US Navy Uses AI-Enhanced Underwater Drones to Detect Threats

US Navy Enhances AI Deployment for Underwater Detection

The US Navy is expanding its use of artificial intelligence (AI) to automatically detect underwater targets after successful trials with underwater drones. This initiative is managed by the Pentagon’s Defense Innovation Unit (DIU), headquartered in California, which focuses on integrating commercial technology into military applications. According to defense officials, this advancement has significantly reduced the time required to scan the ocean floor for underwater mines by half.

Alex Campbell, the unit’s Navy service lead, indicated that the Navy is now rolling out new production contracts to broaden the AI technology’s application. These contracts aim to enhance undersea drones and investigate its potential for identifying enemy ships, planes, and other threats. The testing, spearheaded by US Marines Corps Major Nick Ksiazek from DIU’s AI portfolio, involves machine learning algorithms utilizing sonar sensors to detect underwater objects and navigate the seabed. Initiated two years ago, these AI models have transformed a previously tedious task for sailors of analyzing underwater footage to distinguish between harmless objects and potential explosives.

Instead of manually reviewing the footage, which is sent to sailors for analysis to secure commercial shipping lanes or hostile waters, the AI tools have streamlined the process. This development has allowed the Navy to reduce the personnel engaged in such missions by ten sailors and shorten the mission duration by two days. The Navy is also accelerating the updating process for AI models, pushing updates remotely when the drones surface instead of physically removing them from the water. Previously, deploying such updates could take up to six months; now, it can be accomplished in less than a week.

Adapting AI to Varied Ocean Terrains

AI models necessitate rapid retraining to adapt to different underwater terrains. Identifying objects on a sandy seabed contrasts with a rocky ocean floor or debris-filled seabed. The underwater drones are already operational in the Indo-Pacific region, participating in exercises there. “The bottom of the ocean in the Red Sea looks different from the bottom of the ocean off Hawaii,” Campbell noted, emphasizing the need for adaptable AI to cope with changing terrains and adversarial tactics.

Broader Implications for Military AI

This initiative underscores the Pentagon’s commitment to advancing AI warfare projects amidst an ongoing technological rivalry with China, which prioritizes AI in its military strategy. For this project, DIU collaborates with several tech firms, including Arize AI, Domino Data Lab, Fiddler AI, Latent AI, and Weights & Biases. These firms are working under contracts that have a ceiling of $7.5 million.

The success of this effort marks a significant leap forward in the Navy's operational efficiency and capabilities, promising enhanced security measures and innovation in defense strategies. This pioneering use of AI not only showcases technological evolution but also stands as a strategic maneuver to maintain a competitive edge in modern warfare.