Deep Learning‐Assisted Fingerprint‐Inspired Flexible Pressure Sensor for Tension Monitoring

Full Text in Publisher | Download Paper

This study addresses the challenge of real-time tension monitoring in high-speed carbon fiber production, where traditional sensors cannot track dynamic tension variations across wide fiber bundles. Inspired by human fingerprints’ tactile sensing mechanism, the research team developed a fingerprint-inspired flexible pressure sensor that seamlessly conforms to roller surfaces, forming a distributed monitoring array. The sensor exhibits high sensitivity and wide linear range, enabling precise detection of dynamic tension changes in multi-bundle carbon fiber arrays. By integrating the sensor array with an end-to-end deep learning model, the system achieves a classification accuracy of 96% for abnormal tension events. This approach provides an innovative solution for real-time intelligent tension monitoring in carbon fiber manufacturing, promising to overcome critical bottlenecks in large-scale production and opening new applications for flexible electronics and smart manufacturing sensors.