AI for Cattle Monitoring

The Challenge

Manual surveillance undertaken by feedlot staff is labour intensive, and subtle signs showing distress or disease among animals can be difficult to detect. Furthermore, regular veterinarian checks aren’t guaranteed to identify issues during the earliest stages. The manual intervention to help cast animals can also be risky for both the animal and the handler.

The Solution

This project aims to deliver a fully automated visual surveillance system using state-of-the-art AI technologies tailored for feedlot operations, and designed to provide continuous, accurate, and automated animal health & welfare monitoring. It involves a collaboration between RMIT University, Harmony Agriculture and Food Company Pty Ltd and Food Agility.

By leveraging advanced deep learning techniques, the system can handle the inherent variability in animal appearance and behaviour, offering early disease detection, efficient handling of cast animals, and adaptability to changes over time.

This ensures enhanced animal welfare, reduces economic losses, and strengthens the social license of the meat production industry.

Outcomes

This project will deliver a system that will:

• Use deep learning to adapt to animal appearance and behavior variations.
• Provide real-time anomaly detection for early disease identification and animal distress signals.
• Analyse long-term animal motion patterns to correlate with health and welfare.