Big data made easy: Dairy nutritionist sparks idea for data platform

Project title: “Decision making using DAIRI (Data Automation Interface and Real-Time Interaction): a platform for connecting farmers to their data

The research partnership between Austin Polebitski and Andy Bohnhoff began in an unconventional way. They met at their kids’ sporting events, and like so many parents, struck up conversation to pass the time.

In between baseball games, the two quickly learned that they had a lot in common and started talking about some challenges Bohnhoff saw related to data, farming, and decision making.

Bohnhoff is a veterinarian by training and has worked both in veterinary medicine and dairy nutrition consulting. Currently, he’s the director of nutrition for Prairie Estates Genetics, A family-owned, dairyspecific, seed company.

The idea for DAIRI came from a real-life challenge faced by Andy Bohnhoff when working with farmers. Photo from Bohnhoff.

When he was in the field consulting with farmers, he noticed a deficit of information to help farmers make decisions using milk, feed, and economic data. Bohnhoff says those data sources exist, but they are not accessible in one dashboard, and the data sets are not easy to use.

“My biggest issue was how to take these big data sets and make them useful,” he says. “Farmers have so many sources of data on their farms now, but they can’t make decisions if the data isn’t formatted in a way that’s helpful.”

With Ron Rogers, owner of Prairie Estates Genetics on board, Bohnhoff and Polebitski got to work on a project to create a more user-friendly data interface to help dairy farmers and their consultants make better decisions.

With help from Professional Dairy Producers (PDPW), we started by surveying a group of farmers to help focus our efforts,” Bohnhoff says. “Feed cost is illusive, and it’s often hard to pinpoint, so we decided to focus mostly in this area.”

Once they decided on the appropriate data sets, Polebitski recruited two data-science students to refine the online dashboard and work on backend development. DAIRI uses open-source data science products (think “templates”) that are available for researchers. Multiple Hub-funded projects at both UW–Madison and UW–Platteville use these tools.

“It’s been great to learn about what data matters to dairy farmers. Data-driven decision making is a shift for sure, and we hope this tool helps farmers have a broader view of their business,” says Polebitski.

A screen shot from the DAIRI platform taken July 2023 shows how four lactation groups perform from a milk production and feed standpoint. Farmers and consultants will use this platform, which uses data from multiple sources, to make informed management decisions.

With an operational dashboard and platform, Bohnhoff has beta tested DAIRI with a few of his Prairie Estates Genetics customers. Over the next year, the company will use the tool with a group of 12 farms as a valueadded service and to stay connected to customers all year long. Wider availability of DAIRI could be available as early as 2025.

Bohnhoff is careful to explain who is best suited to use DAIRI. “This platform needs to come to farmers through a service provider like a seed company or a nutrition consultant. For it to work, there needs to be a dedicated staff person inputting the data,” he says. “It’s not realistic for a farmer to do it themselves.”

“I really like what we developed, and I would have utilized it myself if Prairie Estates Genetics wasn’t interested, but thankfully, they are,” says Bohnhoff.