Physics Practical Magic

An AI that spent its life studying water pipes was put on an airplane wing and instantly figured out how to cut fuel-wasting drag by 10%.

April 13, 2026

Original Paper

Physics-guided surrogate learning enables zero-shot control of turbulent wings

Yuning Wang, Pol Suarez, Mathis Bode, Ricardo Vinuesa

arXiv · 2604.09434

The Takeaway

Managing turbulent air is so complex that computers usually need months to figure out a single wing shape. This breakthrough allows a "plug-and-play" AI to handle the chaos of flight instantly, making planes much more efficient.

From the abstract

Turbulent boundary layers over aerodynamic surfaces are a major source of aircraft drag, yet their control remains challenging due to multiscale dynamics and spatial variability, particularly under adverse pressure gradients. Reinforcement learning has outperformed state-of-the-art strategies in canonical flows, but its application to realistic geometries is limited by computational cost and transferability. Here we show that these limitations can be overcome by exploiting local structures of wa