Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Article
  • Published:

Bio-inspired artificial mechanoreceptors with built-in synaptic functions for intelligent tactile skin

Nature Materials volume 24, pages 1100–1108 (2025)Cite this article

Subjects

Abstract

Tactile perception involves the preprocessing of signals from slowly adapting and fast-adapting afferent neurons, which exhibit synapse-like interactions between mechanoreceptors and their dendrites or terminals, transmitting signals to the brain. Emulating these adaptation and sensory memory functions is crucial for artificial tactile sensing systems. Here, inspired by human tactile afferent systems, we present an array of artificial synaptic mechanoreceptors with built-in synaptic functions by vertically integrating synaptic transistors with a reduced graphene oxide channel, an ionogel gate dielectric and an elastomeric fingerprint-like receptive layer in an all-in-one platform. Triboelectric-capacitive gating between the receptive layer and gate dielectric in response to tactile stimulation governs excitatory post-synaptic current patterns, enabling slowly adapting and fast-adapting characteristics for signal preprocessing. The artificial synaptic mechanoreceptor array demonstrated handwriting style, surface pattern and texture discrimination via machine learning using fused slowly adapting and fast-adapting post-synaptic values, offering high data efficiency and potential for intelligent skin.

Access through your institution
Buy or subscribe

This is a preview of subscription content, access via your institution

Access options

Access through your institution

Buy this article

  • Purchase on SpringerLink
  • Instant access to full article PDF
Buy now

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Comparison of human and artificial tactile perception.
Fig. 2: Characterization of synaptic properties of synaptic rGO-FET.
Fig. 3: Structure, mechanism and synaptic properties of ASMRs during tactile stimulation.
Fig. 4: Applications of ASMRs array with built-in synaptic functions.

Similar content being viewed by others

Data availability

All data used for this study are available within the Article and its Supplementary Information. Further data are available from the corresponding author upon reasonable request. Source data are provided with this paper.

Code availability

The codes used to run the simulations and plot the data (excluding the custom-made LabVIEW program) are available from the corresponding author upon reasonable request.

References

  1. Torre, V., Ashmore, J. F., Lamb, T. D. & Menini, A. Transduction and adaptation in sensory receptor cells. J. Neurosci. 15, 7757–7768 (1995).

    CAS  PubMed  PubMed Central  Google Scholar 

  2. Woo, S.-H., Lumpkin, E. A. & Patapoutian, A. Merkel cells and neurons keep in touch. Trends Cell Biol. 25, 74–81 (2015).

    PubMed  Google Scholar 

  3. Hao, J., Bonnet, C., Amsalem, M., Ruel, J. & Delmas, P. Transduction and encoding sensory information by skin mechanoreceptors. Pflug. Arch. Eur. J. Physiol. 467, 109–119 (2015).

    CAS  Google Scholar 

  4. Abraira, V. E. & Ginty, D. D. The sensory neurons of touch. Neuron 79, 618–639 (2013).

    CAS  PubMed  Google Scholar 

  5. Gottschaldt, K.-M. & Vahle-Hinz, C. Merkel cell receptors: structure and transducer function. Science 214, 183–186 (1981).

    CAS  PubMed  Google Scholar 

  6. Zimmerman, A., Bai, L. & Ginty, D. D. The gentle touch receptors of mammalian skin. Science 346, 950–954 (2014).

    CAS  PubMed  PubMed Central  Google Scholar 

  7. Jenkins, B. A. & Lumpkin, E. A. Developing a sense of touch. Development 144, 4078–4090 (2017).

    CAS  PubMed  PubMed Central  Google Scholar 

  8. Maksimovic, S. et al. Epidermal merkel cells are mechanosensory cells that tune mammalian touch receptors. Nature 509, 617–621 (2014).

    CAS  PubMed  PubMed Central  Google Scholar 

  9. Wu, J., Lewis, A. H. & Grandl, J. Touch, tension, and transduction – the function and regulation of piezo ion channels. Trends Biochem. Sci. 42, 57–71 (2017).

