WELCH, Minn. – Seven years ago, Dr. Mike Brouk decided to learn more about robotic milking after a Kansas dairy producer called him and expressed interest in a new way of dairying.
    “At that time in Kansas, we didn’t have any robotic milking herds,” Brouk said. “I thought this is where we’re headed as an industry and will be important to our industry. If I’m going to help producers with this, I really need to take time to immerse myself in it.”
    Brouk, extension dairy specialist and professor in the Kansas State University Department of Animal Sciences and Industry, now has knowledge to share. He presented “Defining and Achieving Success with Robotic Milking” Jan. 16 at the Form-A-Feed Professional Dairy Conference in Welch.
    “This nice shiny box does a really good job of harvesting milk, but it does not produce milk,” Brouk said. “The cow standing in that box does. A lot of times people have forgotten that. If we expect more milk production because we’re going to use a box, we better make sure we put the tools in to produce that extra (milk) because that’s the only way we’re going to harvest it.”
    Brouk talked about the financial aspect of robots. The cost of a regular milking set up is about $350 per cow whereas robots are about $2,500 per cow, he said.
    “I’ve done the math a lot,” Brouk said. “When you look just on cost basis, going to robotic milking is going to change the financial picture of the farm.”
    He said the pounds of milk per day each robot harvests is important to watch. As an example, Brouk said a farm that invests $180,000 per box at a 6% interest rate on a seven-year loan would have 16% of its milk check to cover the loan of the milking equipment if the robots harvested 4,000 pounds of milk per day. At 5,000 pounds, that amount would go down to 12.5%.
    “That’s a huge difference,” Brouk said. “We need to be looking at a baseline of 5,000 pounds (of milk per day) harvested per robot in the United States.”
    In order to do this, it is important to look at the number of cows dedicated to each robot. Brouk said if a farm milks 50 cows per day, it would need to have a herd where each cow produces 100 pounds per day.
    “We don’t have very many 100-pound herds, and we don’t have many herds that milk 100 pounds consistently day in and day out,” he said. “It’s hard.”
    The housing with each robot needs to be able to accommodate the number of cows needed to achieve 5,000 pounds per robot per day.
    “If you designed a barn that only has 50 stalls per box and you move to 65 cows, think of what you did with overstocking,” Brouk said. “Now you limited milk production because you overstocked. If you’re in the process of building a barn, you need to think about how many stalls you have per box.”
    Cow flow is also important, Brouk said. It is influenced by the barn’s design and traffic pattern. There are two types of designs – freeflow and guided or milk first. Freeflow means the cow can move from the stalls to the robot or to the feed bunk freely without going through sort gates; milk first means a cow has to go through a selection gate and get permission to go to the robot. Selection gates are an option for people who are worried about cows that continually circle back to the robot.
    “You’ll find cows that will make 60-80 passes a day,” Brouk said. “They’re keeping animals from coming to the robot.”
    Hunger is the biggest driver to get a cow to visit a robot.
    Four types of feed can be given in the robot depending on the system – pelleted, ground, liquid and rumen inert fat. Many use pellets, Brouk said; however, fines in pellets can be a big issue because they leave residue in the bowl, causing cows to continue to lick and increase box time.
    In a freeflow set up, the average intake of pellets is about 12 pounds per cow while a guided flow set up averages about 6 pounds per cow.
    A partial mixed ration is fed at the bunk. Often times, robotic dairy farms do not have cows separated by age or lactation stage as a traditional parlor set up might.
    “How we balance that becomes a big question,” Brouk said.
    Milk production level will drive hunger.
    “Forage quality has a big impact on how soon her gut is going to empty, impact on milk production, impact on momentum of path of early lactation and a huge impact on whether or not she moves,” Brouk said.
    According to a study, Brouk said BMR corn silage is the most consistent forage to have in a cow’s diet. He said with robots, consistency is important.
    Closely monitoring the nutrition in a robotic barn is important and will set cows up for success. As an example, Brouk shared about a farm that fed 5.5 pounds of pellets per cow per day across all lactations. Despite excellent forages, the dairy averaged 78 pounds of milk per cow per day and had the number of visits to the robot and production drop off at about 25 days in milk.
    “They were ready to sell the cows and the robots, and exit the business,” Brouk said.
    After Brouk made suggestions, the farm made changes, including increasing the amount of pellets available to cows in early lactation. Although this increased the expense of the pellets, it increased milk production to average around 102 pounds per cow per day.
    “It’s important to concentrate on cost per hundredweight of milk produced not cost per cow per day,” Brouk said. “Cost per cow per day is diluted out when milk production increases.”
    Brouk said before installing robots, it is important to take all these factors into consideration.
    “These robots are fun to work with,” he said. “You could spend all day and half the night talking about all these settings and what they should do. But none of these settings make any difference if the cow doesn’t show up to the robot.”