A single-fibre computer enables textile networks and distributed inference

16 min read Original article ↗

Data availability

Data supporting the findings of this study are available at Zenodo (https://doi.org/10.5281/zenodo.13874664)52Source data are provided with this paper.

Code availability

Code supporting the findings of this study is publicly available at Zenodo (https://doi.org/10.5281/zenodo.13874664)52, except for code that, pursuant to copyright by Analog Devices, Inc. and Dialog Semiconductor Plc (‘vendors’), is not licenced for sharing. However, in most cases, such code is available for direct download from the vendors and has been annotated accordingly in our repository. Full code repositories are available from the corresponding author upon reasonable request and with permission of the vendors, as applicable.

References

  1. Dunn, J., Runge, R. & Snyder, M. Wearables and the medical revolution. Per Med. 15, 429–448 (2018).

    Article  CAS  PubMed  MATH  Google Scholar 

  2. Yetisen, A. K., Martinez‐Hurtado, J. L., Ünal, B., Khademhosseini, A. & Butt, H. Wearables in medicine. Adv. Mater. 30, e1706910 (2018).

    Article  PubMed  Google Scholar 

  3. Martin, T., Jovanov, E. & Raskovic, D. Issues in wearable computing for medical monitoring applications: a case study of a wearable ECG monitoring device. In Digest of Papers. Fourth International Symposium on Wearable Computers 43–49 (IEEE Computer Society, 2000).

  4. Kim, D.-H. et al. Epidermal electronics. Science 333, 838–843 (2011).

    Article  ADS  CAS  PubMed  MATH  Google Scholar 

  5. Liu, Y., Pharr, M. & Salvatore, G. A. Lab-on-skin: a review of flexible and stretchable electronics for wearable health monitoring. ACS Nano 11, 9614–9635 (2017).

    Article  CAS  PubMed  MATH  Google Scholar 

  6. Yang, G.-Z., Andreu-Perez, J., Hu, X. & Thiemjarus, S. in Body Sensor Networks (ed. Yang, G.-Z.) 301–354 (Springer, 2014).

  7. Muzammal, M., Talat, R., Sodhro, A. H. & Pirbhulal, S. A multi-sensor data fusion enabled ensemble approach for medical data from body sensor networks. Inf. Fusion 53, 155–164 (2020).

    Article  Google Scholar 

  8. Gravina, R., Alinia, P., Ghasemzadeh, H. & Fortino, G. Multi-sensor fusion in body sensor networks: state-of-the-art and research challenges. Inf. Fusion 35, 68–80 (2017).

    Article  Google Scholar 

  9. Tamura, T., Maeda, Y., Sekine, M. & Yoshida, M. Wearable photoplethysmographic sensors—past and present. Electronics (Basel) 3, 282–302 (2014).

    MATH  Google Scholar 

  10. Mathie, M. J., Coster, A. C. F., Lovell, N. H. & Celler, B. G. Accelerometry: providing an integrated, practical method for long-term, ambulatory monitoring of human movement. Physiol. Meas. 25, R1–R20 (2004).

    Article  ADS  PubMed  Google Scholar 

  11. Wicaksono, I. et al. A tailored, electronic textile conformable suit for large-scale spatiotemporal physiological sensing in vivo. npj Flexible Electron. 4, 5 (2020).

    Article  Google Scholar 

  12. Shi, J. et al. Smart textile‐integrated microelectronic systems for wearable applications. Adv. Mater. 32, e1901958 (2020).

    Article  PubMed  Google Scholar 

  13. Chen, G. et al. Electronic textiles for wearable point-of-care systems. Chem. Rev. 122, 3259–3291 (2022).

    Article  CAS  PubMed  MATH  Google Scholar 

  14. Yan, W. et al. Advanced multimaterial electronic and optoelectronic fibers and textiles. Adv. Mater. 31, e1802348 (2019).

    Article  PubMed  Google Scholar 

  15. Yan, W. et al. Thermally drawn advanced functional fibers: new frontier of flexible electronics. Mater. Today 35, 168–194 (2020).

    Article  CAS  MATH  Google Scholar 

  16. Bayindir, M., Abouraddy, A. F., Arnold, J., Joannopoulos, J. D. & Fink, Y. Thermal‐sensing fiber devices by multimaterial codrawing. Adv. Mater. 18, 845–849 (2006).

    Article  CAS  Google Scholar 

  17. Zhang, T. et al. High-performance, flexible, and ultralong crystalline thermoelectric fibers. Nano Energy 41, 35–42 (2017).

