Abstract
The increasing global demand for efficient water management has underscored the importance of smart irrigation solutions in agriculture. This research introduces a Smart Irrigation System that integrates the Internet of Things (IoT) with an advanced Machine Learning (ML) framework to optimize water usage while ensuring sustainability. The proposed model employs an ensemble of Decision Tree Classifier (DTC) and Random Forest Classifier (RFC) algorithms to analyse critical environmental parameters, including soil moisture, temperature, pH value, and soil variants. Utilizing the meticulously processed Great Time dataset for training and evaluation, the model demonstrates exceptional applicability across diverse agricultural scenarios. Traditional irrigation models exhibit lower accuracy and limited adaptability to varying environmental conditions, creating a need for more robust and efficient approaches. Addressing this gap, the IoT-enabled system leverages real-time data from connected devices and advanced analytics to adapt to dynamic environmental changes. By offering precise irrigation scheduling, the proposed framework promotes resource-efficient water usage, contributing to sustainable farming practices. The ensemble model achieved an impressive accuracy of 98.7%, significantly outperforming conventional methods while maintaining computational efficiency. This study highlights the strength of combining IoT and ML to advance agricultural practices. Experimental outcomes emphasize the scalability, robustness, and reliability of the proposed model, presenting it as a viable solution to tackle water scarcity challenges and enhance crop productivity sustainably.