Related Experiment Video
Updated: Sep 11, 2025
04:06
Author Spotlight: Enhancing Remote Rehabilitation with Virtual Reality and Electromyography
Published on: January 12, 2024
706
Deep Neural-Assisted Flexible MXene-Ag Composite Strain Sensor with Crack Dual Conductive Network for Human Motion
Junheng Fu1, Zichen Xia1, Haili Zhong1
1College of Water Conservancy and Hydropower Engineering, Sichuan Agricultural University, Ya'an 625014, China.
Materials (Basel, Switzerland)
|August 14, 2025
Summary
This study developed a novel multilevel microcrack strain sensor (MAP) for health electronics. The sensor achieves high sensitivity and a wide linear range, enabling accurate detection of human activities using deep learning.
Area of Science:
- Materials Science and Engineering
- Biomedical Engineering
- Wearable Technology
Background:
- Developing stretchable strain sensors with high sensitivity and wide linear range is crucial for health electronics.
- Current sensors face challenges in meeting the practical demands of daily health monitoring.
- Existing technologies often compromise biocompatibility or comfort for performance.
Purpose of the Study:
- To propose a novel heterogeneous surface strategy for creating high-performance stretchable strain sensors.
- To construct a multilevel microcrack strain sensor (MAP) with enhanced detection capabilities.
- To integrate sensor arrays with deep learning for intelligent recognition of human activities.
Main Methods:
- A heterogeneous surface strategy involving in situ silver deposition on modified PDMS (polydimethylsiloxane).
- MXene spray coating to construct a multilevel microcrack strain sensor (MAP) with silver nanoparticles and MXene.
- Seamless integration of sensor arrays and application of deep learning algorithms for activity recognition.
Main Results:
- The MAP sensor demonstrated excellent detection performance with a gauge factor (GFmax) of 487.3 and a response time of approximately 65 ms.
- The dual conductive network within the multilevel heterogeneous microcrack structure enabled robust performance across various deformation variables.
- Deep learning algorithms achieved up to 95% accuracy in identifying different joint movements using the integrated sensor arrays.
Conclusions:
- The developed multilevel heterogeneous microcrack strain sensor offers a promising solution for high-performance stretchable strain sensing.
- The sensor exhibits excellent biocompatibility and comfort, suitable for daily health monitoring applications.
- The integration with deep learning provides an intelligent system for recognizing human activities, advancing wearable health electronics.
