3 Technologies for Finding Your Most Profitable Cows
The most successful and sustainable dairy producers are looking for ways to pinpoint profitability, and perhaps one of the greatest opportunities that lies ahead is in technology that can predict with even greater precision which cows are truly the most profitable in the herd.
Kyle O’Brien, owner of Michigan Dairy Tech and Western Dairy Tech, sees the challenges of market volatility and tight margins that producers are facing, as well as the opportunities for automation, data and machine learning to empower decisions on the individual cow level that lead to the overall ability to run a better dairy farming business.
“They need to find the most profitable cows,” O’Brien says. “And it is probably not the cow making the most solids.” With a background in computer technology, O’Brien has worked directly with dairy producers for more than two decades to bring networking, software and automation to the farm. He believes we have yet to harness the possibilities of what integration can achieve.
“When you combine all of these cow data points with computers automating the decisions of how to handle each cow, that is when this gets powerful,” O’Brien adds.
These are the three technologies that he believes house the untapped potential for per-cow profitability:
- Automation – “Automation is supposed to control the things you can control. That makes sense in a manufacturing environment,” O’Brien says. “One of the benefits of automation is using data from two ‘dumb’ systems to make a smart decision elsewhere.”
- Big data – “Big data is the ability to collect data on every aspect of your operation,” he adds. “The more data points you have, the more informed decisions you can make.”
- Machine learning – Machine learning takes automation and big data to the next level. “Automation used to be a binary operation. With machine learning, it can now make inferences in real time,” he explains. Likewise, machine learning simplifies big data. “Big data without machine learning is just a pile of data, “ he adds. “There is too much information for the human brain to weed through and find patterns.
So what could be possible for a more profitable cow with the combined power of automation, big data and machine learning? Decisions based on real-time biological data, like using the individual cow’s energy balance as an indicator for breeding versus a standard voluntary waiting period. Or accurate dry matter intake information to prompt pen moves from the fresh to high production group, instead of moves made at a pre-determined days in milk (DIM).
“If the humans can trust the machines, I feel they can do better for the cows,” O’Brien says. “It’s our job in the technology sector to make a product they can trust.”
