The HydroGym reinforcement learning platform for fluid dynamics
Nature, vol. 657, pp. 369–376
Abstract
Effective control of fluid flows is critical across transportation, energy and medicine, where it can increase lift, reduce drag, enhance mixing and attenuate noise1–3. Yet fluids are notoriously difficult to control because they involve high-dimensional, nonlinear and multiscale dynamics that resist conventional approaches4–6. Reinforcement learning has driven remarkable progress in fields such as protein folding and complex games, which have shared benchmarks and standardized environments7–10. Fluid dynamics has lacked such infrastructure, so each controller is typically tuned to a single geometry and operating condition, making progress difficult to accumulate, transfer and compare11–13. Here we introduce HydroGym, a solver-independent reinforcement learning platform providing more than 60 validated, openly available flow control environments spanning from canonical laminar flows to complex turbulent flows, with systematic progression in the Reynolds number up to Re = 4 × 105, and Mach number variations in two and three dimensions. Across these environments, agents repeatedly discover robust control principles, including boundary layer manipulation, disruption of acoustic feedback and reorganization of turbulent wakes. Critically, we demonstrate a proof of concept for zero-shot transfer, in which agents that are trained exclusively in inexpensive surrogate environments are deployed to challenging real-world scenarios such as a three-dimensional wing section. We achieve a 38% reduction in local skin friction while reducing exploration costs by four orders of magnitude compared with direct on-wing optimization. As this transfer exploits shared near-wall physics, the breadth of generalization remains open, suggesting a new pathway for research toward policy generalization across computationally prohibitive simulation environments. By offering a common, extensible foundation for reproducible research, HydroGym moves flow control from isolated case studies toward a cohesive community effort. HydroGym offers over 60 standardized reinforcement-learning flow-control environments and demonstrates zero-shot transfer to a 3D wing, cutting local skin friction by 38% and exploration costs by four orders of magnitude.
Authors 21
-
Christian Lagemann corresponding
Affiliation as printed
AI Institute in Dynamic Systems, University of Washington, Seattle, WA, USA. christian.lagemann@rwth-aachen.de
Department of Mechanical Engineering, University of Washington, Seattle, WA, USA. christian.lagemann@rwth-aachen.de
-
Affiliation as printed
AI Institute in Dynamic Systems, University of Washington, Seattle, WA, USA
Department of Mechanical Engineering, University of Washington, Seattle, WA, USA
-
Affiliation as printed
Chair of Fluid Mechanics and Institute of Aerodynamics, RWTH Aachen University, Aachen, Germany
-
Affiliation as printed
Data-Driven Fluid Engineering Laboratory, Inha University, Incheon, South Korea
Department of Mechanical Engineering, State University of New York, Incheon, South Korea
-
Affiliation as printed
Department of Aerospace Engineering, University of Michigan, Ann Arbor, MI, USA
-
KTH Royal Institute of Technology
Affiliation as printed
FLOW, Engineering Mechanics, KTH Royal Institute of Technology, Stockholm, Sweden
-
Technical University of Munich
Affiliation as printed
School of Engineering and Design, Technical University of Munich, Munich, Germany
-
Technical University of Munich
Affiliation as printed
School of Engineering and Design, Technical University of Munich, Munich, Germany
-
Technical University of Munich
Affiliation as printed
School of Engineering and Design, Technical University of Munich, Munich, Germany
-
Affiliation as printed
AI Institute in Dynamic Systems, University of Washington, Seattle, WA, USA
Department of Mechanical Engineering, University of Washington, Seattle, WA, USA
-
Affiliation as printed
AI Institute in Dynamic Systems, University of Washington, Seattle, WA, USA
Department of Mechanical Engineering, University of Washington, Seattle, WA, USA
-
Affiliation as printed
AI Institute in Dynamic Systems, University of Washington, Seattle, WA, USA
Department of Mechanical Engineering, University of Washington, Seattle, WA, USA
-
Affiliation as printed
Chair of Fluid Mechanics and Institute of Aerodynamics, RWTH Aachen University, Aachen, Germany
-
Conservatoire National des Arts et Métiers · Arts et Métiers · HESAM Université · Laboratoire de Dynamique des Fluides
Affiliation as printed
Arts et Métiers Institute of Technology, CNAM, DynFluid, HESAM Université, Paris, France
-
Technical University of Munich
Affiliation as printed
Munich Institute of Integrated Materials, Energy and Process Engineering, Technical University of Munich, Munich, Germany
School of Engineering and Design, Technical University of Munich, Munich, Germany
-
Matthias Meinke Aachen Chair of Fluid Mechanics Institute of Aerodynamics JARA Center for Simulation and Data Science
Affiliation as printed
Chair of Fluid Mechanics and Institute of Aerodynamics, RWTH Aachen University, Aachen, Germany
JARA Center for Simulation and Data Science, RWTH Aachen University, Aachen, Germany
-
Wolfgang Schröder Aachen Chair of Fluid Mechanics Institute of Aerodynamics JARA Center for Simulation and Data Science
Affiliation as printed
Chair of Fluid Mechanics and Institute of Aerodynamics, RWTH Aachen University, Aachen, Germany
JARA Center for Simulation and Data Science, RWTH Aachen University, Aachen, Germany
-
German Center for Neurodegenerative Diseases
Affiliation as printed
MediaTek Research, London, UK
Statistics and Machine Learning, German Center for Neurodegenerative Diseases, Bonn, Germany
-
Affiliation as printed
AI Institute in Dynamic Systems, University of Washington, Seattle, WA, USA
Department of Mechanical Engineering, University of Washington, Seattle, WA, USA
-
Ricardo Vinuesa corresponding
University of Michigan · KTH Royal Institute of Technology
Affiliation as printed
Department of Aerospace Engineering, University of Michigan, Ann Arbor, MI, USA. rvinuesa@umich.edu
FLOW, Engineering Mechanics, KTH Royal Institute of Technology, Stockholm, Sweden. rvinuesa@umich.edu
-
Steven L. Brunton corresponding
Affiliation as printed
AI Institute in Dynamic Systems, University of Washington, Seattle, WA, USA. sbrunton@uw.edu
Department of Mechanical Engineering, University of Washington, Seattle, WA, USA. sbrunton@uw.edu
Cited by 1 stored of 1
1 result
No patents citing this paper on Lens.org (checked 2026-10-06).