Embedded Sensing System for Whispering-Gallery-Mode Optical Resonators

Whispering-gallery-mode (WGM) optical resonators in which light is confined by total internal reflection have found broad applications ranging from optical communications, microlasers, sensing, opto-mechanics to quantum optics. However, overcoming obstacles of practical applications for WGM resonators is absolutely a big challenge. First, we report the first realization of a compact WGM sensing system, which integrates a tunable laser, a current source, a temperature controller, a function generator, an oscilloscope, a photodiode detector, a testing computer with customized testing software, and a packaged WGM sensor into a phone-sized embedded system. Second, we deploy the WGM sensor in the Internet of things (IoT) sensor network. We demonstrate a WGM sensor based embedded IoT device. By connecting the Wi-Fi unit to the internet, a worldwide, real-time control of this system can be realized. We also conducted the aerial thermal mapping experiment by using this wireless sensing system. Third, we report a sensing model with high noise immunity using machine learning algorithm for WGM mode shift sensing. Furthermore, an artificial intelligence at the edge embedded device is designed for real-time inference of sensing results which is running the pre-trained neural network model. This platform is also integrated with a customized iOS application for system control and sensing result display via Wi-Fi antenna. Our algorithm and platform not only exhibit high noise immunity characteristics, but also potential flexibility in configuration for different sensing applications.