Unlocking the future of underwater defense: How the U.S. Navy and DIU are leading AI innovation with AWS

The U.S. Navy has long been at the forefront of innovation, particularly when it comes to securing our nation’s interests beneath the waves. As the challenges of underwater defense continue to evolve, the Navy, in partnership with the Defense Innovation Unit (DIU), has embarked on a groundbreaking initiative to harness the power of artificial intelligence (AI) and machine learning (ML) for underwater target threat detection. At the heart of this effort is Project Accelerated Machine Learning for Maritime Operations (AMMO)—an ambitious program designed to ensure that the Navy’s unmanned underwater vehicles (UUVs) remain agile, adaptive, and operationally effective in even the most complex environments.

The challenge: Adapting to a dynamic underwater battlefield

Traditionally, the Navy has relied on machine learning models to support underwater threat detection by UUVs. However, the static nature of these models posed a significant risk. In an ever-changing underwater landscape where enemy tactics and environmental conditions constantly evolve, the ability to quickly update and redeploy ML models is crucial. Without this capability, the Navy’s mission-critical systems risked becoming obsolete, producing inaccurate or unreliable results that could compromise national security.

Recognizing this challenge, the Navy partnered with DIU to explore how commercial technology could be leveraged to address these limitations. The goal was clear: build an infrastructure that would allow for the continuous monitoring, retraining, and deployment of Automatic Target Recognition (ATR) ML models from the cloud to the edge, ensuring that the Navy’s UUVs could maintain operational superiority.

The solution: Domino Data Lab orchestrates AI innovation on AWS GovCloud

In 2022, the Navy and DIU issued a call to the private sector, seeking innovative solutions to create a robust machine learning operations (MLOps) pipeline that could meet their needs. This call resulted in the formation of Project AMMO, a collaborative effort involving several leading technology companies. At the core of this collaboration is Domino Data Lab, recently recognized with a major award from the U.S. Department of Defense (DoD) Accelerate the Procurement and Fielding of Innovative Technologies (APFIT) program. This recognition underscores Domino’s role as an AI prime, responsible for orchestrating the integration of leading open-source and commercial technologies from vendor partners Weights & Biases, Latent AI, Fiddler AI, and Arize. Each of these companies brought a unique capability to the table, collectively building a comprehensive solution that would transform the Navy’s approach to underwater defense.

As the backbone of the MLOps pipeline for Project AMMO, Domino’s Enterprise AI platform, hosted on AWS GovCloud (US), enables rapid innovation, operational agility, and secure model deployment. By providing a unified environment for data science and machine learning, Domino allows the Navy’s ML developers to track, modify, and redeploy models with unprecedented speed. This orchestration of best-in-class tools ensures that the Navy can continuously adapt to new threats and maintain a decisive edge in underwater operations.

The impact of APFIT and Project Ammo: Strengthening national security through AI

The success of Project AMMO has been profound. During a trilateral Integrated Battle Problem exercise in Fall 2023, the Navy demonstrated its ability to collect sensor data, retrain models against an expanded threat environment, and deploy updated detection models back into the hands of operators—all within a matter of days. This operational capability, supported by Domino’s APFIT-funded platform, reduced the time needed for model updates by 97 percent, from six months to just a few days.

This capability has dramatically enhanced the Navy’s operational effectiveness, providing the tools they need to detect and respond to underwater threats with unprecedented speed and accuracy. As such, Domino and the other vendors in the project received a success memo from the DIU. Read the DIU press release to learn more.

The APFIT award solidifies Domino’s position as a critical partner for scaling AI in national defense. Beyond underwater missions, the methodologies and tools developed through this initiative are now being considered for broader application across the DoD. From subsea warfare to other critical missions, Domino is enabling the DoD to deploy and manage AI at scale, meeting the highest standards of governance, security, and operational impact.

Looking ahead: Domino’s commitment to public sector innovation on AWS

As we look to the future, the success of Project AMMO offers a glimpse into the transformative potential of AI and ML in defense. By leveraging the power of cloud computing and the expertise of leading technology partners, the Navy and DIU have set a new benchmark for what is possible in the realm of underwater defense.

Domino Data Lab’s growing investments in public sector initiatives, including achieving FIPS 140-2 compliance, further demonstrate its commitment to supporting the most sensitive national security missions. As a member of the NIST AI Safety Institute Consortium, Domino is dedicated to advancing responsible AI practices, ensuring that its solutions are not only effective, but also trustworthy and transparent.

For IT and data science leaders in the public sector, the lessons from Project AMMO are clear. The ability to deploy, monitor, and update AI models rapidly is not just a technical challenge—it’s a strategic imperative. Backed by the APFIT award, Domino is leading the way in empowering the U.S. government to navigate the complexities of AI and ML, providing the tools and expertise needed to ensure national defense capabilities remain second to none.

Conclusion

In the rapidly evolving world of AI, robust governance is essential—not just for compliance but for innovation and business value. Domino Data Lab, in partnership with AWS, offers an integrated, automated solution that supports both. As you scale your AI efforts, trust Domino and AWS to provide the governance framework you need to succeed. Make sure to explore our offerings on the AWS Marketplace to see how Domino can help you future-proof your AI initiatives.

Now is the time to take action. Learn more about AI governance in our Playbook for Strategies and Practices for Responsible AI, to discover how our integrated solutions can help your organization not only meet today’s governance challenges, but also prepare for tomorrow’s opportunities. Or, watch our on-demand webinar,  Mitigating AI risks: Governance strategies for GenAI & new threats, featuring Diya Wynn, Responsible AI lead at AWS.