2024-04-17 ノースカロライナ州立大学(NCState)
＜関連情報＞
- https://news.ncsu.edu/2024/04/nc-state-researchers-use-machine-learning-to-create-a-fabric-based-touch-sensor/
- https://www.cell.com/device/fulltext/S2666-9986(24)00162-5
スマートファブリックのためのクリック可能な刺繍トライボエレクトリックセンサー A clickable embroidered triboelectric sensor for smart fabric
Yu Chen,Yali Ling,Yiduo Yang,…,Bao Yang,Xiaoming Tao,Rong Yin
Device Published:April 12, 2024
DOI:https://doi.org/10.1016/j.device.2024.100355
Highlights
- A flexible 3D TENG embroidery sensor for human-machine interface is presented
- A microchip-embedded machine learning model was used for signal recognition
- Dual-channel embroidery sensor with different “gaps” was developed
- A sensing array can be constructed using multiple embroidery sensors
The bigger picture
Most interactive devices are constructed from rigid materials that contrast with the soft and comfortable texture of textiles. Smart textiles, harnessing the potential of flexible electronic technology, offer a promising avenue for the development of comfortable wearable interfaces. However, most flexible electronic devices often still differ from traditional textiles in structure and construction, hindering their integration and scalability. Here, we present an embroidery-based sensor that can achieve touch interaction. By integrating two triboelectric yarns with conventional fabric using a 3D embroidery pattern, a stable triboelectric signal output is achieved. Coupled with machine learning, the embroidery sensor can recognize simple finger gestures for control interfaces. Multiple embroidery sensors can be integrated onto a single piece of fabric in a grid formation for enabling more complicated sensing applications and control interfaces.
Summary
Textile-based human-machine interfaces need to seamlessly integrate electronics with conventional fabrics. Here, we present an embroidery-based device that transforms conventional fabric into a “clickable” button. The device is realized through the integration of dual triboelectric yarns using an embroidery pattern that enables a 3D structure. The design can be customized and optimized by adjusting the gaps between the triboelectric yarns for the needed triboelectric output and other performance metrics, such as consistent contact and separation for the clicking mechanism. Machine learning algorithms are used for signal identification of a diverse range of pressing and swiping gestures on the embroidered device.
Graphical abstract
