CWI Harnesses AI to Tackle Fluid Turbulence with NWO Grant
Amsterdam, Friday, 9 May 2025.
The Centrum Wiskunde & Informatica (CWI) has received a €50,000 grant to develop AI models that simulate fluid turbulence, potentially revolutionizing fields from meteorology to aerospace.
Groundbreaking Approach to Complex Physics
Benjamin Sanderse, head of the Scientific Computing group at CWI, will lead this innovative research project utilizing neural stochastic differential equations to uncover the hidden mathematical patterns behind turbulence [1]. The €50,000 grant, awarded through NWO’s Open Competition Domain Science – XS program, aims to support this speculative yet potentially transformative initiative [2].
The Challenge of Turbulent Systems
Turbulence manifests in numerous forms throughout nature, from smoke patterns in the air to gas movements in distant galaxies [1]. While scientists understand the basic differential equations governing these systems, solving them for real-world applications has proven computationally overwhelming [1]. This limitation has historically restricted our ability to accurately predict and model turbulent flows [GPT].
AI-Powered Innovation
The research introduces a novel stochastic approach inspired by recent advances in generative AI models [1]. This methodology incorporates random components into the equations, enabling the model to predict not just single outcomes but entire distributions of possible turbulent behaviors [1]. The project will partially fund a postdoctoral researcher to develop these new probabilistic turbulence models [2].
Broader Applications and Future Impact
While initially focused on turbulence simulation, the implications of this research extend far beyond fluid dynamics. The combination of physical laws with generative machine learning represents a significant step forward in scientific computing [1]. The project, announced on May 8, 2025, positions CWI at the forefront of combining traditional physics with cutting-edge AI methodologies [2].