Details of the Researcher

PHOTO

Kai Fukami
Section
Graduate School of Engineering
Job title
Associate Professor
Degree

  • Ｐｈ．Ｄ．（UCLA）

  • 修士（工学）（慶應義塾大学）

e-Rad No.
91013704

Research History 8

  • 2025/01 - Present
    Tohoku University　Department of Aerospace Engineering　Associate Professor

  • 2024/03 - 2024/12
    University of California, Los Angeles　Postdoctoral Research Associate

  • 2020/09 - 2024/02
    University of California, Los Angeles　Graduate Research Assistant (Ph.D. student)

  • 2020/09 - 2021/08
    Keio University　Visiting Researcher

  • 2018/04 - 2020/09
    Keio University　Graduate Research Assistant (Master student)

  • 2019/01 - 2019/08
    University of California, Los Angeles　Visiting Research Student

  • 2018/08 - 2018/12
    Florida State University　Visiting Research Student

  • 2017/04 - 2018/03
    Keio University　Undergraduate Research Assistant

Show all Show first 5

Education 3

  • University of California, Los Angeles　Samueli Engineering School　Mechanical and Aerospace Engineering

    2020/09 - 2024/02

  • Keio University　Graduate School of Science and Technology　Science for Open and Environmental System

    2018/04 - 2020/09

  • Keio University　Faculty of Science and Technology　Department of Mechanical Engineering

    2014/04 - 2018/03

Professional Memberships 5

  • Japan Society of Fluid Mechanics

  • American Institute of Aeronautics and Astronautics

  • Society for Industrial and Applied Mathematics

  • American Physics Society

  • Japan Society of Mechanical Engineers

Research Interests 7

  • Data-Oriented Approach

  • Machine Learning

  • Data Science

  • Aerodynamics

  • Turbulence

  • Fluid Mechanics

  • Fluid Dynamics

Awards 12

  1. 2024 Outstanding Reviewer of Journal of Fluid Mechanics

    2025/08　Journal of Fluid Mechanics, Cambridge University Press & Assessment　Journal of Fluid Mechanics Top 1% reviewer

  2. USNC/TAM National Academy of Sciences Travel Fellowship for ICTAM2024

    2024/06　National Academy of Sciences

  3. UCLA Mechanical and Aerospace Engineering Department Outstanding Ph.D. Degree Award

    2024/04　University of California, Los Angeles

  4. Amazon Fellow 2022

    2022/09　Science Hub for Humanity and Artificial Intelligence organized by UCLA and Amazon　Developing artificial-intelligent techniques for turbulence

  5. International Congress of Theoretical and Applied Mechanics Grant

    2021/08　ICTAM 2020+1

  6. Fujiwara award (Best Graduate Research Award at Keio Univ.)

    2021/03　Fujiwara Scholarship Foundation, Keio University　Space-time data recovery of fluid flows using machine learning based super resolution

  7. Certificate of Merit for Thermal Engineering Best Paper in 2nd Pacific Rim Thermal Engineering Conference

    2020/10　Japan Society of Mechanical Engineers Thermal Engineering Division　A machine-learned turbulence generator for the channel flow

  8. 17th Workshop on Turbulence Control

    2019/11　17th Workshop on Turbulence Control　Machine-learning-based super-resolution analysis for spatio-temporal data reconstruction of fluid flows

  9. JSME Fluids Engineering Excellent Presentation Certificate of Merit

    2019/11　Japan Society of Mechanical Engineers Fluids Engineering Division　Machine-learned super-resolution analysis of three dimensional turbulent channel flow

  10. Best Bachelor Thesis Presentation Award

    2018/03　Keio University　Proposal of a turbulence generator using machine learning

  11. Best Presentation Award

    2018/01　12th Workshop on Turbulence Control　Proposal of a turbulence generator using machine learning

  12. Outstanding Performance Award of Independent Studies in Mechanical Engineering

    2016/03　Keio University　SMART HOME with Twitter

Show all ︎Show 5

Papers 39

  1. Machine learning in fluid dynamics: A critical assessment Invited Peer-reviewed

    Kunihiko Taira, Georgios Rigas, Kai Fukami

    Physical Review Fluids　10　(9)　090701　2025/09/16

    Publisher: American Physical Society (APS)

    DOI： 10.1103/8t52-mtb9 　

    eISSN： 2469-990X

    More details Close

    The fluid dynamics community has increasingly adopted machine learning to analyze, model, predict, and control a wide range of flows. These methods offer powerful computational capabilities for regression, compression, and optimization. In some cases, machine learning has even outperformed traditional approaches. However, many fluid mechanics problems remain beyond the reach of current machine learning techniques. As the field moves from its current state toward a more mature paradigm, this article offers a critical assessment of the key challenges that must be addressed. Tackling these technical issues will not only deepen our understanding of flow physics but also expand the applicability of machine learning beyond fundamental research. We also highlight the importance of community-maintained datasets and open-source code repositories to accelerate progress in this area. Furthermore, the future success of machine learning in fluid dynamics will depend on effective training—not only for the next generation of researchers but also for established fluid mechanicians adapting to this evolving landscape. Data-driven fluid dynamics is in its critical transitional state over the next few years to shape its future. This perspective article aims to spark discussions and encourage collaborative efforts to advance the integration of machine learning in fluid dynamics.

  2. Extreme vortex-gust airfoil interactions at Reynolds number 5000 Peer-reviewed

    Kai Fukami, Luke Smith, Kunihiko Taira

    Physical Review Fluids　10　(084703)　2025/08/12

    DOI： 10.1103/vcbd-tvz1 　

  3. Information-Theoretic Machine Learning for Time-Varying Mode Decomposition of Separated Aerodynamic Flows Peer-reviewed

    Kai Fukami, Ryo Araki

    AIAA Journal　2025/08/05

    DOI： 10.2514/1.J065914 　

  4. Compressing fluid flows with nonlinear machine learning: mode decomposition, latent modeling, and flow control Invited Peer-reviewed

    Koji Fukagata, Kai Fukami

    Fluid Dynamics Research　57　(041401)　2025/06/26

    Publisher: IOP Publishing

    DOI： 10.1088/1873-7005/ade8a2 　

    ISSN： 0169-5983

    eISSN： 1873-7005

    More details Close

    Abstract An autoencoder is a self-supervised machine-learning network trained to output a quantity identical to the input. Owing to its structure possessing a bottleneck with a lower dimension, an autoencoder works to achieve data compression, extracting the essence of the high-dimensional data into the resulting latent space. We review the fundamentals of flow field compression using convolutional neural network-based autoencoder (CNN-AE) and its applications to various fluid dynamics problems. We cover the structure and the working principle of CNN-AE with an example of unsteady flows while examining the theoretical similarities between linear and nonlinear compression techniques. Representative applications of CNN-AE to various flow problems, such as mode decomposition, latent modeling, and flow control, are discussed. Throughout the present review, we show how the outcomes from the nonlinear machine-learning-based compression may support modeling and understanding a range of fluid mechanics problems.

  5. Observable-augmented manifold learning for multi-source turbulent flow data Peer-reviewed

    Journal of Fluid Mechanics　1010　(R4)　2025/05/09

    DOI： 10.1017/jfm.2025.383 　

  6. Single-snapshot machine learning for super-resolution of turbulence Peer-reviewed

    Kai Fukami, Kunihiko Taira

    Journal of Fluid Mechanics　1001　2024/12/12

    Publisher: Cambridge University Press (CUP)

    DOI： 10.1017/jfm.2024.1136 　

    ISSN： 0022-1120

    eISSN： 1469-7645

    More details Close

    Modern machine-learning techniques are generally considered data-hungry. However, this may not be the case for turbulence as each of its snapshots can hold more information than a single data file in general machine-learning settings. This study asks the question of whether nonlinear machine-learning techniques can effectively extract physical insights even from as little as a single snapshot of turbulent flow. As an example, we consider machine-learning-based super-resolution analysis that reconstructs a high-resolution field from low-resolution data for two examples of two-dimensional isotropic turbulence and three-dimensional turbulent channel flow. First, we reveal that a carefully designed machine-learning model trained with flow tiles sampled from only a single snapshot can reconstruct vortical structures across a range of Reynolds numbers for two-dimensional decaying turbulence. Successful flow reconstruction indicates that nonlinear machine-learning techniques can leverage scale-invariance properties to learn turbulent flows. We also show that training data of turbulent flows can be cleverly collected from a single snapshot by considering characteristics of rotation and shear tensors. Second, we perform the single-snapshot super-resolution analysis for turbulent channel flow, showing that it is possible to extract physical insights from a single flow snapshot even with inhomogeneity. The present findings suggest that embedding prior knowledge in designing a model and collecting data is important for a range of data-driven analyses for turbulent flows. More broadly, this work hopes to stop machine-learning practitioners from being wasteful with turbulent flow data.

