pith. sign in

arxiv: 2605.19098 · v1 · pith:UXDCRKBGnew · submitted 2026-05-18 · 💻 cs.ET

Embodying Intelligence into Mechanical Metamaterials via Reservoir Computing

Shan He , Steven Kiyabu , Philip R. Buskohl , Patrick Musgrave This is my paper

Pith reviewed 2026-05-20 07:43 UTC · model grok-4.3

classification 💻 cs.ET
keywords mechanical metamaterialsreservoir computingphysical neural networkembodied intelligencenonlinear dynamicsproprioceptionvibration sensing
0
0 comments X

The pith

Mechanical metamaterials with contact nonlinearities can process vibrations as a physical reservoir computer using only linear readouts.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper shows that a network of mechanical unit cells featuring physical contact points can serve as a reservoir that nonlinearly transforms incoming vibrations. A simple linear training step on the resulting sensor signals then solves both standard benchmark computations and tasks tied directly to the structure's own motion, such as proprioception. Experiments compare this nonlinear version against a linear counterpart to establish that the contact-induced nonlinearity is required for strong performance. The work further identifies that the material splits input frequencies into new content distributed across different sensor locations, and that selecting sensors by frequency content improves results.

Core claim

A metamaterial built from unit cells whose contact nonlinearities act as physical leaky ReLU activations functions as a reservoir computer. It maps input vibrations into a higher-dimensional space of spatially separated frequency responses; a trained linear readout on those responses then computes both independent benchmark tasks and embodied tasks coupled to the structure's dynamics.

What carries the argument

The metamaterial reservoir of unit cells with contact nonlinearities that generate ReLU-like transformations and separate input vibrations into new frequency content distributed across sensor readouts.

If this is right

  • Nonlinearity supplied by physical contacts is essential for high performance on both benchmark and embodied tasks.
  • The metamaterial remains effective across inputs that vary in complexity.
  • Frequency separation across sensors directly drives task success, as confirmed by a greedy sensor-selection algorithm.
  • A memory-versus-nonlinearity subspace can be used to predict how well a given metamaterial design will generalize across tasks.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Structures built this way could sense and react to their environment without any external processor.
  • The same contact-nonlinearity principle might be scaled to larger mechanical systems for distributed, real-time adaptation.
  • Designers could use the memory-nonlinearity metric to choose metamaterial geometries for specific sensing goals before fabrication.

Load-bearing premise

The physical contacts inside the unit cells produce consistent, repeatable nonlinear transformations under vibration without being disrupted by fabrication differences, hysteresis, or sensor noise.

What would settle it

Running the same tasks on an otherwise identical metamaterial in which the contacts are replaced by purely linear elastic connections and checking whether task accuracy falls to near-chance levels.

Figures

Figures reproduced from arXiv: 2605.19098 by Patrick Musgrave, Philip R. Buskohl, Shan He, Steven Kiyabu.

Figure 1
Figure 1. Figure 1: Overview of a nonlinear vibratory mechanical metamaterial functioning as a physical reservoir [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: PRC performance of nonlinear and linear metamaterials under varying input complexities. (a) [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Principal Components Analysis of the frequency content of the reservoir’s readout signals to [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Prediction performance as a function of readout sensors under a four-tone input case. (a) [PITH_FULL_IMAGE:figures/full_fig_p010_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Mapping of the reservoir dynamics and the computational tasks in the nonlinearity vs. memory [PITH_FULL_IMAGE:figures/full_fig_p012_5.png] view at source ↗
read the original abstract

