Abstract
The global livestock production system faces significant challenges for sustainable development, including feed resource shortage and environmental pressures. Precision animal nutrition is crucial in addressing these challenges, in which the mathematical model is an indispensable tool. The traditional mathematical models exhibit certain limitations, particularly in accommodating the emerging demands of precision nutrition and feeding for individuals. New technologies, especially big data and artificial intelligence (AI), have shown great potential to mitigate the above shortcomings. This review has summarized the current landscape and applications of big data and AI-powered modeling in animal nutrition and feeding, covering techniques including intelligent data acquisition, in vitro kinetics and multi-omics data mining, data augmentation, advanced and explainable machine learning algorithms, multi-objective and heuristic algorithms, and life cycle assessment-based sustainability evaluation with case studies in pigs and alternative feed ingredients. Furthermore, this review has introduced the next-generation model techniques, including those based on large language models, multi-agents, and embodied AI robots, depicted the potential translation of the advancements from animal nutrition to human health, and discussed the limitations of AI-powered modeling techniques. These pioneering techniques will provide new tools and paradigms for research and practices in animal nutrition and further promote animal husbandry's sustainable development.