Decision Making in Precision Agriculture Using IoT and Machine Learning: A Focus on Rice Diseases

Rice is one of the staple foods consumed by billions, which makes it extremely susceptible to diseases that can decimate yields and quality. Precision agriculture uses the Internet of Things (IoT) and machine learning (ML) to enhance its capability in real-time disease monitoring and decision-making. Sensor data from paddy fields can be aggregated into machine learning models to guide early detection of diseases, such as rice blast, bacterial blight, and sheath blight. The paper presents an IoT-based approach that utilizes environmental sensors to capture critical parameters such as humidity, temperature, rainfall and soil moisture. This data is further processed using several machine learning algorithms, including Decision Tree, Random Forest, Support Vector Machine, XGBoost, LightGBM, Artificial Neural Network (ANN), Logistic Regression, and K-Nearest Neighbors for disease prediction and decision support. XGBoost was the best of all of them. Results show improved accuracy in the detection of diseases using XGBoost Algorithm and timely intervention, which could boost rice productivity. The implications of such work are therefore as immense, ranging from sustainable farm practices and reduced pesticide application to reducing the cost of disease control. Thus a Precision Agriculture based decision support system is proposed which will help the farmers take necessary action prior disease dispersion.