Precision Agriculture Approach for Potato Disease Detection Using ANFIS and Environmental Analytics

India, the world's second-largest potato producer with over 60 million metric tonnes annually, faces significant threats from fungal diseases like late blight and early blight, exacerbated by environmental factors such as temperature, humidity, and soil moisture. This study proposes an ANFIS (Adaptive Neuro-Fuzzy Inference System)-based model for real-time potato disease prediction and classification into Healthy, Early Blight, and Late Blight categories. Sensor-acquired data (temperature, humidity, soil moisture, rainfall, irrigation, crop spacing) is normalized, augmented with a Temperature-Humidity Index (THI), fuzzified, and processed through a two-layer ANFIS architecture for risk assessment and prediction, followed by confidence evaluation and defuzzification. The model is trained, tested, and validated using a structured dataset, and its performance is evaluated using metrics such as RMSE and classification accuracy, achieving an accuracy of approximately 95% with low prediction error. The results highlight the system’s capability to handle uncertainty and variability inherent in agricultural environments. Overall, this work contributes to the advancement of precision agriculture by offering a scalable, adaptive, and data-driven decision-support system that facilitates efficient resource management, improves crop productivity, and supports sustainable farming practices.