    Article  MATH  Google Scholar 

  18. Yan, W. et al. Single fibre enables acoustic fabrics via nanometre-scale vibrations. Nature 603, 616–623 (2022).

    Article  ADS  CAS  PubMed  MATH  Google Scholar 

  19. Gumennik, A. et al. All‐in‐fiber chemical sensing. Adv. Mater. 24, 6005–6009 (2012).

    Article  CAS  PubMed  MATH  Google Scholar 

  20. Pan, Z. et al. All-in-one stretchable coaxial-fiber strain sensor integrated with high-performing supercapacitor. Energy Storage Mater. 25, 124–130 (2020).

    Article  MATH  Google Scholar 

  21. Qu, Y. et al. Superelastic multimaterial electronic and photonic fibers and devices via thermal drawing. Adv. Mater. 30, e1707251 (2018).

    Article  PubMed  MATH  Google Scholar 

  22. Xiong, T. et al. Photo-powered all-in-one energy harvesting and storage fibers towards low-carbon smart wearables. Energy Storage Mater. 65, 103146 (2024).

    Article  Google Scholar 

  23. Khudiyev, T. et al. Thermally drawn rechargeable battery fiber enables pervasive power. Mater. Today 52, 80–89 (2022).

    Article  CAS  Google Scholar 

  24. Dong, C. et al. High-efficiency super-elastic liquid metal based triboelectric fibers and textiles. Nat. Commun. 11, 3537 (2020).

    Article  ADS  CAS  PubMed  PubMed Central  MATH  Google Scholar 

  25. Shi, X. et al. Large-area display textiles integrated with functional systems. Nature 591, 240–245 (2021).

    Article  ADS  CAS  PubMed  MATH  Google Scholar 

  26. Canales, A. et al. Multifunctional fibers for simultaneous optical, electrical and chemical interrogation of neural circuits in vivo. Nat. Biotechnol. 33, 277–284 (2015).

    Article  CAS  PubMed  MATH  Google Scholar 

  27. Hwang, S. et al. Integration of multiple electronic components on a microfibre towards an emerging electronic textile platform. Nat. Commun. 13, 3173 (2022).

    Article  ADS  CAS  PubMed  PubMed Central  MATH  Google Scholar 

  28. Rein, M. et al. Diode fibres for fabric-based optical communications. Nature 560, 214–218 (2018).

    Article  ADS  CAS  PubMed  MATH  Google Scholar 

  29. Danto, S., Ruff, Z., Wang, Z., Joannopoulos, J. D. & Fink, Y. Ovonic memory switching in multimaterial fibers. Adv. Funct. Mater. 21, 1095–1101 (2011).

    Article  CAS  Google Scholar 

  30. Danto, S. et al. Fiber field‐effect device via in situ channel crystallization. Adv. Mater. 22, 4162–4166 (2010).

    Article  CAS  PubMed  MATH  Google Scholar 

  31. Loke, G. et al. Digital electronics in fibres enable fabric-based machine-learning inference. Nat. Commun. 12, 1–9 (2021).

    Article  ADS  MATH  Google Scholar 

  32. Wang, Z. et al. High-quality semiconductor fibres via mechanical design. Nature 626, 72–78 (2024).

    Article  ADS  CAS  PubMed  PubMed Central  MATH  Google Scholar 

  33. Agcayazi, T., Chatterjee, K., Bozkurt, A. & Ghosh, T. K. Flexible interconnects for electronic textiles. Adv. Mater. Technol. 3, 1700277 (2018).

  34. Stanley, J., Hunt, J. A., Kunovski, P. & Wei, Y. A review of connectors and joining technologies for electronic textiles. Eng. Rep. 4, e12491 (2022).

  35. Marion, J. S. et al. Thermally drawn highly conductive fibers with controlled elasticity. Adv. Mater. 34, e2201081 (2022).

    Article  ADS  PubMed  Google Scholar 

  36. Poincloux, S., Adda-Bedia, M. & Lechenault, F. Geometry and elasticity of a knitted fabric. Phys. Rev. X 8, 021075 (2018).

    CAS  MATH  Google Scholar 

  37. Chen, S. et al. Exploring the relationship between applied fabric strain and resultant local yarn strain within the elastic fabric based on finite element method. J. Mater. Sci. 55, 10258–10270 (2020).

    Article  ADS  CAS  MATH  Google Scholar 

  38. Morton, W. E. & Hearle, J. W. S. Physical Properties of Textile Fibres (Woodhead Publishing, 2008).

    Book  MATH  Google Scholar 

  39. Iqbal, W., Jiang, Y., Qi, Y. & Xu, L. Yarn damage evaluation in the flat knitting process. AUTEX Res. J. 21, 272–283 (2021).