This study harnesses the embodied intelligence of mechanical metamaterials to sense and process environmental vibrations with minimal digital computation. Using physical reservoir computing (PRC), we turn the metamaterial and its nonlinear dynamics into a physical neural network that nonlinearly transforms the input vibrations and uses a simple linear training to compute a range of tasks. We introduce a novel metamaterial reservoir composed of a network of unit cells with contact nonlinearities that are the physical equivalent of leaky rectified linear unit (ReLU) activation functions. We experimentally show that the metamaterial reservoir can compute two classes of tasks: independent tasks, such as benchmark functions, and embodied tasks, such as proprioception, which we introduce to describe tasks coupled to the structure's dynamics. By comparing against a linear metamaterial, we demonstrate that nonlinearity is critical for high task performance, and we show that the metamaterial is robust to inputs of varying complexity. Through dimensionality reduction, we uncover the governing information separation mechanism and show that the metamaterial separates the input vibrations into new frequency content spatially distributed across the sensor readouts. We then confirm that frequency content is a key indicator of task performance by conducting an optimal sensor selection study using a frequency-based greedy algorithm. Finally, we demonstrate that a metamaterial's generalized performance for different tasks can be quantified using the memory vs. nonlinearity subspace, providing a design tool for other reservoir abstractions. These results establish the embodied intelligence of mechanical metamaterials and provide a path for sense-assess-response in intelligent systems.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper introduces a mechanical metamaterial reservoir composed of unit cells with contact nonlinearities that emulate leaky ReLU activations. It experimentally demonstrates that this physical reservoir can perform independent benchmark tasks and embodied tasks such as proprioception, shows that nonlinearity is essential via direct comparison to a linear metamaterial control, identifies frequency-content separation across sensor readouts as the governing mechanism, and proposes a memory-versus-nonlinearity subspace metric as a design tool for reservoir performance.

Significance. If the experimental claims hold, the work advances physical reservoir computing by embedding computation directly into passive mechanical structures, enabling low-power vibration sensing and processing for intelligent systems. Credit is given for the explicit linear-metamaterial baseline comparison, the introduction of embodied proprioception tasks, and the frequency-based sensor-selection analysis that links mechanism to performance.

major comments (2)
  1. [Experimental validation] Experimental validation section: the claim that contact nonlinearities furnish stable, history-independent ReLU-like maps (central to the performance gap versus the linear control) is not supported by quantitative repeatability tests, hysteresis-loop measurements, or multi-trial statistics under sustained vibration; without these, fabrication scatter or micro-slip could produce the observed advantage as an artifact rather than the intended nonlinearity.
  2. [Results] Results on task performance and robustness: the reported superiority for benchmark and proprioceptive tasks lacks error bars, trial counts, or statistical significance tests, which is load-bearing for the assertion that the metamaterial is robust to inputs of varying complexity and that frequency separation is a reliable performance indicator.
minor comments (2)
  1. [Introduction] The definition of the embodied proprioception task could be stated more formally with explicit input-output mapping to avoid ambiguity with standard reservoir benchmarks.
  2. [Results] Figure captions for the frequency-separation visualizations should include the exact sensor locations and frequency bins used in the dimensionality-reduction analysis.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback on our manuscript. The comments highlight important aspects of experimental rigor and statistical reporting that we have addressed through targeted revisions. Below we respond point by point to the major comments.

read point-by-point responses

  1. Referee: [Experimental validation] Experimental validation section: the claim that contact nonlinearities furnish stable, history-independent ReLU-like maps (central to the performance gap versus the linear control) is not supported by quantitative repeatability tests, hysteresis-loop measurements, or multi-trial statistics under sustained vibration; without these, fabrication scatter or micro-slip could produce the observed advantage as an artifact rather than the intended nonlinearity.

    Authors: We agree that additional quantitative characterization strengthens the central claim regarding the contact nonlinearities. In the revised manuscript we have expanded the Experimental validation section to include hysteresis-loop measurements obtained over repeated loading cycles under sustained vibration. We also report multi-trial statistics (n=10 independent trials per unit cell) demonstrating consistent ReLU-like maps with low variance across trials. These data confirm stability and history independence while indicating that fabrication scatter and micro-slip do not account for the observed performance differences relative to the linear control. revision: yes

  2. Referee: [Results] Results on task performance and robustness: the reported superiority for benchmark and proprioceptive tasks lacks error bars, trial counts, or statistical significance tests, which is load-bearing for the assertion that the metamaterial is robust to inputs of varying complexity and that frequency separation is a reliable performance indicator.

