I made it through much of my life without learning how to golf. Those who have been on a golf course with me might say I have made it through all of my life without learning how to golf. My dad enjoyed golfing, and I begged him to take me to a driving range for my 12th birthday. I don’t think he was impressed with my natural ability. Over the years, he taught me many things, and we enjoyed a lot of things together, but he never took me near a golf course again. Many, many years later, I belonged to an organization that had an annual golf outing, and I organized it for three years without participating. People made it look so easy that I finally decided I had to try again. I started with a $75 set of clubs and the fundamentals – grip, stance and follow-through. It’s not as easy as it looks.
Managing an automated milking barn is like that. There is a trend toward increased stocking density in automated milking barns. Quite a few farmers have made it work and made it look easy. It has worked for them because they have the fundamentals right.
Milk per robot is a popular way to measure the performance of automated milking systems. The average is about 5,000 pounds per robot, with top robots harvesting over 7,000 pounds. Milk per robot is maximized in barns designed with one stall per cow and matching robot capacity. That’s why barns have been built with about 60 stalls per robot.
Example A in Table 1 shows what that barn might look like with 2.8 milkings, 90 pounds per cow and 5,400 pounds per robot. You could multiply it by four for an eight-unit barn or by eight for a 16-unit barn. If you walk through that barn, it might look like there is room for a lot more cows because some cows are eating, some cows are resting and some cows are waiting to be milked.
Example B is what it might look like if stalls are overstocked by 10% and that empty space is filled with 12 more cows, without adding robots. Milk per cow and total milk go down because there is no longer the capacity for 2.8 milkings. All the per-cow costs, including feed, go up, but total milk goes down, and you end up with less income.
In example C, the barn is stocked at 1.2 cows per stall, but an extra robot is added, allowing extra milkings and more milk per cow. Milk per robot is lower than A or B, but there is 30% more milk per stall and 40% more total milk. Example D goes a step further to 1.4 cows per stall with three robots. Cows are milked 2.9 times instead of 3.4, as in C, so milk per cow is 95 pounds instead of 100. But total milk from the facility went up another 10%.
Of course, all these numbers are hypothetical, but they are realistic. Furthermore, the extra milk must cover the investment in the third robot.
When example D works, it is like the follow-through in the golf analogy – watching the ball go high in the air for 300 yards and drop in the middle of the green. It looks easy. That 300-yard shot doesn’t happen without good fundamentals – the correct grip and stance to ensure control, stability and power. You also must have the fundamentals right to put 1.4 cows per stall in an automated milking barn.
The barn has to flow all the time. As mentioned earlier, automated milking barns look empty with one cow per stall because some cows are eating, some are resting and some are milking – as long as some cows are eating, some are resting and some are milking. It will begin to look crowded very quickly as soon as any of those activities are interrupted. If feed is delivered late or a stop alarm shuts the robots down, there will be more cows in one place. If the wash takes two hours on milk pick-up day, there will be more cows in one place. Any activity that disrupts cow flow will disrupt it more with higher stocking density.
The freestalls have to be comfortable and available. Even at 1.4 cows per stall, there must be empty stalls available all the time. When you watch a cow lying in a freestall and chewing her cud, it is easy to think that time does not matter to her. The fact is, she really doesn’t have any free time. If we want her to spend 12 hours resting, four hours eating and one half hour drinking, and still have time for milking, socializing and grooming, she does not have time to wait. Since a cow will prioritize lying over eating, she certainly doesn’t have time to wait for a freestall. And anytime one area gets backed up with cows waiting, the next area will be backed up a few hours later. Every stall must be useable. Every stall must be comfortable. With more cows using every stall, it will take more bedding to keep the stalls comfortable.
All the systems in the barn will need more capacity. First, look at the alleys and crossovers. Are they wide enough for timid cows to get to the robot with 30 more cows in the pen? Can the ventilation system maintain fresh air with 30 more cows contributing to the humidity? Will the scrapers be able to handle the extra manure, and can the feed pusher move the larger pile of feed? Does the well have the capacity to water more cows in the heat of summer? The headlocks will no longer have room for all the cows. Those are just the bottlenecks in the pen. Extra cows also need more feed storage, more manure storage, more calf housing and more labor. What will the new bottleneck be when robot capacity is no longer limiting?
Adding robots to allow stocking automated milking barns at more than one cow per stall has increased profitability on real farms. The key production metric on those farms is milk per stall. They make it work because they get the fundamentals right.
In case you are wondering, I tried to give the $75 clubs to my son last summer. He got a better set and gave mine back to me – so I am still working on the fundamentals. Fore!
