Highly sensitive, breathable, and superhydrophobic dome structure nonwoven-based flexible pressure sensor utilizing machine learning for handwriting recognition.

Wearable devices that incorporate flexible pressure sensors have shown great potential for human-machine interaction, speech recognition, health monitoring, and handwriting recognition. However, achieving high sensitivity, durability, wide detection range, and breathability through cost-effective fabrication remains challenging. Through ultrasound-assisted modification and impregnation-drying, dome-structured nonwovens/rGO/PDMS flexible pressure sensors were developed. The sensor exhibits high sensitivity (up to 0.65 kPa in the range of 0-1.12 kPa), rapid response/recovery times (73/98 ms), a wide detection range (0-202 kPa), and superhydrophobic properties with a water contact angle of 166°. Additionally, it demonstrates excellent breathability (514.8 mm/s) and stability (>9000 cycles). With these excellent properties, the sensor is able to detect different pressure signals, allowing encrypted information to be transmitted; it can also be used for health monitoring and motion detection. Additionally, machine learning technology is also successfully used to recognize various handwritten words, with a 94 % accuracy rate in all cases. Consequently, the developed flexible pressure sensor has a wide range of potential applications, including information encryption, medical monitoring, motion detection, human-computer interaction, and handwriting recognition.