    Authors: We concur that error bars, explicit trial counts, and statistical tests are necessary to support the robustness and performance claims. The revised manuscript now includes error bars (standard deviation) on all task-performance figures, with trial counts explicitly stated (n=5 independent trials per task). We have added a statistical analysis subsection reporting p-values from paired t-tests that confirm significant differences between the nonlinear and linear metamaterials. These additions directly support the assertions regarding robustness to input complexity and the reliability of frequency separation as a performance indicator. revision: yes

Circularity Check

0 steps flagged

No circularity: experimental results grounded in physical measurements and baseline comparison

full rationale

The paper reports experimental demonstrations of reservoir computing tasks using a physical metamaterial with contact nonlinearities, including direct performance comparisons against a linear metamaterial control. Claims about task computation, frequency separation, and the importance of nonlinearity rest on measured sensor readouts and empirical robustness checks rather than any derivation that reduces by construction to fitted parameters, self-definitions, or self-citation chains. The work is self-contained against external benchmarks (physical experiments) with no load-bearing steps that equate predictions to inputs via the paper's own equations or prior author work.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 1 invented entities

The central claim rests on the domain assumption that physical nonlinear dynamics can serve as an effective reservoir whose states are adequately captured by a modest number of sensors, plus the standard reservoir-computing premise that linear readouts suffice once the reservoir provides sufficient nonlinearity and memory. No explicit free parameters or invented physical entities are described; the 'proprioception' task label is a new framing rather than a new physical entity.

axioms (2)
  • domain assumption Linear regression on sensor readouts is sufficient to extract task performance from the nonlinearly transformed physical states
    Core premise of reservoir computing applied here; invoked when the paper states that only simple linear training is used after the physical transformation.
  • domain assumption Contact nonlinearities in the unit cells behave consistently and repeatably under the tested vibration inputs
    Required for the experimental results to generalize; implied by treating the contacts as reliable physical equivalents of leaky ReLU activations.
invented entities (1)
  • Embodied tasks (proprioception) no independent evidence
    purpose: To label computation tasks that are directly coupled to the metamaterial's own dynamics
    New conceptual category introduced to distinguish tasks integrated with the structure's motion from purely external benchmark functions.

pith-pipeline@v0.9.0 · 5803 in / 1652 out tokens · 79781 ms · 2026-05-20T07:43:17.232017+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

44 extracted references · 44 canonical work pages

  1. [1]

    Princi- pal component analysis.Wiley interdisciplinary reviews: computational statistics, 2(4):433–459, 2010

    Herv´ e Abdi and Lynne J Williams. Princi- pal component analysis.Wiley interdisciplinary reviews: computational statistics, 2(4):433–459, 2010

    work page 2010

  2. [2]

    Salvatore Ameduri and A Concilio. Morph- ing wings review: Aims, challenges, and cur- rent open issues of a technology.Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science, 237(18):4112–4130, 2023

    work page 2023

  3. [3]

    A decade of soft robotic manipulators: Advances in design, modeling, control, and emerging chal- lenges.Journal of Bionic Engineering, pages 1– 44, 2025

    Elsayed Atif Aner, Omar M Shehata, Mo- hammed Ibrahim Awad, and Nancy E ElHady. A decade of soft robotic manipulators: Advances in design, modeling, control, and emerging chal- lenges.Journal of Bionic Engineering, pages 1– 44, 2025

    work page 2025

  4. [4]

    Physical reservoir computing with origami and its application to robotic crawling.Scientific Reports, 11(1):13002, 2021

    Priyanka Bhovad and Suyi Li. Physical reservoir computing with origami and its application to robotic crawling.Scientific Reports, 11(1):13002, 2021

    work page 2021

  5. [5]

    Nonlinear mechanical metamaterial cloaks.Advanced Functional Materials, page e22895, 2025

    Giovanni Bordiga, Jean-Gabriel Argaud, Au- drey A Watkins, Vincent Tournat, and Katia Bertoldi. Nonlinear mechanical metamaterial cloaks.Advanced Functional Materials, page e22895, 2025

    work page 2025

  6. [6]

    Ef- ficient optical reservoir computing for parallel data processing.Optics Letters, 47(15):3784– 3787, 2022

    Ting Bu, He Zhang, Santosh Kumar, Mingwei Jin, Prajnesh Kumar, and Yuping Huang. Ef- ficient optical reservoir computing for parallel data processing.Optics Letters, 47(15):3784– 3787, 2022. 15

    work page 2022

  7. [7]

    Combina- torial design of textured mechanical metamate- rials.Nature, 535(7613):529–532, 2016

    Corentin Coulais, Eial Teomy, Koen De Reus, Yair Shokef, and Martin Van Hecke. Combina- torial design of textured mechanical metamate- rials.Nature, 535(7613):529–532, 2016

    work page 2016

  8. [8]

