Design and Implementation of an AI-IoT Integrated Cloud Platform for Real-Time Poultry Environmental Monitoring and Decision Support

Abstract

The rapid digitalisation of livestock production systems has intensified the demand for affordable, scalable, and user-accessible smart farming solutions, particularly in poultry management, where environmental conditions directly influence animal welfare and productivity. This study presents the design, implementation, and real-world deployment of an AI–IoT integrated cloud platform for real-time poultry environmental monitoring and decision support. The proposed system integrates low-cost IoT sensor nodes for temperature, humidity, and ammonia monitoring, along with energy-efficient sleep scheduling mechanisms and machine-learning–based predictive analytics. Environmental data acquired by distributed sensor nodes is transmitted via Wi-Fi to a central processing unit and securely uploaded to the cloud, where it is stored, analysed, and visualised through an interactive Streamlit dashboard. A hybrid Random Forest–Support Vector Classifier model was employed to provide predictive insights into environmental risk conditions, enabling proactive intervention beyond conventional threshold-based alerts. The platform was deployed and evaluated in a real poultry farm environment, demonstrating reliable real-time monitoring, low-latency cloud connectivity, and improved environmental stability. Practical outcomes include enhanced decision-making for non-technical users, improved accessibility via an intuitive web interface, and measurable reductions in environmental stress indicators associated with poultry mortality. The results confirm the system’s effectiveness in democratising smart poultry farming and highlight its scalability potential for broader multi-livestock and precision agriculture applications.