    PubMed  Google Scholar 

  10. Cobo, R., García-Piqueras, J., Cobo, J. & Vega, J. A. The human cutaneous sensory corpuscles: an update. J. Clin. Med. 10, 227 (2021).

    CAS  PubMed  PubMed Central  Google Scholar 

  11. Song, Z., Banks, R. W. & Bewick, G. S. Modelling the mechanoreceptor’s dynamic behaviour. J. Anat. 227, 243–254 (2015).

    CAS  PubMed  PubMed Central  Google Scholar 

  12. Mendelson, M. & Loewenstein, W. R. Mechanisms of receptor adaptation. Science 144, 554–555 (1964).

    CAS  PubMed  Google Scholar 

  13. Bag, A. et al. Bio‐inspired sensory receptors for artificial‐intelligence perception. Adv. Mater. https://doi.org/10.1002/adma.202403150 (2024).

  14. Khacef, L. et al. Spike-based local synaptic plasticity: a survey of computational models and neuromorphic circuits. Neuromorph. Comput. Eng. 3, 042001 (2023).

    Google Scholar 

  15. Lee, Y. R., Trung, T. Q., Hwang, B.-U. & Lee, N.-E. A flexible artificial intrinsic-synaptic tactile sensory organ. Nat. Commun. 11, 2753 (2020).

    CAS  PubMed  PubMed Central  Google Scholar 

  16. Tien, N. T. et al. A flexible bimodal sensor array for simultaneous sensing of pressure and temperature. Adv. Mater. 26, 796–804 (2014).

    PubMed  Google Scholar 

  17. Jeon, S., Lim, S.-C., Trung, T. Q., Jung, M. & Lee, N.-E. Flexible multimodal sensors for electronic skin: principle, materials, device, array architecture, and data acquisition method. Proc. IEEE 107, 2065–2083 (2019).

    CAS  Google Scholar 

  18. Ni, Y. et al. Visualized in-sensor computing. Nat. Commun. 15, 3454 (2024).

    CAS  PubMed  PubMed Central  Google Scholar 

  19. Ding, G., Han, S. T., Roy, V. A. L., Kuo, C. C. & Zhou, Y. Triboelectric nanogenerator for neuromorphic electronics. Energy Rev. 2, 100014 (2023).

    Google Scholar 

  20. Wan, C. et al. An artificial sensory neuron with tactile perceptual learning. Adv. Mater. 30, 1801291 (2018).

    Google Scholar 

  21. Jiang, C. et al. Neuromorphic antennal sensory system. Nat. Commun. 15, 2109 (2024).

    CAS  PubMed  PubMed Central  Google Scholar 

  22. Wang, W. et al. Neuromorphic sensorimotor loop embodied by monolithically integrated, low-voltage, soft e-skin. Science 380, 735–742 (2023).

    CAS  PubMed  Google Scholar 

  23. Yu, J. et al. Contact-electrification-activated artificial afferents at femtojoule energy. Nat. Commun. 12, 1581 (2021).

    CAS  PubMed  PubMed Central  Google Scholar 

  24. Yu, J. et al. Bioinspired mechano-photonic artificial synapse based on graphene/MoS2 heterostructure. Sci. Adv. 7, eabd9117 (2021).

    CAS  PubMed  PubMed Central  Google Scholar 

  25. Zhao, J., Wei, Z., Yang, X., Zhang, G. & Wang, Z. Mechanoplastic tribotronic two-dimensional multibit nonvolatile optoelectronic memory. Nano Energy 82, 105692 (2021).

    CAS  Google Scholar 

  26. Jia, M. et al. Multibit tribotronic nonvolatile memory based on van der Waals heterostructures. Nano Energy 83, 105785 (2021).

    CAS  Google Scholar 

  27. Zhang, C. et al. Bioinspired artificial sensory nerve based on Nafion memristor. Adv. Funct. Mater. 29, 1808783 (2019).

    Google Scholar 

  28. Chortos, A., Liu, J. & Bao, Z. Pursuing prosthetic electronic skin. Nat. Mater. 15, 937–950 (2016).

    CAS  PubMed  Google Scholar 

  29. Hoffman, B. U. et al. Merkel cells activate sensory neural pathways through adrenergic synapses. Neuron 100, 1401–1413 (2018).

