The wave is now being asked to file its own power budget
the wave is now being asked to file its own power budget: a machine learned parameterization called nlml stays stable in long term wave model runs and makes operational forecasting practical, a physics informed controller learns wave force predictions for grid compatible converters, and a digital twin pairs floaters with nvidia tools to anticipate loads.
the sea now files its own energy forecast.
Context
NLML (Journal of Geophysical Research: Machine Learning and Computation, January 2026, Los Alamos and Argonne authors) is a neural network that emulates the exact nonlinear wave interactions inside the WAVEWATCH III wave model. The abstract reports up to 136 times faster than the exact method, about 1.04 times slower than the fast approximation (DIA), twice the accuracy of DIA in global wave energy and mean parameters (up to 7 times in some regions), and stability through a year-long standalone simulation.
The physics-informed control paper (Ocean Engineering, 2025, Southampton and UCL) combines Physics-Informed Neural Networks with model predictive control for wave energy converters. It was tested on a point absorber in 35 sea wave scenarios from 1.25 m and 1 s to 2.50 m and 12 s.
Eco Wave Power's digital twin work: an NVIDIA blog (June 22, 2026) describes floaters attached to breakwaters with computers and electronics kept on land. A company press release (September 2, 2026) says it signed with AI engineering GmbH, both NVIDIA Inception members, to build a physics and data driven digital twin. Phase 1 models floater behavior, structural loads and energy input, then plans machine learning for forecasting energy yield and system loads.
The note says NLML 'makes operational forecasting practical'. The paper's claim is narrower: it bridges accuracy and speed in a year-long standalone WW3 run, and does not report operational use. The controller paper is about maximising power from a point absorber by predicting wave force, in simulation scenarios, and the abstract text read does not mention 'grid compatible converters', so that wording is unsupported here, not refuted. The digital twin is at Phase 1 per the press release: it only plans forecasting of yield and loads, and the 'anticipate loads' in the note is a stated goal, not a result. The NVIDIA link is membership in the Inception startup program and use of NVIDIA AI infrastructure, per the sources. Both digital twin sources come from the company or its sponsor.
Related work
- NLML: A Deep Neural Network Emulator for the Exact Nonlinear Interactions in a Wind Wave Model (OSTI, Journal of Geophysical Research: Machine Learning and Computation) ↗Source for the 136x speedup, accuracy and stability results.
- WW3-NLML code (Zenodo) ↗Released code that couples NLML to WAVEWATCH III.
- Non-causal model predictive control for rigid-body wave energy converters based on physics-informed neural networks (Ocean Engineering listing) ↗Source for the PINN and MPC controller and its 35 scenarios.
- Eco Wave Power Turns Waves Into Watts With NVIDIA AI Infrastructure and Digital Twins (NVIDIA blog) ↗Source for the floater setup and the NVIDIA Inception connection.
- Eco Wave Power Signs Agreement with AI engineering GmbH to Advance AI-Powered Digital Twin Platform for Wave Energy (company press release) ↗Source for the Phase 1 digital twin plan for loads and yield forecasting.
Watch next
- Look for the controller paper's power gain numbers against a standard controller. Check whether NLML has been run in an operational wave forecast cycle.
Sources
- NLML: A Deep Neural Network Emulator for the Exact Nonlinear Interactions in a Wind Wave Model (OSTI, Journal of Geophysical Research: Machine Learning and Computation)osti.gov
- Non-causal model predictive control for rigid-body wave energy converters based on physics-informed neural networks (Ocean Engineering listing)exa.ai
- Eco Wave Power Turns Waves Into Watts With NVIDIA AI Infrastructure and Digital Twins (NVIDIA blog)blogs.nvidia.com
- Eco Wave Power Signs Agreement with AI engineering GmbH to Advance AI-Powered Digital Twin Platform for Wave Energy (company press release)ecowavepower.com
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 4 October 2026 at 00:33 IST. Sources are the papers and datasets the note draws on.
View the original post ↗Embed this note
More notes
The air is now being asked to keep its own ledger
the air is now being asked to keep its own ledger: ecmwf’s aifs compo becomes the first ai model to forecast atmospheric composition globally every three hours, cleanair simulates 365 days of pm2.5 over china in ten seconds, and a unified framework maps six pollutants at one kilometer across the whole country. the air now files its own composition report.
read the note →The current is now being asked to draw its own map
the current is now being asked to draw its own map: china’s langya 2.0 predicts six ocean phenomena including internal waves and mesoscale eddies, a deep net called wenhai resolves eddies globally with air sea flux formulas built in, and scripps infers surface currents from the way temperature patterns deform in satellite images. the ocean now files its own circulation report.
read the note →The soil is now being asked to report its own carbon
the soil is now being asked to report its own carbon: a nix color sensor paired with generative data augmentation predicts soil organic carbon without a lab, random forest drives 74 percent of soil health mapping studies, and sentinel 2 tracks five year carbon change across france and italy from 922 samples. the dirt now files its own carbon account.
read the note →