<![CDATA[Hydroponic farming using the Nutrient Film Technique (NFT) system is widely implemented due to its efficiency in nutrient management and water use. Spinach and lettuce are leafy commodities widely cultivated using this system because they have a relatively short growth cycle and high economic value. However, determining harvest time is still often done manually based on experience, potentially leading to inaccurate decisions that impact the quality and quantity of production. This study aims to develop a prediction model for harvest time for hydroponic spinach and lettuce plants based on Artificial Neural Network (ANN) by utilizing environmental and physiological parameters of the plant. The parameters used include water temperature, air humidity, light intensity, pH, Electrical Conductivity (EC), and plant age. The dataset used consists of 1,200 observation data of NFT hydroponic cultivation results from January to July 2025. The data went through a preprocessing stage in the form of cleaning, normalization, and dividing training data and test data with a ratio of 80:20. The ANN model was built using the backpropagation method with training parameter optimization. Data was obtained from plant growth monitoring, then normalized and divided into training and test data. Test results showed a prediction accuracy of 92.8% based on MAPE, MAE, and R-squared. This model was implemented in a Streamlit-based web application to facilitate farmer use, making harvest timing more objective, measurable, and data-driven.]]>
Copyrights © 2026
