Milking data to harvest information
Purdue’s Department of Animal Sciences assists Homestead Dairy by turning data into useful information
Nearly 75 years ago, the Houin family began a simple home dairy with nine cows. Two generations later, Homestead Dairy is a technological leader with thousands of cows, teams of robots, and an overwhelming amount of data.
With the advent of digital agriculture, it was only natural that Homestead Dairy leverage its technology to improve the lives of its cows. On the farm, robots perform a variety of tasks including milking cows, feeding them, cleaning manure, and even massaging the cows. Every day, these machines add to Homestead’s current 750,000 lines of data compiled from every calf since 2015. The Housins quickly learned the question changed from how to gather more data to how to use it effectively.
Jill Houin recalls “I personally was looking through the data and thought there had to be a connection between milk consumption in the first 60 days of life and milk production through lactation. I also felt there was a connection with the automatically fed calves and how well they did when they were young, compared to their behavior in the robots. I did not have the time or the tools to analyze all the data I had.”
Previous experiences led Homestead Dairy to seek assistance from Purdue. “Both my husband and I are alumni from Purdue and have always worked well with researchers from the university,” said Houin.
Jackie Boerman, assistant professor of animal sciences, teamed up with her graduate assistant Tabby Steckler to mold the data into a usable format. “Tabby is working on trying to combine information from three or four sources into one, and come up with a way to understand if we should do things differently on this farm to improve growth rates.”
“That’s one of the weaknesses in animal agriculture,” continued Boerman. “We build a lot of stand-alone systems that are great but don’t communicate with any other system that exists.”
Steckler believes the agriculture industry’s need for data analytics will continue to grow after her graduation. “Taking the data from the machines that they use and turning it into something that’s usable is where the gap in technology is on farms now. The farmers get this data, and they don’t know what to do with it. It’s just gibberish. Our biggest goal is to turn these data sets into something that they can use in the end.”
Each system, including the automated feeding system for calves, offers unique information. Boerman explained, “That’s a system where the calves can just come up and drink milk throughout the day. On a typical farm they’re usually just fed two, maybe three times a day. In this case, they can be fed multiple times. They go up to a robot that senses their ID tag and then allows them to drink. From that, we get the daily intake and the speed at which they drink milk. We’re starting to look into if the speed they drink at is a good predictor of health.”
According to Boerman, the collaboration provides unique obstacles and opportunities. “When we do research trials at universities with calves, we may have 30 calves at a time. In this case, we have data from around 10,000 calves. From that, we’re answering questions, but we’re also thinking: We noticed this trend. Are we seeing this happen? Is there a research trial that could answer this question in a more controlled environment? There are multiple benefits for us.”
The benefits are apparent to Houin as well. “Purdue’s support and research have helped me learn vital information about my management and the changes needed to better my herd health for the future. They are great people to work with, and I love learning together about how data and technology can benefit the health of the cows and our food supply!”