    Article  CAS  MATH  Google Scholar 

  40. Boussu, F., Trifigny, N., Cochrane, C. & Koncar, V. in Smart Textiles and Their Applications 375–400 (Elsevier, 2016).

  41. Brunnschweiler, D. The structure and tensile properties of braids. J. Text. Inst. Trans. 45, T55–T77 (1954).

    Article  MATH  Google Scholar 

  42. Rawal, A., Saraswat, H. & Kumar, R. Tensile response of tubular braids with an elastic core. Composites, Part A: Appl. Sci. Manuf. 47, 150–155 (2013).

    Article  CAS  MATH  Google Scholar 

  43. Siitonen, S., Laakkonen, P., Vahimaa, P., Kuittinen, M. & Tossavainen, N. White LED light coupling into light guides with diffraction gratings. Appl. Opt. 45, 2623–2630 (2006).

    Article  ADS  PubMed  Google Scholar 

  44. Miller, J. M., de Beaucoudrey, N., Chavel, P., Turunen, J. & Cambril, E. Design and fabrication of binary slanted surface-relief gratings for a planar optical interconnection. Appl. Opt. 36, 5717–5727 (1997).

    Article  ADS  CAS  PubMed  Google Scholar 

  45. Cornacchia, M., Ozcan, K., Zheng, Y. & Velipasalar, S. A survey on activity detection and classification using wearable sensors. IEEE Sens. J. 17, 386–403 (2017).

    Article  ADS  Google Scholar 

  46. Loke, G. et al. Computing fabrics. Matter 2, 786–788 (2020).

    Article  MATH  Google Scholar 

  47. Guo, Y. et al. Polymer composite with carbon nanofibers aligned during thermal drawing as a microelectrode for chronic neural interfaces. ACS Nano 11, 6574–6585 (2017).

    Article  CAS  PubMed  MATH  Google Scholar 

  48. Loke, G., Yan, W., Khudiyev, T., Noel, G. & Fink, Y. Recent progress and perspectives of thermally drawn multimaterial fiber electronics. Adv. Mater. 32, e1904911 (2020).

    Article  PubMed  Google Scholar 

  49. Branscomb, D., Beale, D. & Broughton, R. New directions in braiding. J. Eng. Fibers Fabr. 8, 11–24 (2013).

  50. Lalitha, V. & Srinivasan, K. A review of Manchester, Miller, and FM0 encoding techniques. Smart Comput. Rev. 4, 481–490 (2014).

  51. Yurtman, A. & Barshan, B. Activity recognition invariant to sensor orientation with wearable motion sensors. Sensors 17, 1838 (2017).

    Article  ADS  PubMed  PubMed Central  MATH  Google Scholar 

  52. Gupta, N., Cheung, H. & Payra, S. Fibre Computer Repository. Zenodo https://doi.org/10.5281/zenodo.13874664 (2025).

Download references

Acknowledgements

This study was sponsored by the Army Research Office and was accomplished under Cooperative Agreement Nos. W911NF-18-2-0048 and W911NF-23-2-0121. The views and conclusions contained in this document are those of the authors and should not be interpreted as representing the official policies, either expressed or implied, of the Army Research Office of the US Government. The US Government is authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation herein. We acknowledge the support from DTRA (Award No. HDTRA1-20-2-0002) Interaction of Ionizing Radiation with Matter University Research Alliance. S.P. acknowledges the support from the NSF Graduate Research Fellowship Program, the Fannie and John Hertz Foundation Fellowship, the Paul & Daisy Soros Foundation Fellowship for New Americans, the Knight-Hennessy Scholars Program at Stanford University and the Astronaut Scholarship Foundation. We thank K. Mazaheri, O. Schirm and S. Gupta for their contributions to the code implementation. We also thank A. Missakian for her support, inspiration and encouragement.

