Smart irrigation for soil moisture prediction using hybrid AI and fog computing

Irrigation control is essential for optimizing water usage in smart agricultural systems, especially where there is water scarcity. The fast development of IoT and fog computing brings up new opportunities for developing new and advanced systems that are both efficient and flexible. Further integration of fog computing with machine learning will bring more developments to smart agriculture systems. This paper has framed an irrigation supervision algorithm to predict irrigation needs in the smart agriculture system. The methodology uses a hybrid algorithm that incorporates Gradient Boosting (GB), Logistic Regression (LR) and Random Forest (RF). The system accurately forecasts the irrigation requirement by processing ecological information such as humidity, temperature, and time-based factors. 97.78% of accuracy illustrates the reliability of the resultant model, and initial processing of the data illustrates that information is concise and accurate. Recall, precision and F1- score are computed as metrics for the performance. The system is integrated into fog computing and machine learning to manage real-time data processing to achieve faster and more accurate predictions.