  7. Aerodynamics-guided machine learning for design optimization of electric vehicles Peer-reviewed

    Jonathan Tran, Kai Fukami, Kenta Inada, Daisuke Umehara, Yoshimichi Ono, Kenta Ogawa, Kunihiko Taira

    Communications Engineering　3　(174)　2024/11/20

    DOI： 10.1038/s44172-024-00322-0 　

  8. Data-driven transient lift attenuation for extreme vortex gust–airfoil interactions Peer-reviewed

    Kai Fukami, Hiroya Nakao, Kunihiko Taira

    Journal of Fluid Mechanics　992　2024/08/10

    Publisher: Cambridge University Press (CUP)

    DOI： 10.1017/jfm.2024.592 　

    ISSN： 0022-1120

    eISSN： 1469-7645

    More details Close

    We present a data-driven feedforward control to attenuate large transient lift experienced by an airfoil disturbed by an extreme level of discrete vortex gust. The current analysis uses a nonlinear machine-learning technique to compress the high-dimensional flow dynamics onto a low-dimensional manifold. While the interaction dynamics between the airfoil and extreme vortex gust are parametrized by its size, gust ratio and position, the wake responses are well captured on this simple manifold. The effect of extreme vortex disturbance about the undisturbed baseline flows can be extracted in a physically interpretable manner. Furthermore, we call on phase-amplitude reduction to model and control the complex nonlinear extreme aerodynamic flows. The present phase-amplitude reduction model reveals the sensitivity of the dynamical system in terms of the phase shift and amplitude change induced by external forcing with respect to the baseline periodic orbit. By performing the phase-amplitude analysis for a latent dynamical model identified by sparse regression, the sensitivity functions of low-dimensionalized aerodynamic flows for both phase and amplitude are derived. With the phase and amplitude sensitivity functions, optimal forcing can be determined to quickly suppress the effect of extreme vortex gusts towards the undisturbed states in a low-order space. The present optimal flow modification built upon the machine-learned low-dimensional subspace quickly alleviates the impact of transient vortex gusts for a variety of extreme aerodynamic scenarios, providing a potential foundation for flight of small-scale air vehicles in adverse atmospheric conditions.

  9. Phase autoencoder for limit-cycle oscillators Peer-reviewed

    Koichiro Yawata, Kai Fukami, Kunihiko Taira, Hiroya Nakao

    Chaos: An Interdisciplinary Journal of Nonlinear Science　2024/06/01

    DOI： 10.1063/5.0205718 　

  10. Data-driven nonlinear turbulent flow scaling with Buckingham Pi variables Peer-reviewed

    Kai Fukami, Susumu Goto, Kunihiko Taira

    Journal of Fluid Mechanics　984　(R4)　2024/04/10

    DOI： 10.1017/jfm.2024.211 　

  11. Reconstructing Three-Dimensional Bluff Body Wake from Sectional Flow Fields with Convolutional Neural Networks Peer-reviewed

    Mitsuaki Matsuo, Kai Fukami, Taichi Nakamura, Masaki Morimoto, Koji Fukagata

    SN Computer Science　5　(3)　2024/03/07

    Publisher: Springer Science and Business Media LLC

    DOI： 10.1007/s42979-024-02602-0 　

    eISSN： 2661-8907

  12. A cyclic perspective on transient gust encounters through the lens of persistent homology Peer-reviewed

    Luke Smith, Kai Fukami, Girguis Sedky, Anya Jones, Kunihiko Taira

    Journal of Fluid Mechanics　980　2024/01/30

    Publisher: Cambridge University Press (CUP)

    DOI： 10.1017/jfm.2024.16 　

    ISSN： 0022-1120

    eISSN： 1469-7645

    More details Close

    Large-amplitude gust encounters exhibit a range of separated flow phenomena, making them difficult to characterize using the traditional tools of aerodynamics. In this work, we propose a dynamical systems approach to gust encounters, viewing the flow as a cycle (or a closed trajectory) in state space. We posit that the topology of this cycle, or its shape and structure, provides a compact description of the flow, and can be used to identify coordinates in which the dynamics evolve in a simple, intuitive way. To demonstrate this idea, we consider flowfield measurements of a transverse gust encounter. For each case in the dataset, we characterize the full-state dynamics of the flow using persistent homology, a tool that identifies holes in point cloud data, and transform the dynamics to a reduced-order space using a nonlinear autoencoder. Critically, we constrain the autoencoder such that it preserves topologically relevant features of the original dynamics, or those features identified by persistent homology. Using this approach, we are able to transform six separate gust encounters to a three-dimensional latent space, in which each gust encounter reduces to a simple circle, and from which the original flow can be reconstructed. This result shows that topology can guide the creation of low-dimensional state representations for strong transverse gust encounters, a crucial step towards the modelling and control of aerofoil–gust interactions.

  13. Sparse Pressure-Based Machine Learning Approach for Aerodynamic Loads Estimation During Gust Encounters Peer-reviewed

    Dashuai Chen, Frieder Kaiser, JiaCheng Hu, David E. Rival, Kai Fukami, Kunihiko Taira

    AIAA Journal　2024/01

    DOI： 10.2514/1.J063263 　

  14. Grasping extreme aerodynamics on a low-dimensional manifold Peer-reviewed

    Kai Fukami, Kunihiko Taira

    Nature Communications　14　6480　2023/10/14

    DOI： 10.1038/s41467-023-42213-6 　

  15. Super-resolution analysis via machine learning: a survey for fluid flows Invited Peer-reviewed

    Kai Fukami, Koji Fukagata, Kunihiko Taira

    Theoretical and Computational Fluid Dynamics　37　(4)　421-444　2023/08

    DOI： 10.1007/s00162-023-00663-0 　

    ISSN： 0935-4964

    eISSN： 1432-2250

  16. Sparse sensor reconstruction of vortex-impinged airfoil wake with machine learning Peer-reviewed

    Yonghong Zhong, Kai Fukami, Byungjin An, Kunihiko Taira

    Theoretical and Computational Fluid Dynamics　37　(2)　269-287　2023/04

    DOI： 10.1007/s00162-023-00657-y 　

    ISSN： 0935-4964

    eISSN： 1432-2250

  17. Image and video compression of fluid flow data Peer-reviewed

    Vishal Anatharaman, Jason Feldkamp, Kai Fukami, Kunihiko Taira

    Theoretical and Computational Fluid Dynamics　37　(1)　61-82　2023/02

    DOI： 10.1007/s00162-023-00643-4 　

    ISSN： 0935-4964

    eISSN： 1432-2250

  18. Machine-Learning-Based Reconstruction of Turbulent Vortices From Sparse Pressure Sensors in a Pump Sump Peer-reviewed

    Kai Fukami, Byungjin An, Motohiko Nohmi, Masashi Obuchi, Kunihiko Taira

    Journal of Fluids Engineering　144　(12)　2022/12

    DOI： 10.1115/1.4055178 　

    ISSN： 0098-2202

    eISSN： 1528-901X

  19. Assessments of epistemic uncertainty using Gaussian stochastic weight averaging for fluid-flow regression Peer-reviewed

    Masaki Morimoto, Kai Fukami, Romit Maulik, Ricardo Vinuesa, Koji Fukagata

    Physica D, Nonlinear Phenomena　440　2022/11

    DOI： 10.1016/j.physd.2022.133454 　

    ISSN： 0167-2789

    eISSN： 1872-8022

  20. Generalization techniques of neural networks for fluid flow estimation Peer-reviewed

    Masaki Morimoto, Kai Fukami, Kai Zhang, Koji Fukagata

    Neural Computing & Applications　34　(5)　3647-3669　2022/03

    DOI： 10.1007/s00521-021-06633-z 　

    ISSN： 0941-0643

    eISSN： 1433-3058

  21. Identifying key differences between linear stochastic estimation and neural networks for fluid flow regressions Peer-reviewed

    Taichi Nakamura, Kai Fukami, Koji Fukagata

    Scientific Reports　12　(1)　2022/03

    DOI： 10.1038/s41598-022-07515-7 　

    ISSN： 2045-2322

  22. Global field reconstruction from sparse sensors with Voronoi tessellation-assisted deep learning Peer-reviewed

    Kai Fukami, Romit Maulik, Nesar Ramachandra, Koji Fukagata, Kunihiko Taira

    Nature Machine Intelligence　3　(11)　945-+　2021/11

    DOI： 10.1038/s42256-021-00402-2 　

    eISSN： 2522-5839

  23. Convolutional neural networks for fluid flow analysis: toward effective metamodeling and low dimensionalization Peer-reviewed

