Comparative Analysis of Capacitive Sensors for Leaf Wetness Measurement

One of the biggest challenges to face in the 21st century is the exponential increase in the world population that brings numerous problems, such as overexploitation and the consequent depletion of global natural resources. In this context, water resources are subjected to unprecedented pressure as the water demand grows significantly, but its availability tends to decrease increasingly. One of the most impactful sectors is certainly agriculture, as it has almost all the available water resources and ranks as one of the leading global water consumers. With the goal of better managing these resources, Smart Agriculture was born, aiming to improve agricultural yield, increase productivity, and optimize the potential of arable land with the introduction of advanced technologies that allow the monitoring of plants in real-time in order to know their specific needs and avoid waste and excessive consumption. The first water management methods were based on using humidity sensors to indirectly examine the water status of the plant, starting with an analysis of the soil composition. However, this was not always sufficient, especially for deep-rooted plants, as the most superficial layer of soil studied by the sensor could present different characteristics from the underlying layers. For this reason, we moved to a direct measurement, introducing specific sensors applied directly to the plant to measure an important parameter called leaf wetness. Leaf wetness is the moisture on the leaf surface, which is essential in crop protection. Due to dew, rain, or irrigation, water on the leaf favors ideal conditions for developing fungal diseases. If leaf wetness is monitored, protection against infection can be done by optimizing treatment timing and minimizing pesticide applications. The objective of this study is a comparative analysis of two sensors that measure the water status of the plant: FylloClip, which connects to the plant via a clip mechanism, and a Leaf Sensor based on FDC2112, a capacitance-to-digital converter, which simulates the behavior of leaves. The research focuses on fundamental aspects such as sensitivity, non-invasiveness, and durability of the sensors themselves to guarantee accurate measurements and, at the same time, respect for the environment and sustainability. FylloClip is a leaf-mounted sensor that uses an LMC555 timer to measure plant capacitance. This timer generates signals with a controllable frequency that allows us to evaluate the capacity of the leaf indirectly by monitoring changes in the oscillation frequency. On the other hand, the Leaf Sensor is a two-sided capacitive sensor that uses FDC2112, equipped with two independent channels that allow it to acquire data from the two faces separately. The capacitive measurement comes from the variation of the dielectric constant between the two electrodes of the sensor. Both sensors are connected to an STM32WL55JC microcontroller, whose firmware is developed to allow communication via the LoRa communication protocol, wireless and long range, and subsequently installed inside a climatic chamber to be tested under controlled conditions, changing in particular temperature and humidity. Through this testing phase, it is possible to evaluate the performance of each sensor in different configurations, such as day-night cycles or extreme weather conditions, allowing us to highlight how this study can contribute to making the agricultural sector more sustainable and providing significant support to reduce water waste.