Sound monitoring has been widely used in the field of the Internet of Things (IoT), in which the sensors are mainly powered by batteries with high power consumption and limited life. Here, a near-zero quiescent power sound wake-up and identification system based on a triboelectric nanogenerator (TENG) is proposed, in which the sound TENG (S-TENG) is used for ambient sound energy harvesting and system activation. Once the sound intensity is higher than 65 dB, the converted and stored electric energy by the S-TENG can wake up the system within 0.5 s. By integrating a deep learning technique, the system is used for identifying sound sources, such as drilling, child playing, dog barking, and street music. In the active mode, the sound signals are recorded by a microelectromechanical systems (MEMS) microphone and then sent to a remote computer for sound recognition through a wireless transmitter within 2.8 s. In the standby mode, the ambient sound is not enough to wake up the system, and the quiescent power consumption is only 55 nW. This work provides a triboelectric sensor-based ultralow quiescent power sound wake-up system, which has shown excellent application prospects in smart homes, unmanned monitoring, and the Internet of Things.
Keywords: near-zero quiescent power; sound identification; sound wake-up system; triboelectric nanogenerator; triboelectric sensor.