PlumX Metrics
Embed PlumX Metrics

Electrowetting Enhanced Analog Self-Powered Touch Panel with Deep Learning Enabled Digit Recognition

SSRN, ISSN: 1556-5068
2024
  • 0
    Citations
  • 62
    Usage
  • 0
    Captures
  • 0
    Mentions
  • 0
    Social Media
Metric Options:   Counts1 Year3 Year

Metrics Details

  • Usage
    62

Article Description

Human-machine interfaces, with contact positioning as a critical functionality, are driving the development of next-generation intelligent wearable electronics. Despite the energy-saving advantages of self-powered triboelectric touch panels, their widespread application is hindered by challenges such as poor signal consistency, limited long-term signal strength enhancement methods, and the complexity of managing numerous signal channels. In this study, we propose an analog triboelectric contact positioning method that significantly reduces signal channel requirements and eliminates amplitude dependence by using a resistance bar design to correlate the touch position with the voltage ratio across just four signal channels. Additionally, a fluorinated acrylate-polyethylene terephthalate double-layer structure is introduced to facilitate electrowetting-assisted charge injection, enabling efficient charge storage due to the enhanced charge penetration characteristics. Our results show that the triboelectric performance decreases by only 26% after droplet wetting and by 15% after 5000 contact-separation cycles, demonstrating excellent wetting resistance and long-term stability. Furthermore, we apply deep learning techniques to process data from the four-channel touch panel, enabling handwritten digit recognition with an accuracy of 98.5% among five digits through cross-channel feature extraction and comparison, and offering insights into the integration of artificial intelligence in triboelectric signal processing.

Bibliographic Details

Wei Xu; Yuzhe Gu; Shiyu Zheng; Yuxin Wang; Yang Li; Qingyun Chen; Qingying Ren; Wei Li; Wei He; Yinghai Zhang; Qiumeng Chen; Yannan Xie

Elsevier BV

Multidisciplinary; electrowetting; triboelectric touch panel; charge injection; digit recognition; convolutional neural networks

Provide Feedback

Have ideas for a new metric? Would you like to see something else here?Let us know