    Masaki Morimoto, Kai Fukami, Kai Zhang, Aditya G. Nair, Koji Fukagata

    Theoretical and Computational Fluid Dynamics　35　(5)　633-658　2021/10

    DOI： 10.1007/s00162-021-00580-0 　

    ISSN： 0935-4964

    eISSN： 1432-2250

  24. Sparse identification of nonlinear dynamics with low-dimensionalized flow representations Peer-reviewed

    Kai Fukami, Takaaki Murata, Kai Zhang, Koji Fukagata

    Journal of Fluid Mechanics　926　2021/09

    DOI： 10.1017/jfm.2021.697 　

    ISSN： 0022-1120

    eISSN： 1469-7645

  25. Experimental velocity data estimation for imperfect particle images using machine learning Peer-reviewed

    Masaki Morimoto, Kai Fukami, Koji Fukagata

    Physics of Fluids　33　(8)　2021/08

    DOI： 10.1063/5.0060760 　

    ISSN： 1070-6631

    eISSN： 1089-7666

  26. Machine-learning-based spatio-temporal super resolution reconstruction of turbulent flows Peer-reviewed

    Kai Fukami, Koji Fukagata, Kunihiko Taira

    Journal of Fluid Mechanics　909　2021/02

    DOI： 10.1017/jfm.2020.948 　

    ISSN： 0022-1120

    eISSN： 1469-7645

  27. Convolutional neural network and long short-term memory based reduced order surrogate for minimal turbulent channel flow Peer-reviewed

    Taichi Nakamura, Kai Fukami, Kazuto Hasegawa, Yusuke Nabae, Koji Fukagata

    Physics of Fluids　33　(2)　2021/02

    DOI： 10.1063/5.0039845 　

    ISSN： 1070-6631

    eISSN： 1089-7666

  28. CNN-LSTM based reduced order modeling of two-dimensional unsteady flows around a circular cylinder at different Reynolds numbers Peer-reviewed

    Kazuto Hasegawa, Kai Fukami, Takaaki Murata, Koji Fukagata

    Fluid Dynamics Research　52　(6)　2020/12

    DOI： 10.1088/1873-7005/abb91d 　

    ISSN： 0169-5983

    eISSN： 1873-7005

  29. Probabilistic neural networks for fluid flow surrogate modeling and data recovery Peer-reviewed

    Romit Maulik, Kai Fukami, Nesar Ramachandra, Koji Fukagata, Kunihiko Taira

    Physical Review Fluids　5　(10)　2020/10

    DOI： 10.1103/PhysRevFluids.5.104401 　

    ISSN： 2469-990X

  30. Convolutional neural network based hierarchical autoencoder for nonlinear mode decomposition of fluid field data Peer-reviewed

    Kai Fukami, Taichi Nakamura, Koji Fukagata

    Physics of Fluids　32　(9)　2020/09

    DOI： 10.1063/5.0020721 　

    ISSN： 1070-6631

    eISSN： 1089-7666

  31. Assessment of supervised machine learning methods for fluid flows Invited Peer-reviewed

    Kai Fukami, Koji Fukagata, Kunihiko Taira

    Theoretical and Computational Fluid Dynamics　34　(4)　497-519　2020/08

    DOI： 10.1007/s00162-020-00518-y 　

    ISSN： 0935-4964

    eISSN： 1432-2250

  32. Machine-learning-based reduced-order modeling for unsteady flows around bluff bodies of various shapes Invited Peer-reviewed

    Kazuto Hasegawa, Kai Fukami, Takaaki Murata, Koji Fukagata

    Theoretical and Computational Fluid Dynamics　34　(4)　367-383　2020/08

    DOI： 10.1007/s00162-020-00528-w 　

    ISSN： 0935-4964

    eISSN： 1432-2250

  33. Nonlinear mode decomposition with convolutional neural networks for fluid dynamics Peer-reviewed

    Takaaki Murata, Kai Fukami, Koji Fukagata

    Journal of Fluid Mechanics　882　2020/01

    DOI： 10.1017/jfm.2019.822 　

    ISSN： 0022-1120

    eISSN： 1469-7645

  34. Super-resolution reconstruction of turbulent flows with machine learning Peer-reviewed

    Kai Fukami, Koji Fukagata, Kunihiko Taira

    JOURNAL OF FLUID MECHANICS　870　106-120　2019/07

    DOI： 10.1017/jfm.2019.238 　

    ISSN： 0022-1120

    eISSN： 1469-7645

  35. Synthetic turbulent inflow generator using machine learning Peer-reviewed

    Kai Fukami, Yusuke Nabae, Ken Kawai, Koji Fukagata

    Physical Review Fluids　4　(6)　2019/06

    DOI： 10.1103/PhysRevFluids.4.064603 　

    ISSN： 2469-990X

  36. Super-resolution analysis with machine learning for low-resolution flow data

    Kai Fukami, Koji Fukagata, Kunihiko Taira

    11th International Symposium on Turbulence and Shear Flow Phenomena, TSFP 2019　2019

    Publisher: International Symposium on Turbulence and Shear Flow Phenomena, TSFP

  37. Machine-learned super-resolution analysis of three-dimensional turbulent channel flow

    Kai Fukami

    The Proceedings of the Fluids engineering conference　2019

    DOI： 10.1299/jsmefed.2019.os8-01 　

    ISSN： 2424-2896

  38. DATA-DRIVEN REDUCED ORDER MODELING OF FLOWS AROUND TWO-DIMENSIONAL BLUFF BODIES OF VARIOUS SHAPES

    Kazuto Hasegawa, Kai Fukami, Takaaki Murata, Koji Fukagata

    PROCEEDINGS OF THE ASME/JSME/KSME JOINT FLUIDS ENGINEERING CONFERENCE, 2019, VOL 2　2019

    ISSN： 2333-2506

  39. CNN-SINDY BASED REDUCED ORDER MODELING OF UNSTEADY FLOW FIELDS

    Takaaki Murata, Kai Fukami, Koji Fukagata

    PROCEEDINGS OF THE ASME/JSME/KSME JOINT FLUIDS ENGINEERING CONFERENCE, 2019, VOL 2　2019

    ISSN： 2333-2506

Show all ︎Show first 5

Misc. 34

  1. Super-resolution analysis of turbulence with machine learning Invited

    Kai Fukami, Koji Fukagata, Kunihiko Taira

    Nagare-Journal of Japan Society of Fluid Mechanics　44　(3)　216-221　2025/09/24

  2. Plunging Airfoil Wakes in Low-Order Latent Space Coordinates

    Hiroto Odaka, Kai Fukami, Kunihiko Taira

    AIAA AVIATION FORUM AND ASCEND 2025　2025-3869　2025/07/16

    Publisher: American Institute of Aeronautics and Astronautics

    DOI： 10.2514/6.2025-3869 　

  3. Data-Driven Modeling, Sensing, and Control of Extreme Vortex-Airfoil Interactions

    Kai Fukami, Kunihiko Taira

    AIAA AVIATION Forum and Ascend 2024　2024-4531　2024/07/27

    DOI： 10.2514/6.2024-4531 　

  4. Extreme aerodynamics of vortex impingement: Machine-learning-based compression and situational awareness Peer-reviewed

    Kai Fukami, Kunihiko Taira

    13th International Symposium on Turbulence and Shear Flow Phenomena (TSFP13)　114　2024/06

  5. Learning the nonlinear manifold of extreme aerodynamics Peer-reviewed

    Kai Fukami, Kunihiko Taira

    NeurIPS 2022　105　2022/12

  6. Machine-learning-based reconstruction of transient vortex-airfoil wake interaction

    Yonghong Zhong, Kai Fukami, Byungjin An, Kunihiko Taira

    AIAA AVIATION 2022 Forum　2022/06/20

    Publisher: American Institute of Aeronautics and Astronautics

    DOI： 10.2514/6.2022-3244 　

  7. Machine-learning-based turbulent state estimation from pressure sensors in a pump sump

    深見開, 安炳辰, 能見基彦, 大渕真志, 平邦彦

    ターボ機械協会講演会(CD-ROM)　86th　2022

  8. Visualization of nonlinear modal structures for three-dimensional unsteady fluid flows with customized decoder design Peer-reviewed

    Kazuto Hasegawa, Kai Fukami, Koji Fukagata

    NeurIPS 2021　30　2021/12

  9. Inserting machine-learned virtual wall velocity for large-eddy simulation of turbulent channel flows

    Naoki Moriya, Kai Fukami, Yusuke Nabae, Masaki Morimoto, Taichi Nakamura, Koji Fukagata

    2021/06/17

    More details Close

    We propose a supervised-machine-learning-based wall model for coarse-grid wall-resolved large-eddy simulation (LES). Our consideration is made on LES of turbulent channel flows with a first grid point set relatively far from the wall ($\sim$ 10 wall units), while still resolving the near-wall region, to present a new path to save the computational cost. Convolutional neural network (CNN) is utilized to estimate a virtual wall-surface velocity from $x-z$ sectional fields near the wall, whose training data are generated by a direct numerical simulation (DNS) at ${\rm Re}_{\tau}=180$. The virtual wall-surface velocity is prepared with the extrapolation of the DNS data near the wall. This idea enables us to give a proper wall condition to correct a velocity gradient near the wall. The estimation ability of the model from near wall information is first investigated as a priori test. The estimated velocity fields by the present CNN model are in statistical agreement with the reference DNS data. The model trained in a priori test is then combined with the LES as a posteriori test. We find that the LES can successfully be augmented using the present model at both the friction Reynolds number ${\rm Re}_{\tau}=180$ used for training and the unseen Reynolds number ${\rm Re}_{\tau}=360$ even when the first grid point is located at 5 wall units off the wall. We also investigate the robustness of the present model for the choice of sub-grid scale model and the possibility of transfer learning in a local domain. The observations through the paper suggest that the present model is a promising tool for recovering the accuracy of LES with a coarse grid near the wall.

