mlsanityData to 7 Oct 2026

Physical Design using Differentiable Learned Simulators

arXiv 2202.00728 ↗First posted 1 Feb 2022cs.LGcited by 37

Not on the map. Why?

Abstract (arXiv)

Designing physical artifacts that serve a purpose - such as tools and other functional structures - is central to engineering as well as everyday human behavior. Though automating design has tremendous promise, general-purpose methods do not yet exist. Here we explore a simple, fast, and robust approach to inverse design which combines learned forward simulators based on graph neural networks with gradient-based design optimization. Our approach solves high-dimensional problems with complex physical dynamics, including designing surfaces and tools to manipulate fluid flows and optimizing the shape of an airfoil to minimize drag. This framework produces high-quality designs by propagating gradients through trajectories of hundreds of steps, even when using models that were pre-trained for single-step predictions on data substantially different from the design tasks. In our fluid manipulation tasks, the resulting designs outperformed those found by sampling-based optimization techniques. In airfoil design, they matched the quality of those obtained with a specialized solver. Our results suggest that despite some remaining challenges, machine learning-based simulators are maturing to the point where they can support general-purpose design optimization across a variety of domains.

Key facts. Cited by 37 papers in the map (data through 7 Oct 2026). Builds on 35 of its 77 references that are in the map. First posted on arXiv 1 Feb 2022.

Builds on

35 of 77 references.

References · 35

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2017Neural Message Passing for Quantum Chemistrycited by 3,460
  3. 2020Fourier Neural Operator for Parametric Partial Differential Equationscited by 2,806
  4. 2018Relational inductive biases, deep learning, and graph networkscited by 1,341
  5. 2020Learning to Simulate Complex Physics with Graph Networkscited by 867
  6. 2020Learning Mesh-Based Simulation with Graph Networkscited by 840
  7. 2016Training Deep Nets with Sublinear Memory Costcited by 825
  8. 2016Automatic chemical design using a data-driven continuous representation of moleculescited by 762
  9. 2019When to Trust Your Model: Model-Based Policy Optimizationcited by 753
  10. 2018Deep Reinforcement Learning in a Handful of Trials using Probabilistic Dynamics Modelscited by 748
  11. 2019The frontier of simulation-based inferencecited by 490
  12. 2021Machine learning accelerated computational fluid dynamicscited by 461
23 more
  1. 2021Brax -- A Differentiable Physics Engine for Large Scale Rigid Body Simulationcited by 375
  2. 2019DiffTaichi: Differentiable Programming for Physical Simulationcited by 248
  3. 2018Graph networks as learnable physics engines for inference and controlcited by 242
  4. 2018Generalisation in humans and deep neural networkscited by 237
  5. 2018Learning Particle Dynamics for Manipulating Rigid Bodies, Deformable Objects, and Fluidscited by 219
  6. 2019Towards Physics-informed Deep Learning for Turbulent Flow Predictioncited by 203
  7. 2019Benchmarking Model-Based Reinforcement Learningcited by 158
  8. 2021Embodied Intelligence via Learning and Evolutioncited by 154
  9. 2018Deep Learning Methods for Reynolds-Averaged Navier-Stokes Simulations of Airfoil Flowscited by 147
  10. 2019Prediction of Aerodynamic Flow Fields Using Convolutional Neural Networkscited by 144
  11. 2019PHYRE: A New Benchmark for Physical Reasoningcited by 137
  12. 2020Combining Differentiable PDE Solvers and Graph Neural Networks for Fluid Flow Predictioncited by 137
  13. 2018Flexible Neural Representation for Physics Predictioncited by 91
  14. 2021Learned Coarse Models for Efficient Turbulence Simulationcited by 90
  15. 2018SPNets: Differentiable Fluid Dynamics for Deep Neural Networkscited by 62
  16. 2019Rapid trial-and-error learning with simulation supports flexible tool use and physical reasoningcited by 60
  17. 2020On the role of planning in model-based deep reinforcement learningcited by 57
  18. 2021An End-to-End Differentiable Framework for Contact-Aware Robot Designcited by 53
  19. 2019Neural reparameterization improves structural optimizationcited by 44
  20. 2019Deep Imitation Learning of Sequential Fabric Smoothing From an Algorithmic Supervisorcited by 44
  21. 2020Model-Predictive Control via Cross-Entropy and Gradient-Based Optimizationcited by 32
  22. 2020Data-driven topology optimization of spinodoid metamaterials with seamlessly tunable anisotropycited by 22
  23. 2021Learning Dynamics Models for Model Predictive Agentscited by 21

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 7; 2023: 14; 2024: 7; 2025: 8; 2026: 1.

