Panel: AI-enhanced Reasoning and Agentic Systems in Fluid Dynamics Research
ORAL · Invited
Abstract
The panel will bring together domain experts to address critical questions: How might we better incorporate physics-informed approaches in the development and use of AI systems for fluids research? How can we standardize the documentation of researcher’s reasoning in AI models and AI reasoning chains in scientific publications? How can we avoid encoding existing theoretical biases into AI systems? Can we push past fundamental limitations of data-driven approaches? How do we document and disseminate knowledge in the era of AI? How can we handle the potential erosion of physical intuition as researchers increasingly rely on AI-generated insights? The panel aims to catalyze a new paradigm where fluid dynamicists become architects of AI reasoning about physical systems, potentially yielding breakthroughs in previously intractable problems while maintaining scientific integrity. Join us in reimagining how fluid dynamics knowledge can be created, formalized, validated, disseminated and leveraged in the age of AI.
–
Presenters
-
Karthikeyan Duraisamy
University of Michigan
Authors
-
Karthikeyan Duraisamy
University of Michigan
-
Peter J Schmid
King Abdullah University of Science and Technology
-
Gianluca Iaccarino
Stanford University
-
Venkatramanan Raman
University of Michigan
-
Luca Magri
Imperial College London, The Alan Turing Institute, PoliTo, Imperial College London
-
Beverley J McKeon
Stanford University