Ultra-flexible Transparent Self-powered Triboelectric Sensors for Eyelash-Guided Human-Machine Interaction

• • Classification accuracy for eyelash guidance signals exceeds 98.5% using a CNN-LSTM hybrid model, enabling reliable, low-latency human-machine interaction without optical tracking hardware. • • The PF-TENG device employs PDMS-FDTS thin films, achieving ultra-flexibility and transparency that permit attachment to eyeglasses without visual obstruction, a critical requirement for continuous driver monitoring. • • Real-time fatigue detection capability was validated by monitoring ocular states, with the system accurately identifying fatigue signs and feeding back driver fatigue levels, directly addressing safety-critical scenarios in intelligent driving. • • The system eliminates dependence on video recording, infrared tracking, and image recognition, reducing hardware complexity and power consumption while maintaining high classification performance, thus offering a cost-effective alternative to conventional eye-tracking systems.