    CAS  PubMed  PubMed Central  Google Scholar 

  30. Kandel, E. R., Schwartz, J. H., Jessell, T. M., Siegelbaum, S. A. & Hudspeth, A. J. Principles of Neural Science (McGraw-Hill, 2012).

  31. Geim, A. K. & Novoselov, K. S. The rise of graphene. Nat. Mater. 6, 183–191 (2007).

    CAS  PubMed  Google Scholar 

  32. Khan, U., Kim, T.-H., Ryu, H., Seung, W. & Kim, S. Graphene tribotronics for electronic skin and touch screen applications. Adv. Mater. 29, 1603544 (2017).

    Google Scholar 

  33. Wang, D. et al. Recent advanced applications of ion-gel in ionic-gated transistor. npj Flex. Electron 5, 13 (2021).

    CAS  Google Scholar 

  34. Bhunia, R., Kim, J. S., Kweon, H., Kim, D. J. & Kim, D. H. Ferroelectric ion gel-modulated long-term plasticity in organic synaptic transistors. Mater. Chem. Phys. 287, 126227 (2022).

    CAS  Google Scholar 

  35. Kong, L. et al. Ion-gel gated field-effect transistors with solution-processed oxide semiconductors for bioinspired artificial synapses. Org. Electron. 39, 64–70 (2016).

    CAS  Google Scholar 

  36. Gong, C., Chen, L., Liu, W. & Zhang, G. Study of short-term synaptic plasticity in ion-gel gated graphene electric-double-layer synaptic transistors. J. Semicond. 42, 014101 (2021).

    Google Scholar 

  37. Gouras, P. & Bishop, P. O. Neural basis of vision. Science 177, 188–189 (1972).

    CAS  PubMed  Google Scholar 

  38. Lee, H.-B. et al. Mogul‐patterned elastomeric substrate for stretchable electronics. Adv. Mater. 28, 3069–3077 (2016).

    CAS  PubMed  Google Scholar 

  39. Ba, Y.-Y. et al. Single-layer triboelectric nanogenerators based on ion-doped natural nanofibrils. ACS Appl. Mater. Interf. 12, 42859–42867 (2020).

    CAS  Google Scholar 

  40. Chung, J. et al. Ion‐enhanced field emission triboelectric nanogenerator. Adv. Energy Mater. 9, 1901731 (2019).

    CAS  Google Scholar 

  41. Ye, B. U. et al. Electrospun ion gel nanofibers for flexible triboelectric nanogenerator: electrochemical effect on output power. Nanoscale 7, 16189–16194 (2015).

    CAS  PubMed  Google Scholar 

  42. Yi, Z., Zhang, Y. & Peters, J. Biomimetic tactile sensors and signal processing with spike trains: a review. Sens. Actuat. A 269, 41–52 (2018).

    CAS  Google Scholar 

  43. Amoli, V. et al. A bioinspired hydrogen bond-triggered ultrasensitive ionic mechanoreceptor skin. Nat. Commun. 10, 4019 (2019).

    PubMed  PubMed Central  Google Scholar 

  44. Lee, J. I. et al. Visco‐poroelastic electrochemiluminescence skin with piezo‐ionic effect. Adv. Mater. 33, 2100321 (2021).

    CAS  Google Scholar 

  45. Ha, M. et al. Skin-inspired hierarchical polymer architectures with gradient stiffness for spacer-free, ultrathin, and highly sensitive triboelectric sensors. ACS Nano 12, 3964–3974 (2018).

    CAS  PubMed  Google Scholar 

  46. Kim, W.-G. et al. Triboelectric nanogenerator: structure, mechanism, and applications. ACS Nano 15, 258–287 (2021).

    CAS  PubMed  Google Scholar 

  47. Hu, W., Zhang, C. & Wang, Z. L. Recent progress in piezotronics and tribotronics. Nanotechnology 30, 042001 (2019).

    CAS  PubMed  Google Scholar 

  48. Fan, F.-R., Tian, Z.-Q. & Wang, Z. L. Flexible triboelectric generator. Nano Energy 1, 328–334 (2012).

    CAS  Google Scholar 

  49. Katz, L. C. & Shatz, C. J. Synaptic activity and the construction of cortical circuits. Science 274, 1133–1138 (1996).