    Computing with networks of nonlinear mechanical oscillators.PloS one, 12(6):e0178663, 2017

    Jean C Coulombe, Mark CA York, and Julien Sylvestre. Computing with networks of nonlinear mechanical oscillators.PloS one, 12(6):e0178663, 2017

    work page 2017

  9. [9]

    Information pro- cessing capacity of dynamical systems.Scientific reports, 2(1):514, 2012

    Joni Dambre, David Verstraeten, Benjamin Schrauwen, and Serge Massar. Information pro- cessing capacity of dynamical systems.Scientific reports, 2(1):514, 2012

    work page 2012

  10. [10]

    Arameh Eyvazian, Yooseob Song, Chibukhchyan Hovhannes, Ardeshir Savari, and Narinderjit Singh Sawaran Singh. State-of-the-art soft robotic systems for unstructured and real-world environments: A systematic review.Engineering Science and Technology, an International Jour- nal, 73:102264, 2026

    work page 2026

  11. [11]

    Twin vortex computer in fluid flow.New Journal of Physics, 23(6):063051, 2021

    Ken Goto, Kohei Nakajima, and Hirofumi Notsu. Twin vortex computer in fluid flow.New Journal of Physics, 23(6):063051, 2021

    work page 2021

  12. [12]

    Programmable elastic metamaterials

    Babak Haghpanah, Hamid Ebrahimi, Davood Mousanezhad, Jonathan Hopkins, and Ashkan Vaziri. Programmable elastic metamaterials. Advanced Engineering Materials, 18(4):643–649, 2016

    work page 2016

  13. [13]

    Physical reser- voir computing on a soft bio-inspired swimmer

    Shan He and Patrick Musgrave. Physical reser- voir computing on a soft bio-inspired swimmer. Neural Networks, 181:106766, 2025

    work page 2025

  14. [14]

    The role of nonlinearity, dimensionality, memory, and input on a mechanical oscillator reservoir computer

    Shan He and Patrick F Musgrave. The role of nonlinearity, dimensionality, memory, and input on a mechanical oscillator reservoir computer. Neurocomputing, 652:131158, 2025

    work page 2025

  15. [15]

    Bird-inspired robotics principles as a framework for developing smart aerospace materials.Journal of Composite Materials, 57(4):679–710, 2023

    Kenneth AW Hoffmann, Tony G Chen, Mark R Cutkosky, and David Lentink. Bird-inspired robotics principles as a framework for developing smart aerospace materials.Journal of Composite Materials, 57(4):679–710, 2023

    work page 2023

  16. [16]

    Reservoir computing beyond memory- nonlinearity trade-off.Scientific reports, 7(1):10199, 2017

    Masanobu Inubushi and Kazuyuki Yoshimura. Reservoir computing beyond memory- nonlinearity trade-off.Scientific reports, 7(1):10199, 2017

    work page 2017

  17. [17]

    Short term memory in echo state networks

    Herbert Jaeger. Short term memory in echo state networks. 2001

    work page 2001

  18. [18]

    Harnessing nonlinearity: Predicting chaotic systems and saving energy in wireless communication.sci- ence, 304(5667):78–80, 2004

    Herbert Jaeger and Harald Haas. Harnessing nonlinearity: Predicting chaotic systems and saving energy in wireless communication.sci- ence, 304(5667):78–80, 2004

    work page 2004

  19. [19]

    Soft robotics: a bioinspired evolu- tion in robotics.Trends in Biotechnology, 2013

    Sangbae Kim. Soft robotics: a bioinspired evolu- tion in robotics.Trends in Biotechnology, 2013

    work page 2013

  20. [20]

    Optomechanical reservoir computing.Proceedings of the National Academy of Sciences, 122(29):e2424991122, 2025

    Steven Kiyabu, Daniel Nelson, John Thomson, Benjamin Schultz, Timothy Vincent, Nathan Hertlein, Andrew Gillman, Amanda Criner, and Philip R Buskohl. Optomechanical reservoir computing.Proceedings of the National Academy of Sciences, 122(29):e2424991122, 2025

    work page 2025

  21. [21]

    Perspectives on intelligence in soft robotics.Advanced Intelligent Systems, 7(1):2400294, 2025

    Vera Gesina Kortman, Barbara Mazzolai, Aime´ e Sakes, and Jovana Jovanova. Perspectives on intelligence in soft robotics.Advanced Intelligent Systems, 7(1):2400294, 2025

    work page 2025

  22. [22]

