| Abstract: |
Pre-slaughter handling of cattle has become an area of critical concern within contemporary livestock production systems due to the impact on meat quality, food safety and regulatory compliance for animal welfare. Standard stress assessment involves invasive physiological sampling, or ranking behavioral responses, both of which are labor-intensive, subjective and may compound animal distress. The current review focuses on the latest non-invasive remote sensing technologies in combination with IoT sensor networks, AI and IRT in novel advances of stress detection system of bovines at the point of slaughter. Drawing upon data from peer-reviewed literature in animal physiology, thermal imaging, machine learning, and precision livestock farming, this study assesses there where the technology currently stands, highlights essential methodological voids, and describes a proposed conceptual AI-IRT-IoT framework for the scalable implementation of this technology in abattoir settings. Conclusions The review studies support the consistence of variations of body surface temperature, especially in ocular, nasal and cervical regions as valid biomarkers for acute psychological and physiological stress in bovines. Convolutional neural networks (CNNs) and ensemble classifiers have shown over 90% classification rates in controlled environments, making them suitable candidates for thermal pattern recognition (TPR) and automated TPR tasks. The IoT-supported edge computing architecturally has also enhanced real-time data acquisition and wireless data transmission while avoiding any setbacks with animal handling. The paper ends with discussions for further research progresses such as standardization of thermal protocols, cross-breed validation and multi-modal sensor fusion to boost up the robustness of the framework. This review provides a baseline reference for researchers, veterinarians and regulatory bodies. Solution technology developers can benefit from the insights provided to guide future selection and application of technology for automated pre-slaughter animal welfare assessment. |
