The US Navy wants to integrate artificial intelligence and machine learning into navigation aid systems on manned ships.

Apart from improving existing bridge navigation aids, the move is intended to act as a testbed for unmanned naval vessels in the pipeline.

As part of the decision, Naval Sea Systems Command (NAVSEA) has sought information on technologies that can be integrated into existing bridge systems such as electronic chart display and information systems, and develop a “human-in-the-loop safety of navigation bridge decision aid.”

Making Crew AI-ready

NAVSEA also requires interested parties to help build “user trust” in the technologies and train the crew in operating AI as a digital twin of present systems.

The “RFI [request for information] seeks to understand how interested parties would use existing sensors…to make maneuvering recommendations to bridge teams/watchstanders.”

The request states that the technology should be ready for assessment by 2025-27.

MANTAS T38 Devil Ray unmanned surface vehicle
MANTAS T38 Devil Ray unmanned surface vehicle during an operations demonstration at the seven-day Unmanned Integrated Battle Problem 21. Photo: US Navy

Testbed for Unmanned Vessels

The US Navy has been pushing to develop a range of unmanned surface and underwater vessels to execute tasks such as “offensive operations and intelligence, surveillance, and reconnaissance with reduced risk to crew and legacy vessels.”

Tom Reynolds of Huntington Ingalls Industries Technical Solutions was quoted at a conference by Defense One saying that installing autonomous and machine-learning technology on a crewed ship could be one way to prove the tech works. 

‘Wouldn’t Necessarily Replace Sailors’

“We don’t have to jump to unmanned right away,” Reynolds said. 

“A bridge to getting unmanned could be…the integration of autonomy onto manned platforms, and just have it run in the background.”

He said the technology wouldn’t necessarily replace sailors, “but basically aid an overwhelmed bridge crew.”





Source link