Advancing AI-driven marine weather forecasting with GNN and HPC

Three months ago, Neuralio A.I. and AlongRoute: Guiding Every Mile set out on a shared mission—to push the boundaries of marine weather forecasting through cutting-edge AI and HPC. 🌊 Today, we’re thrilled to announce the success of that collaboration.

Together, we’ve developed a state-of-the-art solution for predicting oceanographic conditions, specifically Significant Wave Height (SWH), using our Graph Neural Networks (GNN) and Gated Recurrent Units (GRU) model. This model, trained on over 500 million data points with the support of the Vienna Scientific Cluster (VSC5), has demonstrated exceptional accuracy, helping to improve the safety and efficiency of maritime operations. 🚢

A huge thank you to EuroCC Austria for providing the HPC infrastructure that made this possible, and to the dedicated teams from both Neuralio A.I. and AlongRoute: Guiding Every Mile, including Paraskevi Vourlioti, Theano Mamouka, Maria Banti, Charalampos Paraskevas, Stelios Kotsopoulos, Vasileios Alexandridis and Georgia Kalantzi.

This is just the beginning—we're excited for what's to come as we continue to explore new frontiers in AI-driven marine forecasting.

easychair.org/publications/preprint/RLRL