A Modular Distributed IoT Architecture for Closed-Loop Hydroponic Monitoring and Control with Image-Based Plant Observation

The main unique contribution of the research reported is a modular, distributed IoT based hydroponic architecture that combines different functions together, rather than treating sensing, control and planting observation as distinct subsystems. The system consists of a Raspberry Pi supervisory controller with distributed ESP32 field nodes, providing multi-parameter sensing and closed-loop control to the parameters such as nutrient dosing, pH, water circulation, humidity, ventilation and lighting in a callable manner. It also contains a first time non-contact plant segmentation, projected plant area, and colour-based features with camera usage as well as MQTT/Modbus communication and persistent data and remote visualisation. End-to-end integration including sensing, communication and monitoring, logging and actuator control verified in prototype.