Development of an artificial intelligence-supported compost tea production system integrated with greenhouse automation

This doctoral thesis focuses on the development of an innovative compost tea production system that can be integrated with greenhouse automation systems, aiming to support sustainable agriculture and reduce chemical fertilizer usage. The study employs a multidisciplinary approach, combining agricultural machinery, environmental engineering, and information technologies. The developed system incorporates sensors, actuators, and automation software that continuously monitor and control the compost tea production process. Critical parameters such as pH, electrical conductivity (EC), dissolved oxygen, temperature, turbidity, and total dissolved solids (TDS) are measured and recorded in real-time. Three distinct experiments (DP1, DP2, DP3) were conducted to determine the effects of different production strategies on compost tea quality. These experiments elucidated the impacts of recirculation pump, blower, and peristaltic pump operation durations on the physicochemical properties of compost tea. One of the innovative aspects of the system is the integration of a machine learning model to optimize the production process. The developed model demonstrated the ability to determine optimal production conditions by learning the effects of various production parameters. Furthermore, the system possesses the capability to integrate with existing greenhouse automation systems. This integration facilitates more widespread and effective use of compost tea in agricultural production. In conclusion, the developed system offers a more controlled, optimized, and efficient process for compost tea production, presenting an alternative solution for reducing chemical fertilizer use. Future studies should include experiments with different compost materials and nutrients, further development of the machine learning model, and testing the system in larger-scale applications.