Researchers at NC State University have developed a fabric-based sensor that integrates 3D embroidery techniques with machine learning. This sensor enables users to control electronic devices through touch, making it a valuable tool for the growing field of wearable electronics. By incorporating this sensor into clothing.
The device consists of an embroidered pressure sensor and a microchip to process data. The sensor is triboelectric, using friction to generate electric charge. Yarns with positive and negative charges are incorporated into textile fabrics using embroidery machines.
Rong Yin emphasizes the importance of the 3D structure of the pressure sensor, which is crucial for its functionality as a triboelectric sensor. The sensor requires two layers with a gap in between, a challenging task when using embroidery typically limited to two dimensions.
Data from the pressure sensor is processed by the microchip, enabling it to give specific instructions to connected devices. Machine learning algorithms play a crucial role in ensuring the system functions seamlessly. The device must differentiate between various gestures and filter out unintended inputs caused by the fabric’s movements.
Researchers showcased input recognition through a music app connected to a sensor via Bluetooth. The app featured functions like play/pause, next/last song, volume control, and mute, each activated by specific gestures. The sensor was also used for tasks like setting passwords and maneuvering video games, showcasing its versatility.
