Enhancing Smart Farming through ANFIS IoT Framework for Sugarcane Nutrient Prediction

Nutrient management systems in precision agriculture remain limited in processing real-time field conditions, causing inefficiencies in fertilizer application, crop cultivation, and productivity. Sugarcane requires balanced macronutrients (N, P, K) and micronutrients (Zn, Mn, B) to achieve optimal growth and meet industrial quality standards such as a Brix level of 18%. However, the absence of integrated and intelligent monitoring systems often leads to nutrient imbalance, which reduces yield and affects product quality. This study aims to develop a nutrient prediction framework for sugarcane by analyzing macro-and microelement requirements using an ANFIS-based classification model supported by IoT sensor data. The research involved IoT-based data collection from multiple field nodes measuring soil pH, humidity, temperature, and air humidity. The workflow consists of data acquisition, preparation, model training, testing, and evaluation. The proposed ANFIS model was compared with machine learning algorithms including Random Forest, Ridge, Logistic Regression, K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Decision Tree, and Gradient Boosting to validate predictive performance. Evaluation using accuracy, precision, recall, and F1-score showed that the ANFIS model produced strong prediction results, achieving accuracy above 70% with a relatively low error rate. Its performance also demonstrated higher stability and consistency compared to the baseline classification algorithms.