Multi-Stage Data Processing for Enhancing Korean Cattle (Hanwoo) Weight Estimations by Automated Weighing Systems.

Area of Science:

• Agricultural Engineering
• Animal Science
• Data Science

Background:

• Automated weighing systems (AWS) are crucial for modern smart livestock farming.
• High measurement variability from environmental factors and animal activity challenges AWS accuracy.
• Accurate weight monitoring is fundamental for effective cattle management.

Purpose of the Study:

• To develop and validate an algorithm for enhancing the reliability of steer weight measurements from AWS.
• To ensure automated weight measurements closely align with actual cattle body weight.
• To improve data-driven precision feeding strategies in livestock management.

Main Methods:

• A three-stage algorithm was developed: outlier detection/removal, weight estimation, and post-processing adjustment.
• Tukey's fences were used for outlier detection, with mean-based estimation.
• Post-processing incorporated daily weight gain recommendations from the National Institute of Animal Science.

Main Results:

• The best-performing algorithm achieved a root mean square error of 12.35 kg.
• The developed algorithm demonstrated an error margin of less than 10% for individual steers.
• The system reliably measured steer weight, confirming its suitability for smart livestock farming.

Conclusions:

• The developed algorithm significantly enhances the reliability of AWS for steer weight measurement.
• Accurate, automated weight data supports intelligent precision feeding and data-driven livestock management.
• This technology contributes to the advancement of smart livestock farming practices.