The assessment of sheep in saleyards by agents, buyers and market reporters is subjective; that is the estimate of liveweight, fat score & carcase weight is a best guess.
Given the importance of accurate assessment, especially to buyers and reporters, wouldn’t it be good if we could add some real time objective measurement to enhance this important sale function at major sheep and lamb auction markets?
This is the first in a regular series of agtech insights, aiming to provide a vision on the future of technology in agriculture.
There is always conjecture and debate from buyers and market reporters around a pen of sheep or lambs. “What is the average weight”, and then more importantly “what will this translate into as a carcase weight”?
Cattle in Australia are predominately sold with live-weight published; this has not progressed to the saleyards transactions of sheep and lambs. Logistically it is more difficult to run the thousands of sheep and lambs sold at major saleyards across a weighbridge.
One solution to improving on the hoof assessment would be to use “a machine vision system”, or in simple terms a camera.
American Society of Agricultural and Biological Engineers, St. Joseph, Michigan, in 1994 set up a booth with a video camera connected to a computer.
This method allowed the “mean central projected area” of pigs to be calculated and related to their live weight. This produced a live weight accuracy of ±0.9 kg.
Advances including Electronic Identification (EID) of sheep in Victoria are providing the necessary tracking of animals, so it is a natural progression to integrate this with other “machine learning” to improve the assessment of sheep.
“What is the average weight”, and then more importantly “what will this translate into as a carcase weight”?
In the initial stages, any photographic derived sheep weights or carcase attributes would surely be posted as “guidance” only.
However, if we look at the progress of the objective measurement of wool, we shouldn’t rule out that in the future the information provided from a camera could become the benchmark for determining the “fair value” of live sheep.
We would expect resistance to any method that may provide improved accuracy of weight estimates, especially we should expect opposition from those who currently consider they have an existing advantage such as livestock buyers.
What if this use of a camera also provided thermal imagery that identified disease, dog bites or other carcase imperfections? Again, this would assist in determining “fair value” based on saleable meat yield but confirmed in the live sheep! This information would be valuable to the buyer, identifying possible meat yield losses prior to slaughter.
Machine vision is already measuring cattle & pigs on the hoof; providing updated weight reports in real time.
While applying this innovation to sheep will present unique challenges, the science & technology in this space is moving rapidly.
This addition information could provide a quantum leap in improving accuracy of live sheep assessment prior to the carcase hanging in the chiller, and assist the industry along the path to “fair value” calculated with objective data alongside subjective assessment.