  10. Clues for noise robustness of state estimation: Error-curve quest of neural network and linear regression Peer-reviewed

    Taichi Nakamura, Kai Fukami, Koji Fukagata

    ICLR 2021　3　2021/05

  11. Supervised convolutional networks for volumetric data enrichment from limited sectional data with adaptive super resolution Peer-reviewed

    Mitsuaki Matsuo, Kai Fukami, Taichi Nakamura, Masaki Morimoto, Koji Fukagata

    ICLR 2021　4　2021/05

  12. Supervised convolutional network for three-dimensional fluid data reconstruction from sectional flow fields with adaptive super-resolution assistance

    Mitsuaki Matsuo, Taichi Nakamura, Masaki Morimoto, Kai Fukami, Koji Fukagata

    2021/03/16

    More details Close

    The recent development of high-performance computing enables us to generate spatio-temporal high-resolution data of nonlinear dynamical systems and to analyze them for deeper understanding of their complex nature. This trend can be found in a wide range of science and engineering communities, which suggests that detailed investigations on efficient data handling in physical science must be required in future. To this end, we introduce the use of convolutional neural networks (CNNs) to achieve an efficient data storage and estimation of scientific big data derived from nonlinear dynamical systems. The CNN is utilized to reconstruct three-dimensional data from a few numbers of two-dimensional sections in a computationally friendly manner. The present model is a combination of two- and three-dimensional CNNs, which allows users to save only some of the two-dimensional sections to reconstruct the volumetric data. As an example of three-dimensional data, we consider a fluid flow around a square cylinder at the diameter-based Reynolds number $Re_D$ of 300, and show that volumetric fluid flow data can successfully be reconstructed with the present method from as few as five sections. Furthermore, we also propose a combination of the present CNN-based reconstruction with an adaptive sampling-based super-resolution analysis to augment the data compression capability of the present methods. Our report can be a significant bridge toward practical data handling for not only the fluid mechanics field but also a vast range of physical sciences.

  13. Neural network-based anomaly detections for nonlinear dynamical systems

    森本将生, 深見開, 深見開, 中村太一, 深潟康二

    日本機械学会関東支部総会・講演会講演論文集(CD-ROM)　27th　2021

    ISSN： 2424-2691

  14. Machine learning based state estimation of turbulent flows: robustness for noisy input

    中村太一, 深見開, 深見開, 深潟康二

    日本機械学会関東支部総会・講演会講演論文集(CD-ROM)　27th　2021

    ISSN： 2424-2691

  15. Towards an Innovative Flow Control with Machine Learning-Based Reduced-Order Modeling

    深潟康二, 深見開

    伝熱　60　(253)　2021

    ISSN： 1344-8692

  16. Convolutional neural network-based global field recovery from sparse sensors of transitional boundary layer flow

    中村太一, 深見開, 深潟康二

    数値流体力学シンポジウム講演論文集(CD-ROM)　35th　2021

    ISSN： 2433-2674

  17. Fluid flow state estimation from sparse sensor measurements using convolutional neural network

    中村太一, 深見開, 深潟康二

    日本機械学会年次大会講演論文集(CD-ROM)　2021　2021

    ISSN： 2424-2667

  18. Probabilistic neural network-based reduced-order surrogate for fluid flows Peer-reviewed

    Kai Fukami, Romit Maulik, Nesar Ramachandra, Koji Fukagata, Kunihiko Taira

    NeurIPS 2020　7　2020/12/16

    More details Close

    In recent years, there have been a surge in applications of neural networks (NNs) in physical sciences. Although various algorithmic advances have been proposed, there are, thus far, limited number of studies that assess the interpretability of neural networks. This has contributed to the hasty characterization of most NN methods as "black boxes" and hindering wider acceptance of more powerful machine learning algorithms for physics. In an effort to address such issues in fluid flow modeling, we use a probabilistic neural network (PNN) that provide confidence intervals for its predictions in a computationally effective manner. The model is first assessed considering the estimation of proper orthogonal decomposition (POD) coefficients from local sensor measurements of solution of the shallow water equation. We find that the present model outperforms a well-known linear method with regard to estimation. This model is then applied to the estimation of the temporal evolution of POD coefficients with considering the wake of a NACA0012 airfoil with a Gurney flap and the NOAA sea surface temperature. The present model can accurately estimate the POD coefficients over time in addition to providing confidence intervals thereby quantifying the uncertainty in the output given a particular training data set.

  19. Model order reduction with neural networks: Application to laminar and turbulent flows

    Kai Fukami, Kazuto Hasegawa, Taichi Nakamura, Masaki Morimoto, Koji Fukagata

    2020/11/20

    DOI： 10.1007/s42979-021-00867-3 　

    More details Close

    We investigate the capability of neural network-based model order reduction, i.e., autoencoder (AE), for fluid flows. As an example model, an AE which comprises of a convolutional neural network and multi-layer perceptrons is considered in this study. The AE model is assessed with four canonical fluid flows, namely: (1) two-dimensional cylinder wake, (2) its transient process, (3) NOAA sea surface temperature, and (4) $y-z$ sectional field of turbulent channel flow, in terms of a number of latent modes, a choice of nonlinear activation functions, and a number of weights contained in the AE model. We find that the AE models are sensitive against the choice of the aforementioned parameters depending on the target flows. Finally, we foresee the extensional applications and perspectives of machine learning based order reduction for numerical and experimental studies in fluid dynamics community.

