Installing a new milking system of any kind is a significant investment. As more farms install automated milking systems, various methods are used to measure profitability.

Lee karen
Managing Editor / Progressive Dairy

Jim Salfer, a dairy extension educator with the University of Minnesota, along with his university colleagues, has surveyed dairy producers with automated milking systems and gathered some key financial considerations for anyone thinking about installing one of their own.

At the Vita Plus Dairy Summit, Dec. 6 in Madison, Wisconsin, Salfer cited a Holland study pairing 31 conventional and 31 robotic farms that invested in new milking systems in the same year. When comparing the two types of systems, the farms with robots were less profitable overall compared to the conventional farms.

However, when the systems were compared based on full-time-equivalent personnel, there was no difference in profitability, and gross margin was higher for the robot farms. “You can milk more cows and potentially make more money with the same amount of labor [in a robotic system],” Salfer said.

To see how U.S. farms with robots fare, he analyzed data from Finbin, a farm financial management database supported by the University of Minnesota. Finbin contains information from farms in Minnesota and a couple from Ohio. Salfer compared average automated milking system herds to the top 40 percent of automated milking system herds based on net margin per cow, as well as conventional herds with 100 to 200 cows. Across the board, the average herd size for the three groups was within 10 cows.


The conventional herds were likely to be milking 2X, and the automated milking system herds recorded higher milk production and more milk per full-time equivalent than the conventional farms. Automated milking system herds also had increased feed costs. However, feed costs in automated milking system herds were quite consistent compared to the highly variable feed costs for conventional herds.

All operations were consistent in milk price per hundredweight received, but the automated milking system herds did have slightly lower milkfat.

“Net margin per cow has been tough for our robot herds,” Salfer said. This was mostly driven by the milk price. He added the top 40 percent of robot herds were doing better than average conventional herds.

In direct costs per cow, repairs and utilities were higher for the automated milking system herds. Labor per hundredweight was $1.40 to $1.50 for the automated milking system herds and $2.34 for the average conventional herd. However, Salfer said many of these farms are mixed enterprises and, in some instances, calculating the labor specific to the dairy was a best guess.

When comparing overhead costs, automated milking system herds had higher interest and depreciation.

Online model

Salfer and his team developed an online spreadsheet to help dairy producers run different financial scenarios if thinking of adopting robotic milking. “Don’t use it to determine cash flow,” he cautioned. “Use it to play around, and see what happens if milk goes up or down.”

The numbers he used were $16 per hour for chore labor, $25 per hour for management labor, $17 per hundredweight for milk and 11 cents per pound of feed. The model looks at a 20-year timeframe, so he also added a factor for wage inflation over time.

The first example was a 240-cow farm, comparing four robots to a double-eight parlor. The total investment for a new barn with robots was calculated at $2.4 million and a new barn and parlor at $1.352 million. Salfer said part of the reason the automated milking system facility was figured higher was because producers are buying an entire technology system and are likely to want to add automation to feed push-up and manure removal.

“If you’re going to take maximum advantage of robots, you should try to make it as good of a robot barn as you can,” Salfer said. “That’s why building costs are so high.”

For parlor numbers, he used 3.6 turns per hour and figured 2.5 hours per day of labor used for milking-related tasks like fetching cows and scraping stalls.

In this scenario, the robots were more profitable compared to a parlor milking 2X. “That’s part of the reason our smaller dairies are putting in robots,” Salfer said. “It’s a pretty good economic decision.”

He noted models do include a lot of assumption and, here, he assumed milk would go up 5 pounds per cow. Automated milking allows more small farms to put up hay on time, take better care of their calves and better technology to stay on top of breeding. “Depending on what you can do with your labor, it wouldn’t surprise me if your numbers can go up a lot,” he said.

In this 2X scenario, the revenue comes from more milk as well as labor savings, but if it’s all family labor, that savings comes out of family living. “This is the cost of improved quality of life – either that or you need to get more milk.” He adds, “It all depends on what you do with that labor. If you go to the casino or watch TV or play video games, you can’t make a robot payment with that.”

In another scenario, he looked at a 1,500-cow farm with 25 robots or a double-24 parlor milking 3X and running almost 24 hours a day. Here the robots are less profitable than the parlor because, when compared to 3X milking, cows in robots produce 2 pounds less milk, an assumption he used based on what was found in the literature.

Looking at the scenario from another direction, Salfer said to make the profitability of the two different systems – parlor versus robot – to break even, you would currently be paying your milkers $27.05 per hour. Another way to get the systems to break even is to gain about 3 pounds more milk in robots than what the farm is currently getting 3X.

He said there is no data to support that potential increase in milk, but the cows won’t be spending time in the holding area, could benefit from consistent groups and could potentially live longer.

Milk per robot

One metric automated milking system farms use to monitor profitability is milk per robot. Salfer said the farms in his dataset average 4,325 pounds per robot per day, but the variation is extremely wide.

Some of the biggest factors he found leading to increased milk per robot are more milking visits per day, milking speed, more cows per robot and feed consumption, which could just be a factor of having high-producing cows. Salfer also found herds with automated feed push-up had more milk per robot. Herds with automated feed push-up had 80.2 pounds of milk per cow, compared to those with manual push-up at 69.4 pounds per cow.

Negative influences on milk per robot are residual feed (feed offered in the robot that isn’t consumed), higher failed visits and refused visits.

In addition to normal items that drive milk production on any farm (i.e., well-balanced diet, excellent transition cow program, high reproductive efficiency, low somatic cell count and good cow comfort), there are several things farms with robots can do to increase milk per robot.

Salfer suggested fetching the early lactation cows more frequently to maximize lactation potential, minimizing box time per cow with cows that attach and milk out fast, keeping free time to 10 percent or less, culling cows for not adapting to robots, keeping equipment in top working order, singeing udders and trimming tail switches.

Some farms manage robots by carefully thinking about milking permissions. “They don’t want to milk cows that aren’t milking well too often,” Salfer said. The industry commonly looks at high average visits as a good metric, but the farms with the highest milk per robot in Salfer’s study averaged just 2.6 to 2.8 visits per cow per day.

One group that could improve its number of visits to the robot is first-lactation cows. By three weeks postpartum, multiparous cows are averaging more than three times a day at the robot, but it takes heifers at least a month to average three times a day. Salfer said it is not unusual to see heifers in the first 40 days average 1.8 visits.

“I think we really need to be thinking about how we can improve that,” he said, noting one farm in his area has started to train heifers by walking them through the robot several times prior to calving.

Whether benchmarking to other farms, using an online model or focusing on milk per robot, it is good to focus on financial factors from many viewpoints before investing in new technology. When it comes to robots, Salfer said the main factors affecting profitability are milk production per robot, milk production increases compared to other milking systems and trading manual labor for more productive labor.  end mark

PHOTO: One group that could improve its number of visits to the robot is first-lactation cows. By three weeks postpartum, multiparous cows are averaging more than three times a day at the robot, but it takes heifers at least a month to average three times a day. Photo by Mike Dixon.

Karen Lee