Systematic Literature Review on Hydroponic Automation Models: IoT and Machine Learning Approaches

The significance of Hydroponic automation is imperative as it can increase productivity while also being environmentally sustainable. Seeing how IoT and Machine Learning work hand in hand allows us to devise better ways to monitor and cultivate plants in real time. This paper presents a Systematic Literature Review (SLR) of modern hydroponic systems, with a focus on models that are IoT- and ML-driven. The focus of the inquiry is to explore the primary methodologies utilized in practice, how well the models of prediction work, and the fundamental variables influencing the growth of hydroponic crops. Using the PRISMA guidelines, we identified, reviewed, and synthesized eighteen studies relevant to the topic from prominent databases. The findings reveal that 72% of studies investigate both IoT and Machine Learning, 22% address IoT-based automation unrelated to ML, and 6% address ML-based automation without IoT. The most commonly cited predictive models include ARIMA, Random Forest, Support Vector Regression, and Deep Learning. These models demonstrate high accuracy in predicting shifts in environmental conditions. The report also notes that among the many controllable parameters in a hydroponic system, temperature and pH are the two most important to control. This indicates their importance for crops grown in hydroponic systems. Inadequate datasets, limited computing resources, and the difficulty of assembly remain issues in data-centric automation. This document surveys the state of the art in automated hydroponics and offers suggestions for further research, especially on improving the application of IoT machine learning to real-time control of hydroponic systems.