Optimizing water condition in tropical brackish water aquaponics: A novel internet of things and machine learning approach

Abstract

Abstract This research presents a novel water condition regulation system specifically designed for brackish water aquaponics in the Aceh tropical climates. The study utilizes cutting-edge sensor technologies, intelligent feedback mechanisms, and IoT-integrated precise water condition adjustments. The aim of this approach is to optimize critical water condition indicators including pH, salinity, and dissolved oxygen for shrimp cultivation while enabling the simultaneous growth type of vegetable plants. The methodology encompasses a comprehensive model training procedure, wherein Random Forest and Genetic Algorithms are applied on a computer to predict optimize water condition indicators from historical data. The optimize water condition indicators then will be utilized in real time control and inference by Internet-of-Things (IoT) system based on Arduino microcontroller and dispensing actuators. A forty-day experimental phase in a brackish water aquaponics setup was conducted and verified that IoT system has effectively preserved optimal water conditions to the growth of both fishes and leafy green vegetables. The findings of the research suggest that the proposed IoT regulation system has maintained pH stability, salinity and Dissolved Oxygen at optimize level. The finding also recommend that the lettuce, spinach, and kale have emerged as the most appropriate plants for brackish water aquaponic system. The integration of Random Forest and Genetic Algorithm with IoT technology designate a significant advancement in the field of brackish water aquaponics, providing intelligent alternatives for agriculture in Aceh coastal areas.