Author information

Author notes

  1. These authors contributed equally: Nikhil Gupta, Henry Cheung, Syamantak Payra

Authors and Affiliations

  1. Research Laboratory of Electronics, Massachusetts Institute of Technology, Cambridge, MA, USA

    Nikhil Gupta, Gabriel Loke, Jenny Li, Yongyi Zhao & Yoel Fink

  2. Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, MA, USA

    Nikhil Gupta, Gabriel Loke & Yoel Fink

  3. Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, Cambridge, MA, USA

    Henry Cheung, Syamantak Payra, Jenny Li, Yongyi Zhao & Samuel Kravitz

  4. Department of Electrical Engineering, Stanford University, Stanford, CA, USA

    Syamantak Payra

  5. Textiles Department, Rhode Island School of Design, Providence, RI, USA

    Latika Balachander & Ella Son

  6. Department of Computer Science, Brown University, Providence, RI, USA

    Vivian Li

  7. Department of Mechanical Engineering, Massachusetts Institute of Technology, Cambridge, MA, USA

    Sehar Lohawala

  8. Department of Physics, Massachusetts Institute of Technology, Cambridge, MA, USA

    John Joannopoulos

  9. Institute for Soldier Nanotechnologies, Massachusetts Institute of Technology, Cambridge, MA, USA

    John Joannopoulos

Authors

  1. Nikhil Gupta
  2. Henry Cheung
  3. Syamantak Payra
  4. Gabriel Loke
  5. Jenny Li
  6. Yongyi Zhao
  7. Latika Balachander
  8. Ella Son
  9. Vivian Li
  10. Samuel Kravitz
  11. Sehar Lohawala
  12. John Joannopoulos
  13. Yoel Fink

Contributions

G.L., S.P., N.G. and Y.F. conceived of the study. Y.F. and J.J. supervised the study. S.P., H.C. and N.G. designed and implemented the custom interposers for packaging the microelectronic devices. G.L., N.G. and S.L. performed the experiments for fabricating the fibres with the packaged devices. N.G. performed the mechanical characterization of the fibres. S.P., N.G. and H.C. performed the electrical characterization of the in-fibre devices and communications protocol. N.G. performed power draw measurements of the fibre computer and capacity measurements of the embedded battery fibre. S.P., H.C., J.L. and N.G. developed the code and circuitry for programming the MCUs to interface with peripheral devices and run customized algorithms. S.P., H.C. and S.K. wrote the code for multi-fibre optical communications. N.G. performed experiments for the characterization of optical communications between fibres. H.C. and J.L. implemented code on the BLE SoCs for communications with multiple fibres and external devices. L.B., E.S. and V.L. integrated the fibres into fabrics through weaving, knitting and sewing. Y.Z., H.C., J.L. and N.G. contributed to the data collection and code for neural network training and deployment in-fibre. N.G., S.P., H.C., Y.F. and J.J. wrote and revised the paper with input from all authors.

Corresponding author

Correspondence to Yoel Fink.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature thanks Tae-Wook Kim and Xiaoming Tao 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 figures and tables

Extended Data Fig. 1 Flexible Interposer Method for Device Packaging.

a. Images of the components placed on their corresponding interposers. b. Images showing examples of the workflow for the device packaging process for the microcontroller, with the device soldered onto the topside of the interposer with an auxiliary non-foldable region for initial programming (left), which is then cut down to the size of the device (central), and then folded around the device such that the contact pads on the bottom side of the interposer align with each of the faces of the device (right).

Extended Data Fig. 2 Pre-Connection Draw Process.

a. Images of devices pre-connected on an I2C bus on a spool, with 4 helically-wrapped wires for the intra-fibre communication protocol, and one guide wire for holding the tension of the draw. b. Image of preform with PMMA cladding encasing hollow ECOC cylinder. c. Image of a bare ECOC fibre after the PMMA cladding has been removed, with a microcontroller (MCU), accelerometer (acc), temperature sensor (temp), light sensor (LS), light emitting diode (LED), low dropout regulator (LDO), PPG sensor, and BLE module embedded inside the fibre.

Extended Data Fig. 3 Fibre Microscope Images.

Images of a a. microcontroller, b. LED, c. light sensor, d. temperature sensor, e. PPG sensor, f. accelerometer, g. low-dropout voltage regulator, h. BLE device, and i. Li-ion battery embedded in fibre. j. Closer-up image of the helical structure of the I2C conductors in the fibre.

Extended Data Fig. 4 Fatigue Testing.

Tensile fatigue testing of the fibre computer under different conditions. In each measurement, the resistance includes the four I2C wires and guide wire measured in parallel. a. Tensile fatigue testing between 0% and 5% strain at a strain rate of 1 Hz for a section of fibre with an embedded device. Failure in the first I2C wire occurs at 1697 cycles. b. Tensile fatigue testing between 0% and 15% strain at a strain rate of 0.2 Hz, for a section of fibre with an embedded device in the centre. Failure in the first I2C wire occurs at 122 cycles. c. Tensile fatigue testing between 0% and 15% strain at a strain rate of 0.2 Hz, for a section of fibre without an embedded device. Failure in the first I2C wire occurs at 542 cycles. d. Force-displacement curves during the first two cycles of fatigue testing of the section of fibre with the embedded device, showing the difference in tensile properties after the strain-limiting wire breaks during the first cycle.