  20. Improvement of PIV by data augmentation based on machine learning

    森本将生, 深見開, 長谷川一登, 村田高彬, 村上光, 深潟康二

    ながれ　39　(2)　2020

    ISSN： 0286-3154

  21. Toward Turbulence Big Data Analysis Using Machine Learning

    深潟康二, 深見開

    計測と制御　59　(8)　2020

    ISSN： 0453-4662

  22. Extraction of nonlinear modes in fluid flows using a hierarchical convolutional neural network autoencoder

    中村太一, 深見開, 深潟康二

    ながれ　39　(6)　2020

    ISSN： 0286-3154

  23. Supervised machine learning for wall-modeling in large-eddy simulation of turbulent channel flow

    守矢直樹, 深見開, 難波江佑介, 森本将生, 中村太一, 深潟康二

    数値流体力学シンポジウム講演論文集(CD-ROM)　34th　2020

    ISSN： 2433-2674

  24. Toward practical machine learning and fluid flow regressions: perspecitve on interpretability and generalizability

    森本将生, 深見開, ZHANG Kai, 深潟康二

    数値流体力学シンポジウム講演論文集(CD-ROM)　34th　2020

    ISSN： 2433-2674

  25. Machine learning-aided state estimation in a turbulent channel flow and its robustness for sensor information

    中村太一, 深見開, 深潟康二

    数値流体力学シンポジウム講演論文集(CD-ROM)　34th　2020

    ISSN： 2433-2674

  26. Three-dimensional flow field reconstruction from two-dimensional sectional data using machine learning

    松尾光昭, 森本将生, 中村太一, 深見開, 深潟康二

    数値流体力学シンポジウム講演論文集(CD-ROM)　34th　2020

    ISSN： 2433-2674

  27. 注目研究in CFD32 機械学習を用いた円柱周り流れのレイノルズ数依存性の予測

    長谷川一登, 深見開, 村田高彬, 深潟康二

    ながれ　38　(2)　2019

    ISSN： 0286-3154

  28. 機械学習に基づくデータ拡張によるPIVの精度向上

    森本将生, 深見開, 長谷川一登, 村田高彬, 村上光, 深潟康二

    数値流体力学シンポジウム講演論文集(CD-ROM)　33rd　2019

    ISSN： 2433-2674

  29. オートエンコーダを用いたチャネル乱流の機械学習

    中村太一, 深見開, 長谷川一登, 村田高彬, 難波江佑介, 深潟康二

    数値流体力学シンポジウム講演論文集(CD-ROM)　33rd　2019

    ISSN： 2433-2674

  30. Applications of a machine-learned super-resolution algorithm to two-dimensional flow fields

    深見開, 深潟康二, 平邦彦

    ながれ　38　(6)　2019

    ISSN： 0286-3154

  31. Machine-learned super-resolution analysis of three-dimensional turbulent channel flow

    深見開, 深潟康二, 平邦彦

    日本機械学会流体工学部門講演会講演論文集(CD-ROM)　97th　2019

    ISSN： 1348-2882

  32. 機械学習を用いた円柱周り流れにおける低次元モードの抽出と時間発展予測

    村田高彬, 深見開, 深潟康二

    数値流体力学シンポジウム講演論文集(CD-ROM)　32nd　2018

    ISSN： 2433-2674

  33. 機械学習を用いた円柱周り流れのレイノルズ数依存性の予測

    長谷川一登, 深見開, 村田高彬, 深潟康二

    数値流体力学シンポジウム講演論文集(CD-ROM)　32nd　2018

    ISSN： 2433-2674

  34. 機械学習を用いた乱流生成器の提案

    深見開, 河合謙, 深潟康二

    日本機械学会関東支部総会・講演会講演論文集(CD-ROM)　24th　2018

    ISSN： 2424-2691

Show all ︎Show first 5

Presentations 130

  1. Data-driven analysis of extremely gusty aerodynamic flows Invited

    Kai Fukami

    The JSME-KSME Joint Symposium on Computational & CAE 2025　2025/08/22

  2. Plunging airfoil wakes in low-order latent space coordinates

    Hiroto Odaka, Kai Fukami, Kunihiko Taira

    AIAA Aviation Forum 2025　2025/07/25

  3. Observable-augmented manifold learning for unsteady flow analysis Invited

    Kai Fukami

    IUTAM Symposium on Machine Learning in Diverse Fluid Mechanics　2025/05/16

  4. Identifying interpolatory and extrapolatory vortical structures of data-driven fluid dynamics Invited

    Kai Fukami

    3rd Workshop on Data-Driven Fluid Dynamics　2025/03/18

  5. Data-driven analysis of highly unsteady flows: progress and outlook

    Kai Fukami

    The seminar at Spacecraft Thermal and Fluids Systems Laboratory, Tohoku University　2025/02/26

  6. Extreme Aerodynamic Manifold: Data-Driven Modeling and Control of Highly Gusty Flows Invited

    Kai Fukami

    18th CCMR symposium at Toyo University　2025/02/17

  7. Generalized Super-Resolution Analysis with Machine Learning of Turbulence Invited

    Kai Fukami

    1st Workshop for Digital Twin and AI-Integrated Design for Mechanical Systems　2025/02/13

  8. Taming highly unsteady flows with data-oriented approaches: progress and outlook Invited

    Kai Fukami

    Interdisciplinary Scientific Computing Laboratory (ISCL) Seminar Series at Pennsylvania State University　2025/01/17

  9. Quick mitigation of extreme-gust effects with phase-amplitude modeling on a low-dimensional manifold

    Kai Fukami, Hiroya Nakao, Kunihiko Taira

    38th CFD symposium　2024/12/11

  10. Data-driven automotive aerodynamic shape optimization

    Jonathan Tran, Kai Fukami, Kenta Inada, Daisuke Umehara, Yoshimichi Ono, Kenta Ogawa, Kunihiko Taira

    77th Annual Meeting of the APS Division of Fluid Dynamics　2024/11/24

  11. Single-snapshot machine learning for super-resolution analysis of turbulence

    Kai Fukami, Kunihiko Taira

    77th Annual Meeting of the APS Division of Fluid Dynamics　2024/11/24

  12. Data-driven vehicle design optimization through aerodynamics informed dimensionality reduction

    Jonathan Tran, Kai Fukami, Kenta Inada, Daisuke Umehara, Yoshimichi Ono, Kenta Ogawa, Kunihiko Taira

    SIAM Conference on Mathematics of Data Science (MDS24)　2024/10/22

  13. Latent space representation of plunging airfoil wakes using a drag-augmented autoencoder

    Hiroto Odaka, Kai Fukami, Kunihiko Taira

    1st European Fluid Dynamics Conference (EFDC1)　2024/09/16

  14. Data-driven lift regulation of extreme vortex-airfoil interactions

    Kai Fukami, Hiroya Nakao, Kunihiko Taira

    ICTAM2024　2024/08/27

  15. Data-driven modeling, sensing, and control of extreme vortex-airfoil interactions

    Kai Fukami, Kunihiko Taira

    AIAA Aviation Forum 2024　2024/08/02

  16. Extreme aerodynamics of vortex impingement: Machine-learning-based compression and situational awareness

    Kai Fukami, Kunihiko Taira

    13th International Symposium on Turbulence and Shear Flow Phenomena (TSFP13)　2024/06/26

  17. Super-Resolution Analysis: Revisiting the Training Process/Data for Machine Learning in Fluid Dynamics Invited

    Kunihiko Taira, Kai Fukami

    Advancing fluid and soft-matter dynamics with machine learning and data science: a conference at UW-Madison　2024/06/03

  18. Let us machine-learn fluid dynamics! Invited

    Kai Fukami

    73rd SCJSF & JABA Forum　2024/05/25

  19. Discrete gust encounters through the lens of persistent homology

    Luke R. Smith, Kai Fukami, Girguis Sedky, Anya R. Jones, Kunihiko Taira

    3rd Colloquium on Vortex Dominated Flows (DisCoVor)　2024/04/19

  20. Aerodynamics-informed manifold learning for data-driven design optimization of automobiles

    Jonathan Tran, Kai Fukami, Kunihiko Taira

    17th Southern California Flow Physics Symposium (SoCal Fluids XVII)　2024/04/13

  21. Phase-amplitude model-based control of extreme vortex-airfoil interactions on a low-dimensional manifold

    Kai Fukami, Hiroya Nakao, Kunihiko Taira

    17th Southern California Flow Physics Symposium (SoCal Fluids XVII)　2024/04/13

  22. Taming extreme aerodynamic flows with generalized super resolution and manifold identification Invited

    Kai Fukami

    Online webinar at the Laboratoire de Mécanique des Fluides de Lille (LFML)　2024/02/22

  23. Extreme Aerodynamic Manifold: Vortex-Airfoil Interactions Invited

    Kai Fukami, Kunihiko Taira

    Remote Colloquium on Vortex Dominated Flows (ReCoVor)　2024/01/19

  24. U.S. Ph.D. life as a Japanese Mechanical Engineer, Invited

    Kai Fukami

    Seminar for the Japanese Graduate Student Association in the United States　2023/12/16

  25. Extreme Aerodynamic Manifold: Vortex-Airfoil Interactions

    Kai Fukami, Kunihiko Taira

    76th Annual Meeting of the APS Division of Fluid Dynamics　2023/11/20

  26. Analyzing the Dynamics of Discrete Gust Encounters with Persistent Homology

    Luke R. Smith, Kai Fukami, Girguis Sedky, Anya R. Jones, Kunihiko Taira

    76th Annual Meeting of the APS Division of Fluid Dynamics　2023/11/20

  27. Data-driven compression of plunging airfoil wakes

    Hiroto Odaka, Kai Fukami, Kunihiko Taira

    76th Annual Meeting of the APS Division of Fluid Dynamics　2023/11/19

  28. Grasping extreme aerodynamics on a low-dimensional manifold Invited

    Kai Fukami, Kunihiko Taira

    Science Hub Showcase 2023 hosted by the UCLA-Amazon Science Hub for Humanity and Artificial Intelligence　2023/10/13

  29. Let us machine-learn fluid dynamics: A perspective of global field reconstruction and nonlinear manifold identification

    Kai Fukami

    Seminar at Aerodynamic Design Research Group at Tohoku University　2023/08/24

  30. Global field reconstruction from sparse sensors with Voronoi tessellation-assisted deep learning

    Kai Fukami, Romit Maulik, Nesar Ramachandra, Koji Fukagata, Kunihiko Taira

    Seminar at Structures-Computer Interaction Lab at UCLA　2023/06/07

  31. Discovering the nonlinear manifold of extreme aerodynamic flows

    Kai Fukami, Kunihiko Taira

    16th Southern California Flow Physics Symposium (SoCal Fluids XVI)　2023/04/22

  32. Feature extraction from plunging airfoil wakes using an autoencoder,

    Hiroto Odaka, Kai Fukami, Kunihiko Taira

    16th Southern California Flow Physics Symposium (SoCal Fluids XVI)　2023/04/22

  33. Super-resolving turbulent flows with machine learning: a survey Invited

    Kai Fukami, Koji Fukagata, Kunihiko Taira

    SIAM Conference on Computational Science and Engineering (CSE23)　2023/02/27

  34. Developing artificial-intelligent techniques for turbulence Invited

    Kai Fukami

    Lightning Talks by the Amazon Fellows hosted by the UCLA-Amazon Science Hub for Humanity and Artificial Intelligence　2023/02/23

