most diverse spectrum of participants ever. The “secret sauce” of NAML is that it puts together researchers working on complementary efforts and enables them to discuss their common challenges so they can collaborate to find effective solutions.
NAML 2022 addressed a wide variety of technologies and applications related to machine learning and artificial intelligence, including computer vision, reinforcement learning and natural language processing. Novel applications were discussed related to autonomous systems, platform analytics and battle management aids. Specialized sessions were organized on topics including synthetic data for machine learning and real-time machine learning. One topic that was once again prominent at NAML was radio frequency (RF) machine learning. Unlike the field of computer vision, where machine learning on images has largely been driven by commercial applications, machine learning on RF signals is being pioneered on DoD applications. Applications to radio and radar systems were presented, as well as cutting-edge research on novel computation methods for signal processing.
Another topic that grew in prominence at this year’s NAML event was AI assurance, including robustness of algorithms and the safety of deployed systems. Unlike many commercial applications where machine learning has found success, nearly every DoD application is safety-critical. Robustness and security cannot be an afterthought, and NAML provided a platform to discuss many of the important considerations on this topic all through the technology pipeline.
Building on the success of NAML 2022, AFCEA San Diego and the Naval Information Warfare Center Pacific continue to work to bring together those individuals from a wide array of organizations who are dedicated to ensuring that the U.S. military can leverage AI-related technologies to ensure that warfighters have a decisive edge against any adversary.
Future NAML events are being planned to continue discussing applications of machine learning for a wide array of DoD missions. Through these events, the NAML community has provided an opportunity for government researchers to broaden their views of how their work fits in to the larger picture, to build connections and understand concerns that need to be addressed to transition their work to products and services that directly support the DoD’s platforms, systems, sensors and weapons. The NAML community provides mentorship and professional development opportunities and increases the quality of government-produced research by enabling informal peer review and feedback.
By Katie Rainey, director of Science and Technology, Intelligence, Surveillance & Reconnaissance Department, Naval Information Warfare Center Pacific, and Capt. George Galdorisi, USN (Ret.), director, Strategic Assessments and Technical Futures, Naval Information Warfare Center Pacific.
The views expressed in this article are those of the authors, and do not necessarily represent those of Naval Information Warfare Center Pacific, the Department of the Navy, or the Department of Defense.