    CAS  PubMed  Google Scholar 

  50. Kim, Y. et al. A bioinspired flexible organic artificial afferent nerve. Science 360, 998–1003 (2018).

    CAS  PubMed  Google Scholar 

  51. Shiffrin, R. M. & Atkinson, R. C. Storage and retrieval processes in long-term memory. Psychol. Rev. 76, 179–193 (1969).

    Google Scholar 

  52. McGaugh, J. L. Memory - a century of consolidation. Science 287, 248–251 (2000).

    CAS  PubMed  Google Scholar 

  53. Zang, Y., Shen, H., Huang, D., Di, C.-A. & Zhu, D. A dual-organic-transistor-based tactile-perception system with signal-processing functionality. Adv. Mater. 29, 1606088 (2017).

    Google Scholar 

  54. Bag, A., Moon, D.-B., Park, K.-H., Cho, C.-Y. & Lee, N.-E. Room-temperature-operated fast and reversible vertical-heterostructure-diode gas sensor composed of reduced graphene oxide and AlGaN/GaN. Sens. Actuat. B 296, 126684 (2019).

    CAS  Google Scholar 

Download references

Acknowledgements

This research was supported by the Basic Science Research Program (grant numbers RS-2019-NR040077 and RS-2025-00520659) and the Nanomaterial Technology Development Program (grant number RS-2024-00403639) through the National Research Foundation of Korea (NRF), funded by the Ministry of Science and Technology and the Ministry of Education.

Author information

Author notes

  1. These authors contributed equally: Seok Ju Hong, Yu Rim Lee, Atanu Bag.

Authors and Affiliations

  1. School of Advanced Materials Science & Engineering, Sungkyunkwan University, Suwon, Republic of Korea

    Seok Ju Hong, Yu Rim Lee, Atanu Bag, Hyo Soo Kim, Tran Quang Trung, M. Junaid Sultan, Dong-Bin Moon & Nae-Eung Lee

  2. Research Center for Advanced Materials Technology, Sungkyunkwan University, Suwon, Republic of Korea

    Atanu Bag & Nae-Eung Lee

  3. SKKU Advanced Institute of Nano Technology (SAINT), Sungkyunkwan University, Suwon, Republic of Korea

    Nae-Eung Lee

  4. Samsung Advanced Institute for Health Sciences and Technology (SAIHST), Sungkyunkwan University, Suwon, Republic of Korea

    Nae-Eung Lee

  5. Biomedical Institute for Convergence at SKKU (BICS), Sungkyunkwan University, Suwon, Republic of Korea

    Nae-Eung Lee

Authors
  1. Seok Ju Hong
    View author publications

    Search author on:PubMed Google Scholar

  2. Yu Rim Lee
    View author publications

    Search author on:PubMed Google Scholar

  3. Atanu Bag
    View author publications

    Search author on:PubMed Google Scholar

  4. Hyo Soo Kim
    View author publications

    Search author on:PubMed Google Scholar

  5. Tran Quang Trung
    View author publications

    Search author on:PubMed Google Scholar

  6. M. Junaid Sultan
    View author publications

    Search author on:PubMed Google Scholar

  7. Dong-Bin Moon
    View author publications

    Search author on:PubMed Google Scholar

  8. Nae-Eung Lee
    View author publications

    Search author on:PubMed Google Scholar

Contributions

N.-E.L. and Y.R.L. conceived the study and designed the experiments. S.J.H. and Y.R.L. fabricated the devices and conducted the measurements. A.B. designed and proposed ML experiments and carried out the ML simulations. Y.R.L., H.S.K. and T.Q.T. developed the initial device fabrication protocol. M.J.S. and D.-B.M. supported the electrical measurements. S.J.H., Y.R.L. and A.B. carried out the data analysis and drafted the manuscript. S.J.H., A.B. and N.-E.L. revised the manuscript. N.-E.L. supervised the entire study, provided funding and reviewed and edited the manuscript. All authors approved the final version of the manuscript.

