Algorithm for estimating producer carcass yield

Summary

​The primary objective of this project was to estimate the producer carcass yield based on daily measurements of yield at the meat processing plant, which are derived from multiple producers. We found that it is feasible to estimate the average yields of individual producers from the ensemble measured by the processor. We tested the performance of the algorithm with simulated test data and real data from an Australian lamb meat processor and found that boning room yield for the top supplier to the processor was 83 ± 0.6%. The secondary objective of this project was to estimate the contribution of an individual sire group to carcass yield based on daily measurements of yield at the meat processing plant. Extending the methodology to deal with estimating meat yields of different sire families was also possible. This avoids the high phenotyping costs of measuring the yield of individual animals and will allow for more rapid genetic selection of yield and improved producer management of animals for yield. The algorithm will also allow for more rapid introduction of new sensor technology into meat processing plants. Future validation work will be required to benchmark the algorithm and develop a usable commercial product.