Internet-of- Things Behavior Monitoring System Based on Wearable Inertial Sensors for Classifying Dairy Cattle Health Using Machine Learning

In this work, an Internet-of- Things (IoT) system for monitoring dairy cattle behavior is developed using wearable inertial sensors and machine learning algorithms. A behavior recognition model is established using the K-Nearest Neighbors (KNN) algorithm to monitor feeding and movement behaviors, including standing, lying, walking, resting, feeding, and ruminating, with high accuracy above 80%. Based on the recognized behaviors from leg and collar sensors, a health classification model is further developed using the Support Vector Machine (SVM) machine learning algorithm, achieving an accuracy of up to 70%. By embedding these models into the developed IoT system, online real-time monitoring of dairy cattle's daily behaviors and health score is demonstrated, enabling health status alerting. This work lays the foundation for designing IoT systems for real-time monitoring of dairy cattle behaviors and health status classification towards cost-effective livestock farming.