Microwave Moisture Content Sensors with Embedded Predictive-Model

This contribution presents microwave solutions implementing stand-alone portable Moisture Content (MC) sensors, which can be adopted in precision agriculture and health care contexts. A first solution is based on a microwave self-oscillating antenna loaded by wood, whose dc bias conditions variations are related to the antenna loading conditions, represented by the wood in different moisture conditions. A low-computation Machine-Learning (ML) algorithm is developed, enabling the system to operate in real time and as a stand-alone system. A second solution proposes a portable microwave sensing system capable of automatically acquiring RF scattering parameters with fast ML processing.