Deep Reinforcement Learning-Based Decision Support System for Optimizing Strategies in Precision Livestock Farming
The growing complexity of modern dairy farming requires innovative solutions to balance the simultaneous demands of increasing milk production, improving animal welfare, reducing operational costs, and minimizing environmental impact. Traditional rule-based management approaches, which rely heavily on historical averages, are increasingly inadequate for addressing the dynamic and individualized needs of livestock herds. ; ; This project proposes the development of an Artificial Intelligence (AI)powered Decision Support System (DSS) for precision livestock farming. The system integrates advanced deep reinforcement learning (DRL) techniques with neural network–based state encoding to optimize farm management strategies. By transforming high-dimensional sensor and management data into actionable insights, the platform continuously learns and adapts to improve feeding, breeding, milking, and health interventions. ; ; A data-driven simulation environment, constructed from real-world farm data, enables safe and efficient training and evaluation of the AI models. This approach ensures robust policy development without disrupting farm operations. The system is designed with a multi-objective optimization framework, balancing productivity, animal health, cost-efficiency, and environmental sustainability. ; ; Ultimately, this work aims to deliver a scalable, user-friendly decision support platform that enhances dairy farm profitability and sustainability. Broader impacts include contributions to global food security and alignment with the United Nations Sustainable Development Goals, particularly Zero Hunger and Climate Action. ;
Back to Current Students