Corresponding author

Correspondence to Nae-Eung Lee.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Materials thanks Chiara Bartolozzi, Ye Zhou and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 Schematic of the device structure and photographs of the fabricated single ASMRs and fabrication steps.

a, Design image of a set of ASMRs with slowly adapting (SA)- and fast-adapting (FA)-type characteristics. b, Photographs of the fabricated SA and FA ASMRs on PI substrate and the fabrication process steps. The channel was formed beneath the source and drain electrodes. The source and drain electrodes were encapsulated to prevent gate leakage. The ionogel gate dielectric was coated to connect the channel and gate area for the SA mechanoreceptor, while the ionogel gate dielectric is extended from the gate to the end of the planar electrode for the FA mechanoreceptor. Additionally, an ionogel gate dielectric layer was formed on top of the planar electrode in the touch area. The receptive layer was installed above the touch area. The contact area of the receptive layer in the SA ASMR is positioned above the IG gate dielectric layer of the rGO-FET, covering the entire channel as well as the source and drain regions. In contrast, the receptive layer in the FA ASMR is located on the IG-coated on the square-shaped planar electrode, which is extended up to the separate IG gate dielectric layer on the rGO-FET.

Extended Data Fig. 2 Synaptic properties of IG-gated rGO-FET.

a, The EPSC change ratio (\(\Delta\)IEPSC/IEPSC,i) of long-term plasticity (LTP) properties with a longer retention time lasting over 1830 s. b, Synaptic strength (SS) values at 0, 5, and 10 s after removing the applied pre-synaptic pulse (Vrec) (amplitudes: −0.1, −0.5, −0.7, −1, −1.2, −1.5, and −2 V). The SS10 (10 s after removing Vrec) was almost 52 times higher for the case of an applied Vrec of −2 V (7.29) compared to that with an applied Vrec of −0.7 V (0.14). The SS values were negligible at −0.1 V of Vrec (0.007) and −0.5 V of Vrec (0.03), indicating that ionic transport was very slight, showing STP properties. c, Responses of rGO-FET to positive pre-synaptic pulses with varying amplitude. Vrec amplitudes ranged from 0.1 V to 3 V. Positive Vrec led to decreased post-synaptic current (IIPSC) values compared to negative Vrec, attributed to positive ion migration towards the channel, reducing hole conductivity. d, SS variation with the number of Vrec pulses (5, 10, 20, 50, 100 times) recorded at 0 s, 5 s, and 10 s after pulse removal. e, SS after 0, 5, and 10 s of applied Vrec pulses with various pulse durations (0.05, 0.07, 0.2, 0.5, 1, 3, and 5 s). f, Responses of rGO-FET to different pulse rates. The EPSC change ratio (\(\Delta\)IEPSC/IEPSC,i) of rGO-FET synapse when stimulated by the electric pulses of Vrec (pulse width of 0.01 and amplitude of −0.5 V) at different pulse rates (1, 3, 5, 8, and 10 Hz). The \(\Delta\)IEPSC/IEPSC,i value was increased with an increase in the frequency of spikes.

Source data

Extended Data Fig. 3 Excitatory post-synaptic current (IEPSC) and source leakage current (Ileak) measured for SA ASMR and five different configurations of FA ASMR with varying designs.

IEPSC and Ileak were measured under 8 kPa mechanical stimulation for 5 s. IEPSC was monitored at a drain-to-source voltage (Vds) of 0.2 V, while Ileak was measured at Vds = 0 V. The panels show the schematic design, mechanical stimulation conditions, and measured IEPSC and Ileak for: a, SA ASMR, b, Standard FA ASMR with a planar electrode-channel distance of 0.3 mm, c, Modified FA ASMR with an electrode-channel distance of 1.5 mm, d, Modified FA ASMR with an electrode-channel distance of 3 mm, e, Modified FA ASMR with a direct connection between the source and planar electrode, with an electrode-channel distance of 0.3 mm, and f, Modified FA ASMR without a planar electrode.

