Project Summary
This project develops an AI-powered routing system to help ships with wind-assisted propulsion reduce fuel use and emissions. By combining real-world weather data, vessel performance models, and advanced machine learning, it delivers smarter, greener voyage planning. The solution supports the maritime industry’s shift toward low-carbon, energy-efficient operations.
Project Achievements
This project develops an AI-powered routing system to help ships with wind-assisted propulsion reduce fuel use and emissions. By combining real-world weather data, vessel performance models, and advanced machine learning, it delivers smarter, greener voyage planning. The solution supports the maritime industry’s shift toward low-carbon, energy-efficient operations.
Conclusions
The project successfully demonstrates that integrating Deep Reinforcement Learning (DRL) with physics-informed simulations provides a superior framework for maritime routing. By accurately modeling Wind-Assisted Propulsion (WAP) dynamics, the solution achieves significant fuel savings and emission reductions, directly supporting global maritime decarbonization goals. The developed method ensure operational robustness by effectively managing forecast uncertainties and partial observability in volatile metocean conditions. This verified and scalable platform offers a practical, high-performance alternative to traditional routing techniques.
Next Steps
Futher research and development with funding support to advance the proof-of-concept achieved in this project through routes such as CMDC7, EU Horizon.
