Physics-guided ML to build digital twin for wind-assisted propulsion ships

Poster (konferens), 2025

Wind-assisted propulsion systems (WAPS) (Figure 1) are receiving renewed attention as the shipping industry confronts stricter greenhouse gas (GHG) emission regulations on the pathway toward net-zero emissions by 2050. This project aims to pioneer physics-informed machine learning (PIML)-based WAPS ship modeling1,2 by using key technological enablers, such as physics-informed neural networks (PINNs), and by addressing the following research questions (Figure 2):
1) How can a PIML framework be developed to combine physical principles with real-world sailing data for WAPS ships?
2) How can WAPS devices be controlled to optimize WAPS ship operation?
3) How can onboard crews be trained to operate the new WAPS ships efficiently and safely?