Source data

Extended Data Fig. 4 Synaptic properties of ASMRs with tactile stimulation.

a, SS values of ASMRs with increasing tactile stimulation amplitudes. Different pressures ranging from 0.4 kPa to 33 kPa with a contact time of 5 s were applied to the ASMRs using an Al block. The SS values of both SA and FA ASMRs increased with the tactile stimulation amplitude and were saturated near 30 kPa pressure. In the observed SA-EPSC data, the \(\Delta\)IEPSC/IEPSC,i from touch to release was 2.1 at a pressure of 0.4 kPa and 2.8 at a pressure of 33 kPa. On the other hand, the \(\Delta\)IEPSC/IEPSC,i of FA-EPSC was 0.57 at a pressure of 0.4 kPa and 2.8 at a pressure of 33 kPa. b, SS values of ASMRs with increasing rates of tactile stimulation (0.2, 0.4, 0.6, and 0.8 Hz). 8 kPa of pressure was applied to the ASMRs using an Al block. The SS increased as the tactile stimulation rate increased. c, EPSC change ratio of ASMRs with increasing durations of tactile stimulation (0.1, 0.5, 1, 3, 5, 7 s). A pressure of 8 kPa was applied to the ASMRs using an Al block. The EPSC change ratio increased with longer duration times. d, Linearity of SS between SA- and FA-type mechanoreceptors with different numbers of touches. The linearity of FA-EPSC was higher than that of SA-EPSC, indicating a higher SS value in FA-EPSC (0.85) compared to SA-EPSC (0.58) after 100 stimulations.

Source data

Extended Data Fig. 5 Schematic of the device structure and photographs of the as-fabricated device and fabrication steps.

a, Design image of the ASMRs array consisting of 8 slowly adapting (SA)- and 8 fast adapting (FA)-type ASMRs. Transistors (TRs) for SA and FA mechanoreceptors are noted as SA and FA TRs. b, Photograph of the fabricated ASMRs array on a PI substrate and fabrication process steps of ASMRs array. The receptive contact areas during sliding of textured materials are marked with rectangular boxes. The channel was formed beneath the source and drain electrodes. The source and drain electrodes were encapsulated to prevent gate leakage. The ionogel gate dielectric was coated to connect the channel and gate area for the SA mechanoreceptor, while the ionogel gate dielectric extended from the gate to the end of the planar electrode for the FA mechanoreceptor. Additionally, an ionogel gate dielectric layer was formed on top of the planar electrode in the touch area. The receptive layer was installed above the touch area. The receptive contact areas of the ASMRs array were designed in an alternating interwoven pattern, allowing the signals of FA and SA ASMRs to be generated from the same area of textured materials during sliding motion. This design is expected to enable more accurate identification of the spatial distribution of tactile stimulus information.

Extended Data Fig. 6 Handwriting pattern recognition and tracking of the order of writing.

a, Schematic representation of the expected Korean characters ‘ㄱ’, ‘ㄴ’, and ‘ㄷ’generated by writing on our ASMRs. b, Pattern recognition and order of writing of the Korean character ‘ㄴ’. The writing order is shown as 1, 5, 9, 13, 14, 15, 16 (left), 16, 15, 14, 13, 9, 5, 1 (middle), 1, 5, 9, 13, 16, 15, 14, 13 (right). c, Pattern recognition and order of writing of the Korean character ‘ㄷ’. The writing order is described as 1, 2, 3, 4, 1, 5, 9, 13, 14, 15, 16 (left), 4, 3, 2, 1, 5, 9, 13, 14, 15, 16 (middle), 16, 15, 14, 13, 9, 5, 1, 2, 3, 4 (right).

Source data

Extended Data Fig. 7 Signal processing procedure for ML based classification.

The diagrams illustrate the process of preparation by combining the output signals of SA and FA ASMRs array for distinguishing various surface patterns and textures using ML. a, Schematic illustration depicting the feature selection procedure from the EPSC signals of SA and FA ASMRs. b, Schematic showing the detailed data preparation hierarchy and classification strategy, utilizing the EPSC signals of SA and FA ASMRs separately. c, Illustration of the multimodal fusion technique of SA- and FA-mimicking EPSC signals.

