Early identification of animals in want of administration intervention is essential to maximise animal well being and welfare and reduce points with productiveness. Feeding conduct, captured by automated feeding techniques, can be utilized to observe the well being and welfare standing of particular person pigs.
Goal: Right here, we current a framework for monitoring feeding conduct of grow-finish pigs in actual time, utilizing a low-frequency radio frequency identification (RFID) system. Utilizing historic knowledge, an autoregressive linear mannequin for predicting every day time on the feeder was developed and utilized to detect anomalous decreases in feeding conduct related to well being standing of the pig.
Strategies: A complete of two,826 pigs had been individually monitored with our warning system over your entire grow-finish interval, and well being warnings had been in comparison with caretaker diagnoses.
Outcomes: The system detected 55.7% of the caretaker diagnoses, and on common these occasions had been detected 2.8 d sooner than prognosis by the caretaker. Excessive numbers of probably spurious well being warnings, generated by the system, may be partly defined by the shortage of a dependable and repeatable gold customary reference knowledge set.
Conclusion: Outcomes from this work present a strong foundation for monitoring particular person animals, however additional enhancements to the system are vital for sensible implementation.
Funk TH, Rohrer GA, Brown-Brandl TM, Keel BN. On-line feeding conduct monitoring of particular person group-housed grow-finish pigs utilizing a low-frequency RFID digital feeding system. Translational Animal Science. 2024; 8: txae051. https://doi.org/10.1093/tas/txae051