  35. Machine learning for fluid dynamics -- Part I: Unsupervised learning Invited

    Kunihiko Taira, Kai Fukami

    Seminar at Honda Motor Co.,LTD.　2022/12/15

  36. Machine learning for fluid dynamics -- Part II: Supervised learning Invited

    Kunihiko Taira, Kai Fukami

    Seminar at Honda Motor Co.,LTD.　2022/12/15

  37. Machine learning for fluid dynamics -- Part III: Applications Invited

    Kai Fukami, Kunihiko Taira

    Seminar at Honda Motor Co.,LTD.　2022/12/15

  38. Learning the nonlinear manifold of extreme aerodynamics

    Kai Fukami, Kunihiko Taira

    Machine Learning and the Physical Sciences, Workshop at the 36th Conference on Neural Information Processing Systems (NeurIPS)　2022/12/01

  39. Image and video compression of fluid flow data

    Vishal Anantharaman, Kai Fukami, Kunihiko Taira

    75th Annual Meeting of the APS Division of Fluid Dynamics　2022/11/20

  40. Compact manifold representation of airfoil wake-vortex gust interaction

    Kai Fukami, Kunihiko Taira

    75th Annual Meeting of the APS Division of Fluid Dynamics　2022/11/20

  41. Finding scale-invariant turbulent flow structures for enhanced machine learning

    Kai Fukami, Kunihiko Taira

    SIAM Conference on Mathematics of Data Science (MDS22)　2022/09/26

  42. Physics-inspired machine learning for fluid flow reconstruction and reduced-complexity modeling Invited

    Kai Fukami

    Seminar at Osaka University　2022/09/13

  43. Towards phase-inspired airfoil wake control in autoencoder latent space

    Kai Fukami

    Seminar at Tokyo Institute of Technology　2022/09/12

  44. Reconstructing and modeling unsteady flows with physics-inspired machine learning Invited

    Kai Fukami

    2nd US-Japan Workshop on Data-Driven Fluid Dynamics　2022/09/05

  45. Quantifying uncertainty in deep learning for fluid flow reconstruction Invited

    Romit Maulik, Kai Fukami, Masaki Morimoto, Nesar Ramachandra, Ricardo Vinuesa, Koji Fukagata, Kunihiko Taira

    USACM Thematic Conference on Uncertainty Quantification for Machine Learning Integrated Physics Modeling (MLIP)　2022/08/18

  46. Machine-learning-based reconstruction of transient vortex-airfoil wake interaction

    Yonghong Zhong, Kai Fukami, Byungjin An, Kunihiko Taira

    AIAA Aviation Forum 2022　2022/06/27

  47. Time-varying broadcast mode analysis for airfoil wake dynamics

    Kai Fukami, Vedasri Godavarthi, Yonghong Zhong, Chi-An Yeh, Kunihiko Taira

    IUTAM Symposium on Data-driven modeling and optimization in fluid mechanics　2022/06/15

  48. Machine-learning-based turbulent state estimation from pressure sensors in a pump sump

    Kai Fukami, Byungjin An, Motohiko Nohmi, Masashi Obuchi, Kunihiko Taira

    86th conference of turbomachinery society of Japan　2022/05/12

  49. Broadcasting perturbations over turbulence Invited

    Kunihiko Taira, Chi-An Yeh, Kai Fukami

    Causality in turbulence and transition　2022/05/03

  50. Image and video compression of fluid flow data

    Vishal Anantharaman, Jason Feldkamp, Kai Fukami, Kunihiko Taira

    15th Southern California Flow Physics Symposium (SoCal Fluids XV)　2022/04/23

  51. Machine-learning-based flow reconstruction of gust vortex-airfoil wake interactions

    Yonghong Zhong, Kai Fukami, Byungjin An, Kunihiko Taira

    15th Southern California Flow Physics Symposium (SoCal Fluids XV)　2022/04/23

  52. Network broadcast analysis of airfoil wakes

    Kai Fukami, Vedasri Godavarthi, Chi-An Yeh, Kunihiko Taira

    15th Southern California Flow Physics Symposium (SoCal Fluids XV)　2022/04/23

  53. Flow field reconstruction from sparse sensors with neural networks: Progress and outlook Invited

    Kai Fukami, Yonghong Zhong, Kunihiko Taira

    Advanced Modeling & Simulations seminar at the University of Texas at El Paso (UTEP) - Multi-Scale/Physics Computational Laboratory　2022/04/22

  54. Reconstructing turbulence with deep learning: uncertainty quantification and outlook Invited

    Kai Fukami, Romit Maulik, Nesar Ramachandra, Masaki Morimoto, Ricardo Vinuesa, Koji Fukagata, Kunihiko Taira

    SIAM Conference on Uncertainty Quantification (UQ22)　2022/04/15

  55. Flow field reconstruction from sparse sensors with machine learning Invited

    Kai Fukami, Yonghong Zhong, Kunihiko Taira

    Seminar at Sorbonne University　2022/04/04

  56. Reconstructing turbulent flows with machine-learning-based super-resolution analysis Invited

    Kai Fukami, Koji Fukagata, Kunihiko Taira

    National Science Foundation AI Planning Institute for Data Driven Physics’ Workshop on ``AI Super-Resolution Simulations: from Climate Science to Cosmology"　2022/02/23

  57. Model form uncertainty quantification of neural network-based fluid flow estimation

    Masaki Morimoto, Kai Fukami, Romit Maulik, Ricardo Vinuesa, Koji Fukagata

    35th CFD symposium　2021/12/14

  58. Convolutional neural network-based global field recovery from sparse sensors of transitional boundary layer flow

    Taichi Nakamura, Kai Fukami, Koji Fukagata

    35th CFD symposium　2021/12/14

  59. Visualization of nonlinear modal structures for three-dimensional unsteady fluid flows with customized decoder design

    Kazuto Hasegawa, Kai Fukami, Koji Fukagata

    Machine Learning and the Physical Sciences, Workshop at the 35th Conference on Neural Information Processing Systems (NeurIPS)　2021/12/13

  60. Global field reconstruction from sparse sensors with Voronoi tessellation-assisted convolutional neural network Invited

    Kai Fukami, Romit Maulik, Nesar Ramachandra, Koji Fukagata, Kunihiko Taira

    Remote Colloquium on Vortex Dominated Flows (ReCoVor)　2021/12/10

  61. Robust machine learning of turbulence through generalized Buckingham Pi-inspired pre-processing of training data

    Kai Fukami, Kunihiko Taira

    74th Annual Meeting of the APS Division of Fluid Dynamics　2021/11/21

  62. Data-driven reduced-order modeling for turbulent flow forecast: neural networks and sparse regressions

    Taichi Nakamura, Kai Fukami, Koji Fukagata

    Mechanistic Machine Learning and Digital Twins for Computational Science, Engineering & Technology (MMLDT-CSET2021)　2021/09/26

  63. Convolutional neural network based three-dimensional fluid flow recovery from two-dimensional sectional data with super resolution based data augmentation

    Mitsuaki Matsuo, Taichi Nakamura, Masaki Morimoto, Kai Fukami, Koji Fukagata

    Mechanistic Machine Learning and Digital Twins for Computational Science, Engineering & Technology (MMLDT-CSET2021)　2021/09/26

  64. Latent space based feedback control design: Machine-learning-based reduced-order modeling of unsteady fluid flows

    Shoei Kanehira, Kai Fukami, Kazuto Hasegawa, Taichi Nakamura, Masaki Morimoto, Koji Fukagata

    Mechanistic Machine Learning and Digital Twins for Computational Science, Engineering & Technology (MMLDT-CSET2021)　2021/09/26

  65. Towards practical uses of supervised neural networks for fluid flow regressions

    Masaki Morimoto, Kai Fukami, Kai Zhang, Koji Fukagata

    Mechanistic Machine Learning and Digital Twins for Computational Science, Engineering & Technology (MMLDT-CSET2021)　2021/09/26

  66. Hierarchical neural network for spatial reconstruction of fluid flow fields

    Naoki Moriya, Masaki Morimoto, Kai Fukami, Kazuto Hasegawa, Koji Fukagata

    Japan Society of Fluid Mechanics Annual Meeting 2021　2021/09/21

  67. Applications of convolutional neural network-based nonlinear mode decomposition to three-dimensional fluid flows

    Kazuto Hasegawa, Kai Fukami, Koji Fukagata

    JSME Mechanical Engineering Congress, 2021 Japan (MECJ-21)　2021/09/05

  68. Fluid flow state estimation from sparse sensor measurements using convolutional neural network

    Taichi Nakamura, Kai Fukami, Koji Fukagata

    JSME Mechanical Engineering Congress, 2021 Japan (MECJ-21)　2021/09/05

  69. Time-series analysis of fluid flow dynamics using latent variables with sparse regressions: towards data-driven flow controls