    Mechanical neural networks: Archi- tected materials that learn behaviors.Science Robotics, 7(71):eabq7278, 2022

    Ryan H Lee, Erwin AB Mulder, and Jonathan B Hopkins. Mechanical neural networks: Archi- tected materials that learn behaviors.Science Robotics, 7(71):eabq7278, 2022

    work page 2022

  23. [23]

    Physical reservoir computing with emerging electronics

    Xiangpeng Liang, Jianshi Tang, Yanan Zhong, Bin Gao, He Qian, and Huaqiang Wu. Physical reservoir computing with emerging electronics. Nature Electronics, 7(3):193–206, 2024

    work page 2024

  24. [24]

    Y. Ma. A review of smart materials for morphing structures.Composite Structures, 2019

    work page 2019

  25. [25]

    Rectifier nonlinearities improve neural net- work acoustic models

    Andrew L Maas, Awni Y Hannun, Andrew Y Ng, et al. Rectifier nonlinearities improve neural net- work acoustic models. InProc. icml, volume 30, page 3. Atlanta, GA, 2013

    work page 2013

  26. [26]

    Real-time computing without stable states: A new framework for neural com- putation based on perturbations.Neural compu- tation, 14(11):2531–2560, 2002

    Wolfgang Maass, Thomas Natschl¨ ager, and Henry Markram. Real-time computing without stable states: A new framework for neural com- putation based on perturbations.Neural compu- tation, 14(11):2531–2560, 2002

    work page 2002

  27. [27]

    Md Najmul Mowla, Davood Asadi, Tahir Durhasan, Javad Rashid Jafari, and Moham- madreza Amoozgar. Recent advancements in morphing applications: Architecture, artificial intelligence integration, challenges, and future trends-a comprehensive survey.Aerospace Sci- ence and Technology, 161:110102, 2025

    work page 2025

  28. [28]

    Physical reservoir comput- ing—an introductory perspective.Japanese Journal of Applied Physics, 59(6):060501, 2020

    Kohei Nakajima. Physical reservoir comput- ing—an introductory perspective.Japanese Journal of Applied Physics, 59(6):060501, 2020

    work page 2020

  29. [29]

    A soft body as a reservoir: case studies in a dynamic model of octopus-inspired 16 soft robotic arm.Frontiers in computational neuroscience, 7:91, 2013

    Kohei Nakajima, Helmut Hauser, Rongjie Kang, Emanuele Guglielmino, Darwin G Caldwell, and Rolf Pfeifer. A soft body as a reservoir: case studies in a dynamic model of octopus-inspired 16 soft robotic arm.Frontiers in computational neuroscience, 7:91, 2013

    work page 2013

  30. [30]

    Spectral analysis of mechan- ical reservoir computing with relu spring net- works

    Daniel Nelson, Steven Kiyabu, Timothy Vin- cent, Andrew Gillman, Amanda Criner, and Philip R Buskohl. Spectral analysis of mechan- ical reservoir computing with relu spring net- works. InSmart Materials, Adaptive Structures and Intelligent Systems, volume 88322, page V001T08A004. American Society of Mechanical Engineers, 2024

    work page 2024

  31. [31]

    Recent advances in reservoir comput- ing with a focus on electronic reservoirs

    Fabiha Nowshin, Yuhao Zhang, Lingjia Liu, and Yang Yi. Recent advances in reservoir comput- ing with a focus on electronic reservoirs. In2020 11th International Green and Sustainable Com- puting Workshops (IGSC), pages 1–8. IEEE, 2020

    work page 2020

  32. [32]

    Phase- changing metamaterial capable of variable stiff- ness and shape morphing.Advanced Engineering Materials, 21(12):1900802, 2019

    Ryan Poon and Jonathan B Hopkins. Phase- changing metamaterial capable of variable stiff- ness and shape morphing.Advanced Engineering Materials, 21(12):1900802, 2019

    work page 2019

  33. [33]

    Large-scale optical reservoir computing for spa- tiotemporal chaotic systems prediction.Physical Review X, 10(4):041037, 2020

    Mushegh Rafayelyan, Jonathan Dong, Yongqi Tan, Florent Krzakala, and Sylvain Gigan. Large-scale optical reservoir computing for spa- tiotemporal chaotic systems prediction.Physical Review X, 10(4):041037, 2020

    work page 2020

  34. [34]

