An IoT-driven predictive analytics framework for dynamic irrigation optimization in tomato cultivation

Addressing the critical challenges of global food security and water scarcity, we introduced an Internet of Things (IoT)-driven predictive analytics framework for dynamic irrigation optimization in tomato cultivation. Our primary objective was to develop a robust model that accurately estimates daily water requirements, with the aim of minimizing water consumption while concurrently maintaining optimal soil health. This framework leverages a uniquely comprehensive, experimentally controlled dataset from multi-sensor IoT deployments, covering environmental conditions (air temperature, humidity, CO₂, pressure) and key soil parameters (humidity, temperature, electrical conductivity). Through a rigorous data preprocessing pipeline and a tailored feature engineering approach, critical temporal patterns, inter-variable relationships, and insights from agronomic indicators like Growing Degree Days (GDD), alongside other dynamically derived features, were extracted. A two-part eXtreme Gradient Boosting (XGBoost) regression model, combining classification and regression, was developed and validated to precisely predict the daily water volume needed per hectare. The innovation of this work lies in its ability to harness complex historical IoT data to build a sophisticated intelligence layer for irrigation scheduling. By demonstrating the model's accuracy in identifying optimal water levels under varying conditions and achieving significant water savings in a simulated dynamic optimization, this research provides foundational data-driven insights that can inform highly effective precision irrigation strategies. The model achieved a high R² of 0.9476 and yielded a potential water saving of 50.84% in a simulated dynamic optimization compared to the model's raw prediction. Such intelligence empowers farmers to significantly reduce water waste and prevent harmful over-irrigation, leading to more sustainable and efficient smart agriculture, which is critical for enhancing crop resilience and resource efficiency in a changing climate.