Enhancing Crop Yields in Urban Farming Systems Through Machine Learning

With a shared goal to enhance food security, governments implement various initiatives to increase food production and quality. Although abundant, traditional farming techniques pose the risks of agricultural runoff, soil quality degradation, and unfavorable weather conditions, which may result in crop failure. Machine learning, a subfield of artificial intelligence, can address several challenges inherent in traditional farming practices, including resource inefficiency and crop yield variability. This paper, therefore, thoroughly examines and discusses machine learning algorithms in areas of water and nutrient management for hydroponics, aquaculture, and aquaponics systems. Considering the findings from the literature review, avenues for integrating machine learning into an innovative container-based multi-loop aquaponics system with green internet of things and semi-transparent photovoltaics at Singapore Institute of Technology are also presented and described.