U.S Navy Retrains Deep-Sea AI Sonar In Under 5 Minutes At The World’s Largest Naval Exercise

U.S Navy, along with Lockheed Martin, successfully tested an AI-driven sonar system during Rim of the Pacific 2026 (RIMPAC), the world’s largest naval exercise. In the trials, the system quickly identified and targeted underwater threats while allowing the operators to retrain the system’s machine learning model in less than 5 minutes.

This presents a much more efficient and quicker way to identify submarines by processing real-time acoustic data and by providing real-time updates to the crew members.

According to Lockheed Martin, the technology also allows the operators of naval helicopters to monitor twice as many sonar feeds compared to old systems.

The system has been developed by Lockheed Martin along with software company FUSE Inc.

SensorMAX first analyses live Sonar information and then provides information to flight crews and ground operators.

If the crew finds an unfamiliar sound underwater, they can retrain the AI model in less than 5 minutes and send the updated software to aircraft while they remain in flight. This update process enables operators to enhance detection without impacting the mission.

It functions as a sonar assistant, enabling helicopter crews to handle more information during anti-submarine operations, with a capacity to process a maximum of 8 sonobuoy feeds simultaneously while doubling the capacity of legacy equipment.

By analysing more acoustic streams, operators get a bigger picture of underwater activity without increasing the workload of the crew.

The system software can be integrated into an aircraft, surface vessels, or even ground stations, since it is lightweight and has a platform-independent design.

“Today’s battlespace requires the defence industry to accelerate capability delivery and seamlessly integrate commercial technology; at RIMPAC 2026 our mission-focused engineers demonstrated exactly how we are meeting that challenge,” said Devon Rodgers, vice president and general manager of Undersea Mission Systems at Lockheed Martin.

The successful test supports the U.S Navy’s wider effort to connect manned and unmanned systems across the maritime domain.

Lockheed Martin also said that the demonstration at RIMPAC is a step towards integrating AI into maritime sensors. Instead of depending on fixed software, operators can enhance acoustic detection models and distribute updates across the connected platforms.