Intelligent TDS Monitoring and Anomaly Detection in Hydroponic NFT Systems for Spinach – A Case Study
Pradnya Vishram Kulkarni and Vinaya Gohokar
This research focuses on the development of an AI-driven anomaly detection system for Total Dissolved Solids (TDS) in Nutrient Film Technique (NFT) hydroponics, leveraging IoT technologies. NFT is a soilless farming method where a continuous flow of nutrient-rich water supports plant growth. Maintaining optimal TDS levels at different growth stages is crucial for healthy plant development. The proposed system integrates sensors to monitor realtime TDS levels, with data collected using ESP32 and processed by a Raspberry Pi-4. Isolation Forest algorithm analyses TDS variations across different plant growth stages—germination, early vegetative, late vegetative, and harvest—to detect anomalies that may indicate nutrient imbalances. MimMaxScaler helps validate plant age and TDS requirements with given range of nutrients. The rule-based classification gives the age-wise anomaly is TDS scaling values in the instantaneous sample, together giving the system accuracy up to 90%. By automating anomaly detection, the system helps optimize nutrient dosing, reduce waste by 20%, and improve crop yield. The IoT-enabled framework autonomously monitors and analyses TDS levels, ensuring real-time anomaly detection and nutrient optimization without the need for manual intervention, offering a scalable solution for modern hydroponic farming. This research highlights the potential of AI-powered hydroponic systems in enhancing plant health, maximizing yields, and promoting sustainable agriculture through intelligent nutrient management.
Keywords: agriculture, IoT, hydroponics, automation