AI in Fluid Dynamics Research at Leeds

Fluid dynamics has had a long and distinguished history. Over time, the tools used in this field have changed significantly, but have always kept pace with the times.
Initially, theories were developed using pen and paper, verified by often sparse experimental measurements. The introduction of modern computers ushered in an era of computational modelling. Today, we are in the age of big data, characterised by extensive datasets from sensors used in a wide variety of fields such as medicine, astrophysics and atmospheric science. These large and complex datasets provide the next big challenge for fluid dynamics research.

Advances in machine learning and AI, designed to extract insight from large datasets, represent the very latest additions to the research toolkit. At the University of Leeds, machine learning is a vibrant area of research with cutting-edge applications across the breadth of the institution. Researchers at the Leeds Institute for Fluid Dynamics are leveraging these new techniques to gain new data-driven insights.

To illustrate some of the AI-led research in fluid dynamics at Leeds, we have compiled four cases studies on diverse topics, highlighting how data-led science is at the forefront of our work:

- Modelling blood flow
- Improving air quality in indoor environments
- Short-term storm forecasting in Africa
- Predicting space weather