File(s) under embargo
until file(s) become available
FABRICATION OF SELF-POWERED DEVICES AND THE APPLICATION IN HUMAN-COMPUTER INTERFACE
The development of computation resources, sensing technologies and artificial intelligence has been the critical driving force for the development of next-generation smart devices in this Internet of Things (IoT) era. The requirement of sustainable and reliable sensing operation is becoming more and more important due to the exponential increment in the number of sensors with different functions deployed everywhere. Such a sensing network provides a more convenient way for connection and communication, making the boundary among humans, machines, and the external environment gradually blurred. Self-powered sensors which can directly utilize the external input energy to drive the operation of their own are highly desirable compared with the current sensing technology with a limited life cycle caused by power shortage. The invention of Triboelectric Nanogenerators (TENG) and Piezoelectric Nanogenerators (PENG) provides a novel direction for self-powered sensing technology. Both TENG and PENG can directly convert mechanical energy to an electrical signal, which could be further processed and understood by computers and humans.
In this dissertation, the research efforts have led to the design and fabrication of self-powered devices as well as system integration in the Human-Computer Interface (HCI). Materials modification was carried out to boost the performance of TENG output. A Piezotronic device using new semiconductor material, Tellurium, was fabricated and its fundamental charge transport characteristics were carefully studied for a new understanding in piezotronics. Beyond materials science, a system-level demonstration of using TENG for HCI application was also carried out. With the help of artificial intelligence technology such as machine learning and deep learning, more in-depth information was successfully extracted from the general TENG signal. The combination between Finite Element Analysis (FEA) and deep learning provided a more powerful platform for the development and verification of TENG-based sensing devices with improved working reliability. The presented concepts and results in this dissertation show the potential for the implementation of novel self-powered sensing technology in the future development for smart sensors, virtual/augmented reality (VR/AR) and other HCI-related areas.