Extended Data Fig. 8 Surface pattern classification using SS values recorded after mechanical stimulation.

a-c, Confusion matrices displaying highest classification accuracies after considering features from (a) SA-type, (b) FA- type ASMRs, and (c) multimodal fusion of (SA/FA) features for 4 types of surface patterns (that is, 1: Mogul, 2: Planar, 3: Pyramid, and 4: Square). d-f, Receiver operating characteristic (ROC) curves for the (d) SA-, (e) FA-, and (f) SA/FA-based classification.

Source data

Extended Data Fig. 9 Surface texture classification results using SS values recorded after mechanical stimulation.

For ML, we collected 2560 EPSC signals (1280 SA signals and 1280 FA signals) using the ASMRs array, spanning 16 categories of textures. a-c, Confusion matrices illustrating the surface texture classification accuracy (highest) using features from (a) SA, (b) FA ASMRs, and (c) multimodal fusion of SA and FA (SA/FA) features for the classification of 16 texture categories (that is, 1: Cotton 100%, 2: polyester 100%, 3: rayon 100%, 4: nylon 100%, 5: acetate 100%, 6: cotton 65%-polyester 35%, 7: polyester 85%-spandex 15%, 8: cotton 52%-nylon 48%, 9: polyester 56%-rayon 44%, 10: nylon 54%-rayon 46%, 11: cashmere 100%, 12: linen 100%, 13: cowhide 100%, 14: skin, 15: nitrile 100%, and 16: Al foil). d-f, Receiver operating characteristic (ROC) curves for texture classification using (d) SA features, (e) FA features, and (f) combined SA/FA features.

Source data

Extended Data Fig. 10 Surface texture classification results using EPSC values recorded during mechanical stimulation.

a-c, Confusion matrices illustrating the highest surface texture classification accuracy using features from (a) FA, (b) SA ASMRs, and (c) multimodal fusion of SA and FA (SA/FA) features for 16 texture categories (that is, 1: Cotton 100%, 2: polyester 100%, 3: rayon 100%, 4: nylon 100%, 5: acetate 100%, 6: cotton 65%-polyester 35%, 7: polyester 85%-spandex 15%, 8: cotton 52%-nylon 48%, 9: polyester 56%-rayon 44%, 10: nylon 54%-rayon 46%, 11: cashmere 100%, 12: linen 100%, 13: cowhide 100%, 14: skin, 15: nitrile 100%, and 16: Al foil). d, Average classification accuracy for different surface textures using EPSC signals from FA, SA, and SA/FA fused responses recorded during the initial 20 s of mechanical stimulation. Dot plots display data distribution, with error bars indicating the standard deviation (n = 5).

Source data

Supplementary information

Supplementary Information

Supplementary Notes 1–15, Figs. 1–25, Tables 1–4 and references.

Supplementary Data

Source data for Supplementary Figs. 2–6,10–12, 14, 17–20 and 22–25.

Source data

Source Data Fig. 1

Statistical source data.

Source Data Fig. 2

Statistical source data.

Source Data Fig. 3

Statistical source data.

Source Data Fig. 4

Statistical source data.

Source Data Extended Data Fig. 2

Statistical source data.

Source Data Extended Data Fig. 3

Statistical source data.

Source Data Extended Data Fig. 4

Statistical source data.

Source Data Extended Data Fig. 6

Statistical source data.

Source Data Extended Data Fig. 8

Statistical source data.

Source Data Extended Data Fig. 9

Statistical source data.

Source Data Extended Data Fig. 10

Statistical source data.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Hong, S.J., Lee, Y.R., Bag, A. et al. Bio-inspired artificial mechanoreceptors with built-in synaptic functions for intelligent tactile skin. Nat. Mater. 24, 1100–1108 (2025). https://doi.org/10.1038/s41563-025-02204-y

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue date:

  • DOI: https://doi.org/10.1038/s41563-025-02204-y

This article is cited by

Search

Advanced search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing