Modelling of Machine Learning-based Poultry Farm Environmental Monitoring System with Internet of Things

Abstract

This research presents an environmental monitoring system designed for poultry farms by integrating Internet of Things (IoT) technologies with Machine Learning (ML) algorithms. Conducted in Uyo, Nigeria, from June 2023 to April 2024, the study collected and labeled 8,412 samples of key environmental parameters—including ambient temperature, humidity, air quality, and the average age of chickens in weeks—with guidance from poultry farming experts. These data were used to train and validate three Machine Learning models: Random Forest, K-Nearest Neighbors, and Support Vector Classifier and each was evaluated for accuracy and reliability. The Random Forest model achieved the highest accuracy at 98%, outperforming the other models and indicating its robustness in environmental monitoring tasks. The study highlights the potential of IoT and ML technologies to improve farm productivity and animal welfare through proactive, data-driven management strategies. This work advances existing solutions by enhancing precision through data-driven insights, leveraging IoT to capture real-time data and ML algorithms to analyze environmental conditions with exceptional accuracy unlike traditional monitoring methods that rely solely on manual checks.