    Kai Fukami, Takaaki Murata, Kai Zhang, Shoei Kanehira, Koji Fukagata

    JSME Mechanical Engineering Congress, 2021 Japan (MECJ-21)　2021/09/05

  70. Demonstration of machine learning-based reduced order modeling using unsteady flows around bluff bodies with various shapes

    Kazuto Hasegawa, Kai Fukami, Koji Fukagata

    25th International Congress of Theoretical and Applied Mechanics (XXV ICTAM)　2021/08/22

  71. Extracting nonlinear dynamics of low-dimensionalized flows

    Kai Fukami, Takaaki Murata, Koji Fukagata

    25th International Congress of Theoretical and Applied Mechanics (XXV ICTAM)　2021/08/22

  72. Error-curve analysis of neural network and linear stochastic estimation for fluid flow problems

    Taichi Nakamura, Kai Fukami, Koji Fukagata

    16th U.S. National Congress on Computational Mechanics　2021/07/25

  73. Parameter influence of supervised/unsupervised use of convolutional neural networks for fluid flow analyses

    Masaki Morimoto, Kai Fukami, Kai Zhang, Aditya G. Nair, Koji Fukagata

    16th U.S. National Congress on Computational Mechanics　2021/07/25

  74. Latent variable-based analysis with machine learning for reduced-order modeling and control of fluid flows

    Kai Fukami, Koji Hasegawa, Taichi Nakamura, Shoei Kanehira, Koji Fukagata

    16th U.S. National Congress on Computational Mechanics　2021/07/25

  75. 2D-3D CNN: Enabling neural networks for effective fluid data handling

    Mitsuaki Matsuo, Taichi Nakamura, Masaki Morimoto, Kai Fukami, Koji Fukagata

    22nd Workshop on Turbulence Control　2021/06/18

  76. Machine-learned invariant map for turbulent flow analysis and modeling: interpolation and extrapolation

    Kai Fukami, Kunihiko Taira

    Machine learning methods for prediction and control of separated turbulent flows　2021/06/16

  77. Clues for noise robustness of state estimation: Error-curve quest of neural network and linear regression

    Taichi Nakamura, Kai Fukami, Koji Fukagata

    International Conference on Learning Representation (ICLR) workshop, Deep Learning for Simulation (SIMDL)　2021/05/07

  78. Supervised convolutional networks for volumetric data enrichment from limited sectional data with adaptive super resolution

    Mitsuaki Matsuo, Kai Fukami, Taichi Nakamura, Masaki Morimoto, Koji Fukagata

    International Conference on Learning Representation (ICLR) workshop, Deep Learning for Simulation (SIMDL)　2021/05/07

  79. Voronoi tessellation-assisted convolutional neural network for flow field reconstruction from sparse sensors

    Kai Fukami, Romit Maulik, Nesar Ramachandra, Koji Fukagata, Kunihiko Taira

    14th Southern California Flow Physics Symposium (SoCal Fluids XIV)　2021/04/10

  80. Bends of weight surfaces for noise robustness: linear and nonlinear fluid flow regressions

    Taichi Nakamura, Kai Fukami, Koji Fukagata

    21st Workshop on Turbulence Control　2021/03/19

  81. Voronoi tessellation-aided machine learning for fluid flow data recovery from moving sparse sensors

    Kai Fukami, Romit Maulik, Nesar Ramachandra, Koji Fukagata, Kunihiko Taira

    21st Workshop on Turbulence Control　2021/03/19

  82. Utilization of autoencoder-based nonlinear manifolds for fluid flow forecasts driven with long short-term memory Invited

    Taichi Nakamura, Kai Fukami, Kazuto Hasegawa, Yusuke Nabae, Koji Fukagata

    DataLearning workshop of Data Science Institute, Imperial College London,　2021/03/16

  83. Feedback control of unsteady fluid flows using an autoencoder with sparse identification of nonlinear dynamics

    Shoei Kanehira, Kai Fukami, Kazuto Hasegawa, Taichi Nakamura, Masaki Morimoto, Koji Fukagata

    60th JSME Kanto Student Union Conference　2021/03/10

  84. Convolutional neural network based three-dimensional data reconstruction from two-dimensional sectional data with adaptive sampling

    Mitsuaki Matsuo, Masaki Morimoto, Taichi Nakamura, Kai Fukami, Koji Fukagata

    60th JSME Kanto Student Union Conference　2021/03/10

  85. Supervised machine learning based data-driven wall modeling for large-eddy simulation in a turbulent channel flow

    Naoki Moriya, Kai Fukami, Yusuke Nabae, Masaki Morimoto, Taichi Nakamura, Koji Fukagata

    60th JSME Kanto Student Union Conference　2021/03/10

  86. Machine learning based state estimation of turbulent flows: robustness for noisy input

    Taichi Nakamura, Kai Fukami, Koji Fukagata

    27th JSME Kanto Union conference　2021/03/10

  87. Neural network-based anomaly detections for nonlinear dynamical systems

    Masaki Morimoto, Kai Fukami, Taichi Nakamura, Koji Fukagata

    27th JSME Kanto Union conference　2021/03/10

  88. Toward practical global field reconstruction from sparse sensors with deep learning Invited

    Kai Fukami, Romit Maulik, Nesar Ramachandra, Koji Fukagata, Kunihiko Taira

    DataLearning workshop of Data Science Institute, Imperial College London,　2021/03/09

  89. Convolutional neural network based fluid data enrichment for numerical and experimental studies Invited

    Kai Fukami, Kunihiko Taira, Masaki Morimoto, Koji Fukagata

    SIAM Conference on Computational Science and Engineering (CSE21)　2021/03/01

  90. Autoencoder based extraction of low-dimensional manifolds in fluid flows Invited

    Masaki Morimoto, Kai Fukami, Kazuto Hasegawa, Taichi Nakamura, Koji Fukagata

    SIAM Conference on Computational Science and Engineering (CSE21)　2021/03/01

  91. The use of convolutional neural networks for PIV data augmentation

    Masaki Morimoto, Kai Fukami, Hikaru Murakami, Koji Fukagata

    14th World Congress on Computational Mechanics (WCCM) ECCOMAS Congress 2020　2021/01/11

  92. Low-dimensionalized flow representation with customized autoencoders

    Kai Fukami, Takaaki Murata, Koji Fukagata

    14th World Congress on Computational Mechanics (WCCM) ECCOMAS Congress 2020　2021/01/11

  93. Toward practical machine learning and fluid flow regressions: perspective on interpretability and generalizability

    Masaki Morimoto, Kai Fukami, Kai Zhang, Koji Fukagata

    34th CFD symposium　2020/12/21

  94. Three-dimensional flow field reconstruction from two-dimensional sectional data using machine learning

    Mitsuaki Matsuo, Masaki Morimoto, Taichi Nakamura, Kai Fukami, Koji Fukagata

    34th CFD symposium　2020/12/21

  95. Supervised machine learning for wall-modeling in large eddy simulation of turbulent channel flows

    Naoki Moriya, Kai Fukami, Yusuke Nabae, Masaki Morimoto, Taichi Nakamura, Koji Fukagata

    34th CFD symposium　2020/12/21

  96. Machine learning-aided state estimation in a turbulent channel flow and its robustness for sensor information

    Taichi Nakamura, Kai Fukami, Koji Fukagata

    34th CFD symposium　2020/12/21

  97. Probabilistic neural network-based reduced-order surrogate for fluid flows

    Kai Fukami, Romit Maulik, Nesar Ramachandra, Koji Fukagata, Kunihiko Taira

    Machine Learning and the Physical Sciences, Workshop at the 34th Conference on Neural Information Processing Systems (NeurIPS)　2020/12/07

  98. CNN-AE/LSTM based turbulent flow forecast on low-dimensional latent space

    Taichi Nakamura, Kai Fukami, Kazuto Hasegawa, Yusuke Nabae, Koji Fukagata

    Machine Learning and the Physical Sciences, Workshop at the 34th Conference on Neural Information Processing Systems (NeurIPS)　2020/12/07

  99. Toward latent space based feedback control with CNN-SINDy reduced order modeling of unsteady fluid flows

    Shoei Kanehira, Kai Fukami, Kazuto Hasegawa, Taichi Nakamura, Masaki Morimoto, Koji Fukagata

    20th Workshop on Turbulence Control　2020/12/02

  100. Convolutional neural network based wall modeling for large eddy simulation in a turbulent channel flow

    Naoki Moriya, Kai Fukami, Yusuke Nabae, Masaki Morimoto, Taichi Nakamura, Koji Fukagata

