Gradient Porous Flexible Pressure Sensors with the Relay Effect for High-Accuracy Braille-to-Speech Recognition.

The development of highly sensitive, wide linear-range flexible pressure sensors is crucial for practical applications in human-computer interaction, physiological signal detection, and motion monitoring. However, traditional flexible pressure sensors often suffer from limited compressibility in their structural design, resulting in rapid saturation of the detection range and low sensitivity, which hinder their commercial viability. In this study, we propose a straightforward electrospinning method to fabricate a thermoplastic polyurethane/carboxyl-functionalized multiwalled carbon nanotube/polydimethylsiloxane (TPU/c-MWCNTs/PDMS) piezoresistive sensor featuring a gradient porous structure. Owing to the gradient porous structure and PDMS surface treatment, the device exhibits significantly enhanced sensitivity and an extended linear range. The optimized sensor demonstrates a broad linear detection range of 0-200 kPa, a high sensitivity of 78.41 kPa-1, a low detection limit of 8 Pa, a fast response time of 80 ms, and exceptional durability over 10,000 loading/unloading cycles, enabling precise detection of complex human motions. Furthermore, by integrating a deep-learning algorithm, we developed an intelligent Braille-to-speech recognition system capable of accurately translating Braille alphabets and sentences into speech with recognition accuracies of 94.3 and 93.9%, respectively. This innovative structural design not only addresses the trade-off between sensitivity and linear range in flexible pressure sensors but also paves the way for advanced assistive technologies to enhance the quality of life for individuals with disabilities.