Dairy Students Gain Experience with Edge Computing to Monitor Animal Behavior

By Jim Massey, Freelance writer 

Josh Gerbitz – Dairy Sciences Class of 2022

UW-Madison Dairy Sciences major Josh Gerbitz says before he got involved in a research project with Assistant Professor João Dórea and fellow students, he didn’t necessarily think of research as being something that was all that practical. 

But his thinking changed after working on a project involving the use of edge computing to monitor dairy cattle movement as a predictor of animal health and welfare.  

“This was my first experience with a research project and I thought it was really cool,” Gerbitz says. “It was not like coming out with a scientific paper. This project was very, very practical and made a lot of sense.” 

Dórea, who joined the Department of Animal and Dairy Sciences in August of 2019 when a position was created for precision farming and data analytics, applied for and received an AI for Earth grant from Microsoft to monitor the movement of dairy animals on the UW-Madison campus and at the Marshfield Agricultural Research Station. 

The goal was to use edge computing – a system that allows data to be analyzed close to where it is collected before it is sent to the cloud – to monitor animals as they eat, drink, stand up or lay down. Because only the results of the research predictions are sent to the cloud, it is less expensive to process than if a large amount of data is transmitted. 

João Dórea, faculty member in the UW–Madison Department of Dairy Science, explains results he observed using an automated computer vision system that was developed to monitor the behavior of dairy calves

“The power of the data is tremendous for predicting the behavior of the cow,” Dórea says. “When you know how much a cow is lying down, for example, it can be effective in forecasting the performance of the animal. Animals change their behavior when they’re sick, so if we can determine that a specific animal is lying down more than normal, we might be able to take action to prevent serious disease.” 

The hardware built by Microsoft has a user-friendly interface that allows non-expert quick implementation. Knowing that, Dórea decided to pilot the edge computing system with four dairy science undergraduate students. 

“In our lab we develop a lot of those system and it would be nearly impossible to have non-experts building and deploying such systems,” Dórea says. “When I saw the grant opportunity, I thought that would be the perfect project to introduce key concepts and applications involving edge computing and sensing technology to ag students.”  

The team developed case studies to monitor the movement of dairy calves and the feeding behavior of dairy cows. Cameras were set up aimed at pens where movement of the animals could be recorded. 

Gerbitz says in both studies, students took sample photos of the animals in different positions, and then Dórea programmed the computer to record how much time the animals spent in those various positions. The computer analyzed the information and sent the data points to the cloud. 

Kelsey Maurer – Dairy Sciences Class of 2022

Dairy Science major Kelsey Maurer, one of the undergraduate research assistants, said the technology was somewhat difficult to understand at first, but the students practiced with toy animals on a table in the classroom to figure out how to position the camera and record the images.  

“(Dórea) emphasized that we wouldn’t have to learn computer programming – that was the scary part at first – so he took care of that,” Maurer says. 

“What really drew me to take part in this project was the advanced technology. Over the years we have gained so much technology in the agricultural industry. I thought it would be really cool to work on a project like this and see how this technology works.” 

The monitoring of dairy cattle movement is not necessarily anything new, with neck tags, ankle bracelets and ear tags used on some farms to keep track of what cattle are up to. But Maurer says she believes this system could be used to watch and record the movements of a group of animals rather than just a single calf or cow. 

Dórea says the goal with the small pilot case study was to get students familiar and introduced to cutting-edge technology. 

Gerbitz was active in youth activities while growing up in Milton in Rock County. He is currently the president of the Badger Dairy Club on campus.  

He worked on a dairy farm while in high school and fell in love with the dairy industry. 

“The people that I met along the way, at cattle shows, conferences and other events, prompted me to pursue it further,” he says. “I decided I want to work with these people the rest of my life, so I might as well go to UW-Madison where the dairy industry is happening right now.” 

After he graduates in May, Gerbitz plans to work full-time for Alta Genetics as a dairy business adviser.  

Maurer lives on her family’s 1,200-cow dairy farm in Manitowoc County. She says she doesn’t know if she will return to the family dairy farm in May when she graduates, but that’s her long-term goal. 

“I haven’t fully decided what I’m going to do (after graduation),” Maurer says. “I’m still exploring my options.” 

João Dórea, assistant professor of dairy science (right), and animal sciences graduate student Arthur Fernandes set up a camera over a calf hutch at Arlington Agricultural Research Station in Arlington, Wis.

Dórea, a native of Brazil, spent two years coordinating dairy and beef research in Latin America for DSM, a global supplier of animal health and nutrition products. He developed his expertise in data analytics and sensor technology when he came to the UW-Madison in 2016. Currently, he is an assistant professor in the Department of Animal and Dairy Sciences, focusing on artificial intelligence technologies for livestock phenomics and farm-management decisions.

Read the student’s research report Using Edge Computing to Monitor Animal Behavior in Dairy Farms here.