    73rd Annual Meeting of the APS Division of Fluid Dynamics　2020/11/22

  101. Visualization for internal procedure of neural networks for fluid flows

    Masaki Morimoto, Kai Fukami, Koji Fukagata

    73rd Annual Meeting of the APS Division of Fluid Dynamics　2020/11/22

  102. Unstructured fluid flow data recovery using machine learning and Voronoi diagrams

    Kai Fukami, Romit Maulik, Nesar Ramachandra, Kunihiko Taira, Koji Fukagata

    73rd Annual Meeting of the APS Division of Fluid Dynamics　2020/11/22

  103. Investigation of autoencoder-based low dimensionalization for various flow fields

    Masaki Morimoto, Kai Fukami, Kazuto Hasegawa, Taichi Nakamura, Koji Fukagata

    Japan Society of Fluid Mechanics Annual Meeting 2020　2020/09/18

  104. Extraction of nonlinear modes in fluid flows with a hierarchical convolutional neural network autoencoder

    Taichi Nakamura, Kai Fukami, Koji Fukagata

    Japan Society of Fluid Mechanics Annual Meeting 2020　2020/09/18

  105. ML-PIV: Convolutional neural network based velocity estimator for imperfect particle images

    Masaki Morimoto, Kai Fukami, Koji Fukagata

    18th Workshop on Turbulence Control　2020/06/26

  106. Application of machine learning for reduced order modeling of turbulent channel flow,

    Taichi Nakamura, Kai Fukami, Kazuto Hasegawa, Takaaki Murata, Yusuke Nabae, Koji Fukagata

    59th JSME Kanto Student Union Conference　2020/03/16

  107. Proposal of machine learning based particle image velocimetry

    Masaki Morimoto, Kai Fukami, Kazuto Hasegawa, Takaaki Murata, Hikaru Murakami, Koji Fukagata

    59th JSME Kanto Student Union Conference　2020/03/16

  108. A machine-learned turbulence generator for the channel flow

    Kai Fukami, Yusuke Nabae, Ken Kawai, Koji Fukagata

    2nd Pacific Rim Thermal Engineering Conference　2019/12/13

  109. Machine learning of turbulent channel flows using autoencoders

    Taichi Nakamura, Kai Fukami, Kazuto Hasegawa, Takaaki Murata, Yusuke Nabae, Koji Fukagata

    33rd CFD symposium　2019/11/27

  110. Improvement of PIV by data augmentation based on machine learning

    Masaki Morimoto, Kai Fukami, Kazuto Hasegawa, Takaaki Murata, Hikaru Murakami, Koji Fukagata

    33rd CFD symposium　2019/11/27

  111. Space-time recovery of high-resolution turbulent flow fields with machine learning based super resolution

    Kai Fukami, Koji Fukagata, Kunihiko Taira

    72nd Annual Meeting of the APS Division of Fluid Dynamics　2019/11/23

  112. Machine-learning-based super-resolution analysis for spatio-temporal data reconstruction of fluid flows

    Kai Fukami, Koji Fukagata, Kunihiko Taira

    17th Workshop on Turbulence Control　2019/11/14

  113. Machine-learned super-resolution analysis of three dimensional turbulent channel flow

    Kai Fukami, Koji Fukagata, Kunihiko Taira

    The Japan Society of Mechanical Engineers Fluids Engineering Conference 2019　2019/11/07

  114. Applications of machine-learned super-resolution algorithm for two-dimensional flow fields

    Kai Fukami, Koji Fukagata, Kunihiko Taira

    Japan Society of Fluid Mechanics Annual Meeting 2019　2019/09/13

  115. Super-resolution analysis with machine learning for low-resolution flow data

    Kai Fukami, Koji Fukagata, Kunihiko Taira

    11th International Symposium on Turbulence and Shear Flow Phenomena (TSFP11)　2019/07/30

  116. CNN/SINDy based reduced order modeling of unsteady flow fields

    Takaaki Murata, Kai Fukami, Koji Fukagata

    ASME-JSME-KSME Joint Fluids Engineering Conference 2019　2019/07/28

  117. Data-driven reduced order modeling of flows around two-dimensional bluff bodies flow of various shapes

    Kazuto Hasegawa, Kai Fukami, Takaaki Murata, Koji Fukagata

    ASME-JSME-KSME Joint Fluids Engineering Conference 2019　2019/07/28

  118. Image-based super-resolution analysis with machine learning for two-dimensional turbulence

    Kai Fukami, Koji Fukagata, Kunihiko Taira

    13th Southern California Flow Physics Symposium (SoCal Fluids XIII)　2019/04/20

  119. Prediction of unsteady flows using machine-learned reduced order model

    Kazuto Hasegawa, Kai Fukami, Takaaki Murata, Koji Fukagata

    58th JSME Kanto Student Union Conference　2019/03/18

  120. Machine learning-based prediction of flows around a circular cylinder at different Reynolds numbers

    Kazuto Hasegawa, Takaaki Murata, Kai Fukami, Koji Fukagata

    15th Workshop on Turbulence Control　2019/01/22

  121. Prediction of the Reynolds number dependency of flow around circular cylinder using machine learning

    Kazuto Hasegawa, Kai Fukami, Takaaki Murata, Koji Fukagata

    32nd CFD symposium　2018/12/11

  122. Extraction of low dimensional modes in a flow around a circular cylinder and prediction of their temporal evolutions using machine learning

    Takaaki Murata, Kai Fukami, Koji Fukagata

    32nd CFD symposium　2018/12/11

  123. Extraction of low dimensional modes in a flow around a circular cylinder and prediction of their temporal evolutions

    Takaaki Murata, Kai Fukami, Koji Fukagata

    14th Workshop on Turbulence Control　2018/10/17

  124. Introduction about Fluid dynamics & Machine Learning

    Kai Fukami

    13th Workshop on Turbulence Control　2018/05/25

  125. Proposal of an inflow turbulence generator using machine learning

    Kai Fukami, Ken Kawai, Koji Fukagata

    57th JSME Kanto Student Union Conference　2018/03/16

  126. Proposal of a turbulence generator using machine learning

    Kai Fukami, Ken Kawai, Koji Fukagata

    12th Workshop on Turbulence Control　2018/01/23

  127. Data-oriented analysis of extremely unsteady flows Invited

    Kai Fukami

    The RIKEN next-generation CAE consortium for combustion systems 2025　2025/04/25

  128. Data-oriented approaches for analysis of unsteady flows Invited

    Kai Fukami

    A seminar on “Fluids and Informatics” at the Japan Society of Mechanical Engineers　2025/04/22

  129. Data-driven analysis, modeling, and control of extreme aerodynamic flows

    Kunihiko Taira, Kai Fukami, Luke R. Smith, Yonghong Zhong, Alec J. Linot, Hiroto Odaka, Barbara Lopez-Doriga

    The EuroMech Colloquium on Data-Driven Fluid Dynamics and the 2nd ERCOFTAC Workshop on Machine Learning for Fluid Dynamics　2025/04/02

  130. Large eddy simulation with a data-oriented wall model in turbulent channel flow

    Naoki Moriya, Kai Fukami, Yusuke Nabae, Masaki Morimoto, Taichi Nakamura, Koji Fukagata

    19th Workshop on Turbulence Control　2020/09/08

Show all Show first 5

Research Projects 2

  1. “Japan-originated" compactification project for nonlinear aerodynamic phenomena with an ultra-high degree of freedom

    Kai Fukami, Soshi Kawai

    Offer Organization: Ministry of Education, Culture, Sports, Science and Technology (MEXT)

    System: MEXT Coordination Funds for Promoting AeroSpace Utilization

    2025/10 - 2028/03

  2. Designing super-quick control strategies of unsteady flows based on nonlinear phase-amplitude machine learning Competitive

    Kai Fukami

    Offer Organization: Japan Society for the Promotion of Science (JSPS)

    System: Grant-in-Aid for Research Activity Start-up

    Institution: Tohoku University

    2025/07 - 2027/03

Teaching Experience 4

  1. Design and Drawing II　Tohoku University

  2. Aerodynamics　Tohoku University

  3. Advanced Aero System II　Tohoku University

  4. Introduction to Aerospace Engineering　Tohoku University

Academic Activities 6

  1. JK-FLOW (Japan-Korea Fluid Mechanics Online Workshop)

    2025/03 - Present

    Activity type: Academic society, research group, etc.

  2. JSME RC: Advanced measurements, simulations, and data science for digital twin in thermal and fluid engineering and their industrial applications

    2025 - Present

  3. A local executive committee member, IUTAM Symposium, GA22-04: Machine Learning in Diverse Fluid Mechanics

    2025/05/15 - 2025/05/17

    Activity type: Academic society, research group, etc.

  4. 3rd Data-Driven Fluid Dynamics Workshop

    2025/03/17 - 2025/03/19

    Activity type: Competition, symposium, etc.

  5. SIAM Conference on Computational Science and Engineering (CSE23), Mini-symposium ``Reduced-complexity models for fluid flows"

    2023/02 -

    Activity type: Competition, symposium, etc.

  6. SIAM Conference on Mathematics of Data Science (MDS22), Mini-symposium ``Data-driven analysis and modeling of unsteady flows"

    2022/09 -

    Activity type: Competition, symposium, etc.

Show all Show first 5