AI-Empowered Livestock Monitoring in Captive or Free-Range Breeding: A Review

Timely health monitoring is crucial for preventing livestock diseases throughout the breeding process. Conventional Livestock Health Monitoring (LHM) relies on manual efforts, achieving acceptable accuracy but suffering from low efficiency. Modern LHM integrates artificial intelligence to enhance both accuracy and efficiency in monitoring practices. This study reviews and evaluates over 200 research papers on intelligent monitoring in livestock, focusing on two breeding scenarios: Captive and Free-range. It examines various species, including pigs, cattle, sheep and horses and conducts a comparative analysis of three key tasks: livestock identification, growth estimation and behavior/posture detection. By highlighting common challenges and technical requirements across species, the study provides insights for selecting suitable monitoring technologies and designing tailored systems for each breeding scenario. The scenario-based classification enables customized LHM solutions that improve versatility and adaptability across different farming environments. The study reveals common challenges in cross-species monitoring and outlines future research directions to address technical limitations and practical implementation barriers in AI-driven livestock health management.