A Signal Amplitude-Insensitive\nTriboelectric Touch\nPanel with a Significantly Reduced Signal Channel and Deep-Learning-Enhanced\nRobustness

The self-powered triboelectric touch panel has garnered\nconsiderable\nresearch attention due to its potential to reduce system energy consumption\nand its applications in human–machine interfaces, e-skin, and\nthe Internet of Things. Current methods for achieving triboelectric-based\ntouch positioning in an <i>M</i> × <i>N</i> detection pixel array typically require signal amplitude comparison\nacross at least <i>M</i> + <i>N</i> signal channels,\nthereby limiting lightweight design possibilities. In contrast, our\nnovel “resistor ladder” approach necessitates only 4\nsignal channels for touch positioning. This method leverages a lookup\ntable correlating touch positions with amplitude ratios from different\nchannels, rendering it insensitive to signal amplitude and significantly\nenhancing robustness. We fabricated a transparent touch panel using\nPET tribomaterial, where the surface roughness was enhanced through\nplasma treatment. The panel successfully demonstrated touch positioning\nfor 128 taps within a 4 × 4 pixel detection array and sliding\npositioning using a predefined lookup table. To further enhance device\nrobustness, a 2D convolutional neural network was implemented, which\nachieved an impressive touch positioning accuracy of 97.7% even under\nartificially introduced signal defects. This study represents an initial\nexploration of amplitude-insensitive touch and sliding positioning\nmethods, significantly reducing the number of required signal channels\nand enhancing the robustness of triboelectric touch panels.