Coughs, pecks, slurps.
Average, thin, fat.
Those are some of the real-life sounds and sights that are part of the reality of “precision livestock farming,” according to a European expert who spoke at the Manitoba Swine Seminar.
They can be the key to catching problems quickly before a human would spot them and fixing systems and helping animals before a small problem becomes a major one.
“It can be part of an early warning system,” said Tomas Norton, an Irish researcher, systems developer and professor.
Norton said early precision livestock production systems are being developed in Europe using microphones, cameras and computer analysis to spot problems.
Microphones can be used to pick up a wide array of barn noises, including pig coughs. A computer algorithm can sort through the coughs to alert a farmer if there is anything unusual about the coughs in the barn, which could reveal a disease that is about to break out.
It sounds simple, but creating a system that actually works isn’t easy. For instance, barns are not quiet places.
“You have a lot of noise in a typical pig building … ventilation systems, doors banging, pigs banging things, grunting, coughing, screaming, etc.”
As a result, a computerized system needs to be able to separate the sounds and extract only coughs.
However, all coughs are not the same. Pigs generally cough when they get up, or for various other reasons. Only sick coughs reveal a potential problem.
“We need an algorithm that can extract (the right sounds),” said Norton. “(Fortunately), it’s not that complicated.”
The same goes for cameras that can be set up over pens. Visual data can be analyzed with another algorithm to identify overweight or underweight pigs, which can also be a sign of trouble.
Precision systems like this don’t just require cameras, microphones and computer systems; they also require a human overseer who doesn’t need to be on the farm being monitored.
“We also need somebody who’s listening to what’s going on in the house,” said Norton. “We want to do stuff in real time. We want to send the information as quickly as possible to the farmer.”
The gains can be significant not only for efficiency but also for sustainability. Catching a disease before it becomes a problem stops pig deaths and growth losses early and reduces antibiotic use.
The systems can even spot basic mechanical problems such as plugged feeders and water sources.
Farmers often don’t receive feedback on the health or status of their animals until it is too late to act.
“The problem is that the farmer can’t act on that information,” Norton said about occasional assessments or reports from slaughterhouses.
That’s what precision livestock farming is all about.
“There is a possibility to act on it so you don’t decrease the efficiency of your production,” he said.