37 papers.

Sort:

By area:

  1. 2022Fourier Neural Operator with Learned Deformations for PDEs on General Geometriescited by 416
  2. 2022PDEBENCH: An Extensive Benchmark for Scientific Machine Learningcited by 396
  3. 2023Geometry-Informed Neural Operator for Large-Scale 3D PDEscited by 293
  4. 2022MultiScale MeshGraphNetscited by 94
  5. 2023Operator Learning with Neural Fields: Tackling PDEs on General Geometriescited by 80
  6. 2023Eagle: Large-Scale Learning of Turbulent Fluid Dynamics with Mesh Transformerscited by 54
  7. 2022Learning to Accelerate Partial Differential Equations via Latent Global Evolutioncited by 38
  8. 2023Learning Neural PDE Solvers with Parameter-Guided Channel Attentioncited by 37
  9. 2023Learning Flexible Body Collision Dynamics with Hierarchical Contact Mesh Transformercited by 18
  10. 2023Swarm Reinforcement Learning For Adaptive Mesh Refinementcited by 16
  11. 2023Grounding Graph Network Simulators using Physical Sensor Observationscited by 15
  12. 2023MMGP: a Mesh Morphing Gaussian Process-based machine learning method for regression of physical problems under non-parameterized geometrical variabilitycited by 15
25 more
  1. 2024A review of graph neural network applications in mechanics-related domainscited by 14
  2. 2023Drag-guided diffusion models for vehicle image generationcited by 13
  3. 20233D-IntPhys: Towards More Generalized 3D-grounded Visual Intuitive Physics under Challenging Scenescited by 12
  4. 2022Port-metriplectic neural networks: thermodynamics-informed machine learning of complex physical systemscited by 11
  5. 2022Thermodynamics of learning physical phenomenacited by 10
  6. 2025VLMgineer: Vision Language Models as Robotic Toolsmithscited by 9
  7. 2024OmniArch: Building Foundation Model For Scientific Computingcited by 8
  8. 2024Latent Intuitive Physics: Learning to Transfer Hidden Physics from A 3D Videocited by 8
  9. 2023Learning to Design and Use Tools for Robotic Manipulationcited by 7
  10. 2023Graph Neural Network Based Surrogate Model of Physics Simulations for Geometry Designcited by 5
  11. 2024Inverse analysis of granular flows using differentiable graph neural network simulatorcited by 5
  12. 2025RobotSmith: Generative Robotic Tool Design for Acquisition of Complex Manipulation Skillscited by 4
  13. 2023SURFSUP: Learning Fluid Simulation for Novel Surfacescited by 3
  14. 2023Diffusion Generative Inverse Designcited by 3
  15. 2024Up-sampling-only and Adaptive Mesh-based GNN for Simulating Physical Systemscited by 3
  16. 2025Learning Effective Dynamics across Spatio-Temporal Scales of Complex Flowscited by 2
  17. 2025A Design Co-Pilot for Task-Tailored Manipulatorscited by 2
  18. 2025Accelerating PDE-Constrained Optimization by the Derivative of Neural Operatorscited by 1
  19. 2025Hybrid Physics-Machine Learning Models for Quantitative Electron Diffraction Refinementscited by 1
  20. 2022Deep convolutional surrogates and degrees of freedom in thermal designcited by 0
  21. 2024A Straightforward Gradient-Based Approach for High-Tc Superconductor Design: Leveraging Domain Knowledge via Adaptive Constraintscited by 0
  22. 2024Adaptable Method for Crystal Design across Diverse Constraints and Objectives with Pretrained Property Predictorscited by 0
  23. 2025Accelerated Gradient-based Design Optimization Via Differentiable Physics-Informed Neural Operator: A Composites Autoclave Processing Case Studycited by 0
  24. 2025PLAID: A Unified Data Model for Machine Learning on Heterogeneous Physics Simulationscited by 0
  25. 2026Enabling Structure-Only Initialization and Out-of-Distribution Generalization in GNN-based Molecular Dynamics Simulatorscited by 0