Extended Data Fig. 5 Washability.

a. Images of a fibre computer embedded with a microcontroller, accelerometer, and BLE device undergoing 10 wash cycles. b. Status of the fibre before washing, with the I2C communications allowing for data exchange between the microcontroller and accelerometer, and the on-board LEDs of the microcontroller (red) and BLE (blue) devices indicating that the fibre is functional. Sensors are read at a polling rate of 50 Hz (20 ms between queries), while each sensor read takes 0.3 ms. c. Status of the fibre after washing, where we see the I2C communications between the microcontroller and accelerometer still intact to allow for data exchange, and the on-board LEDs of the microcontroller and BLE devices still indicating fibre functionality.

Extended Data Fig. 6 Tensile Properties of Braided Fibre Computers.

Images of fibre computers braided by a 16-carrier braiding machine using a. 540 denier polyester yarn and b. 360 denier cotton yarn, with 8 devices embedded inside each fibre. Enlarged image of a section of the c. polyester-braided and d. cotton-braided fibre containing an embedded microcontroller e. Force displacement curves for the braided fibre computers, with a notable enhancement in the tensile strength compared to bare fibres.

Extended Data Fig. 7 Optical Transmission Over Waveguide.

a. Images of the setup used for testing optical transmission over waveguide with controlled compression at the interface between the waveguide and fibres. A light sensor and LED fibre are placed a certain distance apart with a waveguide placed perpendicularly on top of the fibres. A screw-in clamp with an in-line load cell is used for adjusting the compressive strain and measuring the compressive force (Supplementary Note 9). In between the fibres, the waveguide is woven in a fabric to simulate the fabric environment. b. Increase in intensity received by the light sensor fibre over time as the compressive strain at the interface between the waveguide and LED fibre is increased in a stepwise manner. c. Amplitude received by the light sensor fibre when the LED fibre is excited with a square wave of different frequencies, showing a 3 dB bandwidth of between 10-20 kHz.

Extended Data Fig. 8 Implementation of Bluetooth Network for Distributed Inference.

a. Schematic showing the timeline of data transmission between fibres, with individual neural network inferences sent to a single fibre (left), which are then averaged to produce a network inference that is sent to another fibre (middle), which is then relayed to an external BLE display for continual monitoring. b. Image of four fibres, each with microcontrollers (red) and BLE modules (blue) for wireless communication of neural network inference between fibres. c. Example image of inference output on an external phone display communicated via BLE from four-fibre network.

Extended Data Fig. 9 In-Fabric Optical Communications.

a. Images of the interfaces between the LED fibre and waveguides (top left), and the light sensor fibre and waveguides (bottom left) for the sample with lower tension. Image of sample when the LED fibre is emitting light, showing weak coupling to the waveguide bus (right). b. Images of the interfaces between the LED fibre and waveguides (top left), and the light sensor fibre and waveguides (bottom left) for the sample with higher tension. Image of sample when the LED fibre is emitting light, showing stronger coupling to the waveguide bus (right). c. Plot of the intensity received by the light sensor in each fabric when the LED fibre is toggled to the on state. The distance between the LED fibre and light sensor fibre is 3 cm in each case. d. Bit transmission at 5 Hz in the fabric sample shown in (b), where the LED and light sensor fibres are separated by 7.5 cm.

Extended Data Fig. 10 Extraction of Weights for Pre-Encoding Anatomical Behavioural Awareness.

An example of the process for extracting the confidence level weights that were pre-encoded on the individual fibres for performing the federated inference algorithm with anatomical behavioural awareness (Supplementary Note 11). a. The validation set confusion matrix for the right leg fibre, with the precision of the predictions for each activity listed below. The precisions from each limb’s fibre on the validation dataset were pre-stored in a corresponding array on each fibre to enable evaluation of the federated inference algorithm evaluation on the test dataset. b. The arrays of confidence weights stored on each fibre.

Supplementary information

Supplementary Information

The supplementary information file contains Supplementary Notes 1–12, Figs. 1–13, Tables 1 and 2 and References.

Source data

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Gupta, N., Cheung, H., Payra, S. et al. A single-fibre computer enables textile networks and distributed inference. Nature 639, 79–86 (2025). https://doi.org/10.1038/s41586-024-08568-6

Download citation

  • Received:

  • Accepted:

  • Published:

  • Version of record:

  • Issue date:

  • DOI: https://doi.org/10.1038/s41586-024-08568-6