    Stable propagation of mechani- cal signals in soft media using stored elastic en- ergy.Proceedings of the National Academy of Sciences, 113(35):9722–9727, 2016

    Jordan R Raney, Neel Nadkarni, Chiara Daraio, Dennis M Kochmann, Jennifer A Lewis, and Katia Bertoldi. Stable propagation of mechani- cal signals in soft media using stored elastic en- ergy.Proceedings of the National Academy of Sciences, 113(35):9722–9727, 2016

    work page 2016

  35. [35]

    Design, fabri- cation and control of soft robots.Nature, 2015

    Daniela Rus and Michael Tolley. Design, fabri- cation and control of soft robots.Nature, 2015

    work page 2015

  36. [36]

    Recent advances in physical reser- voir computing: A review.Neural Networks, 115:100–123, 2019

    Gouhei Tanaka, Toshiyuki Yamane, Jean Benoit H´ eroux, Ryosho Nakane, Naoki Kanazawa, Seiji Takeda, Hidetoshi Numata, Daiju Nakano, and Akira Hirose. Recent advances in physical reser- voir computing: A review.Neural Networks, 115:100–123, 2019

    work page 2019

  37. [37]

    Biomechanics of bird flight

    Bret W Tobalske. Biomechanics of bird flight. Journal of Experimental Biology, 210(18):3135– 3146, 2007

    work page 2007

  38. [38]

    Information processing dynamics with perturbed cylinder-flap fluid-structure interac- tions

    Timothy Vincent, David M Boston, Steven Kiyabu, Alexander M Pankonien, and Philip Buskohl. Information processing dynamics with perturbed cylinder-flap fluid-structure interac- tions. InAIAA SCITECH 2026 Forum, page 0900, 2026

    work page 2026

  39. [39]

    Fluid- based reservoir computing for distributed signal processing.Journal of Intelligent Material Sys- tems and Structures, 36(18-19):1297–1300, 2025

    Timothy Vincent and Alex Pankonien. Fluid- based reservoir computing for distributed signal processing.Journal of Intelligent Material Sys- tems and Structures, 36(18-19):1297–1300, 2025

    work page 2025

  40. [40]

    Advancements in soft robotics: a comprehensive review on actuation methods, materials, and applications.Polymers, 16(8):1087, 2024

    Yanmei Wang, Yanen Wang, Ray Tahir Mush- taq, and Qinghua Wei. Advancements in soft robotics: a comprehensive review on actuation methods, materials, and applications.Polymers, 16(8):1087, 2024

    work page 2024

  41. [41]

    Metamaterial beam with resonators of ac- tive feedback control to broaden and tune the bandgaps.Acta Mechanica, 236(3):2331–2343, 2025

    Yuhang Wang, Lifeng Wang, and Yuqiang Gao. Metamaterial beam with resonators of ac- tive feedback control to broaden and tune the bandgaps.Acta Mechanica, 236(3):2331–2343, 2025

    work page 2025

  42. [42]

    Mechan- ical computing.Nature, 598(7879):39–48, 2021

    Hiromi Yasuda, Philip R Buskohl, Andrew Gillman, Todd D Murphey, Susan Stepney, Richard A Vaia, and Jordan R Raney. Mechan- ical computing.Nature, 598(7879):39–48, 2021

    work page 2021

  43. [43]

    Embodying multifunctional mechano- intelligence in and through phononic metastruc- tures harnessing physical reservoir computing

    Yuning Zhang, Aditya Deshmukh, and Kon-Well Wang. Embodying multifunctional mechano- intelligence in and through phononic metastruc- tures harnessing physical reservoir computing. Advanced Science, 10(34):2305074, 2023

    work page 2023

  44. [44]

    Harnessing physical reservoir computing in nonlinear me- chanical metastructures

    Yuning Zhang and Kon-Well Wang. Harnessing physical reservoir computing in nonlinear me- chanical metastructures. InAIAA Scitech 2022 Forum, page 1464, 2022. 17 Supporting Information: Embodying Intelligence into Mechanical Metamaterials via Reservoir Computing Finite Element Model of Metamaterial Unit Cells The finite element analysis (FEA) of the ReLU b...

    work page 2022