Contribution to development of autonomous manoeuvring controller

Nabtesco has developed Drive Train Controller; DTC for the autonomous container vessel "GENBU" in the MEGURI2040 Fully Autonomous Ship Program, promoted by the Nippon Foundation.

DeepSea developed the vessel motion-prediction model underpinning the DTC of Nabtesco—covering autonomous manoeuvring, including automated berthing and unberthing.

The developed data-driven manoeuvring motion model receives the ship's propulsion and steering signals—propeller revolutions (RPM), rudder angle, and thruster commands—along with prevailing wind, wave, and current conditions, and predicts how the vessel will move under a given set of commands.

By anticipating vessel motion accurately in actual operating conditions, the model helps the manoeuvring system to decide and act safely.

"The data-driven motion prediction capability captures how the specific vessel will respond in real conditions and hence it is key to the autonomous manoeuvring function.Seeing our motion-prediction model running on a commercially operating, government-certified autonomous vessel is a milestone for DeepSea " said Michalis Mathioudakis, Data Quality & Marine Modeling Team Lead, DeepSea Technologies.

Research & Innovation Director, Theodoros Stouraitis, also added, "This demonstrates that data-driven models can reliably anticipate how a ship will respond in real conditions—the foundation for safe, precise maneuvering with reduced crew workload, and a key step toward self-navigating vessels."

Find all about this project here.