Oleaje: AI for Early Warning of Extreme Coastal Waves

Oleaje began as a course project and became a national winner. On 17 September 2025 in Madrid, the tool created by Universidad Politécnica de Madrid students Sara Yagüe Rubio, Nerea Portillo Juan, and Mónica Ferrer Gómez-Cano earned Samsung’s National AI Prize at the SIC Summit: AI for a Better World, after judges highlighted its originality, social impact, and technical viability. Built to forecast coastal wave conditions up to 60 hours in advance, Oleaje targets a clear goal: earlier alerts that reduce damage to ports and coastal infrastructure and, most importantly, protect lives.

At the center of the effort is researcher Dr. Nerea Portillo, who recently defended her PhD at UPM’s School of Civil Engineering and studies maritime climate and port impacts. Portillo connected her doctoral work with the Samsung Innovation Campus curriculum on Python and artificial intelligence, then helped steer the team toward a deployable forecasting tool. “The frequency and intensity of extreme weather events is increasing,” she notes, and accurate anticipation strengthens coastal resilience. Her vision bridges two worlds: established numerical models in maritime engineering and modern machine learning that can add speed and precision.

Technically, Oleaje uses Long Short-Term Memory (LSTM) neural networks designed for time-series prediction. Trained on historical wave data, the networks learn statistical relationships that allow them to project near-term sea states from recent observations. Portillo explains the intuition simply: if the model knows the wave state at time T-1, it can infer conditions at time T. That approach scales to large climate and ocean datasets and promises faster, higher-resolution forecasts than many physics-only workflows, making it a practical complement rather than a replacement.

The project also reflects Portillo’s view that coastal engineering is inherently collaborative. She points to the need for biologists, oceanographers, and port engineers to work together, and she links technical excellence with social purpose. Her commitment extends beyond research: inspired by her physicist mother, she engages with the STEAMxTI program to bring real, relatable science role models into schools that rarely see them, aiming to narrow persistent gender gaps in engineering.

Recognition at the SIC Summit anchors a broader momentum. Samsung’s training pathway emphasizes responsible AI and practical impact, and Spanish officials underscored the value of ethical regulation and a human-centered approach. For Portillo, the next steps are concrete: integrate Oleaje with early-warning systems, couple it more tightly with physical models, and mature it through university-industry collaboration. “I have just finished my PhD,” she says, “and I plan to keep improving this tool and achieving effective integration with other prediction systems.” Oleaje is not an endpoint. It is a credible blueprint for how young, interdisciplinary teams can use AI to help coastal communities adapt to a changing climate.

Read Nerea’s journal articles:

Nerea Portillo Juan, Mónica Ferrer Gómez-Cano, Sara Yagüe Rubio, Vicente Negro Valdecantos. Improving Multi-Variable Wave Forecasting with AI: Integrating LSTM and Random Forest, Using a Window and Flatten Technique. Ocean Modelling, 199: 102638, 2026. https://doi.org/10.1016/j.ocemod.2025.102638

Nerea Portillo Juan, Vicente Negro Valdecantos, Peter Troch. Advancing artificial intelligence in ocean and maritime engineering: Trends, progress, and future directions. Ocean Engineering, 339: 122077, 2025. https://doi.org/10.1016/j.oceaneng.2025.122077

Read the CaminosUPM News

Read more about Coastal engineering and sustainable port infrastructure @ CIVILis