Advancements in machine learning applications in poultry farming: a literature review

Integrating machine learning (ML) in poultry science presents transformative opportunities for optimizing production, enhancing animal welfare, and improving disease management. This review explores the current landscape of ML applications within the poultry sector, encompassing growth prediction, disease detection, behaviour analysis, environmental monitoring, and productivity enhancement. ML techniques, including artificial neural networks, random forests, and deep learning, have demonstrated high predictive accuracy and adaptability in handling complex and nonlinear poultry data. Key innovations include the automated detection of diseased birds via image and audio recognition, the prediction of growth and body weight using environmental and nutritional parameters, and the assessment of animal behavior and welfare. This review also highlights challenges related to data quality, model interpretability, infrastructure limitations, and the generalizability of models across different poultry systems. Despite these hurdles, case studies reported in the literature demonstrate tangible benefits in productivity gains and early disease mitigation through ML applications. Moreover, the emergence of real-time sensing technologies and Internet of Things devices enables more granular data collection, further enhancing ML’s potential impact. Future strategies include fostering closer collaboration between data scientists and poultry specialists, developing explainable machine learning models, and integrating these models into decision-support systems to better assist farmers. The article advocates for scalable, ethical, and transparent ML solutions that align with both commercial viability and animal welfare goals. Overall, ML serves as a promising frontier for addressing the complex biological and operational dynamics of modern poultry farming.