One of the benefits to adopting an automatic milking system is all of the information readily available for users to access. In a webinar with the Penn State Extension Technology Tuesday series, Mat Haan, Penn State Extension dairy educator, and Steve Halahan, a dairy consultant and feed nutritionist for Cargill Animal Nutrition, reviewed the various forms of data generated and how they can be utilized by producers, nutritionists, veterinarians and other members of the farm team.

Lee karen
Managing Editor / Progressive Dairy

“Robots will measure almost 120 variables per cow per day,” Haan says, noting some of these measurements are also becoming more common in conventional herds. He divides the information into the following five categories:

1.Systems management – This includes milkings per cow per day; milking time, also known as the time it takes to milk the cow; and box time, which is the time cows are occupying the robot stall.

2.Milk production variables – Categories such as milk yield, fat, protein and lactose, help to demonstrate farm production. However, not every category is available in all automatic milking systems. In general, this data can be viewed on a per-quarter basis for each cow, as an individual cow measurement or an entire herd average.

3.Udder health and milk quality – Through electrical conductivity, milk color, milk temperature, somatic cell count or some indication thereof, automatic milking systems can offer an udder health report, combining one or more of these readings.


4.Nutrition and general cow health – The data collected here varies by system but can include how much pellet was fed, bodyweight, rumination, MUN, milk enzymes and ketone bodies.

5.Reproduction – Cow activity and milk progesterone levels are two measurements for reproduction, but these, too, are not available in every system.

In addition to data collected by the robots, some systems can connect to external data collection devices, such as a sampling unit for DHIA. This can provide the user with even more information and could also be used to help calibrate the robot’s functions.

Haan adds that some automatic milking system software can connect or sync with other herd management software systems, giving producers additional data to help make management decisions on the farm.

Data from automatic milking systems can be accessed in multiple ways. The robots are equipped with touch screens to access the data cowside, and they likely connect to a central office computer.

“Mobile apps are becoming increasingly more popular,” Haan says. This gives dairy producers the flexibility to make adjustments to the milking system when they are away from home.

Technologies such as Google Glass are continuing to enter the marketplace and give data users more options in their access points.

Lastly, Haan says, remote access is a handy way for robotic herds to share information with their nutritionists and veterinarians.

There are a number of benefits to having access to data collected by robotic milking systems, Haan says. Readily available data improves labor efficiency on the farm.

Rumination and activity monitoring allows users to see what is going on with an individual cow, and producers don’t necessarily have to walk the herd to locate a problem. A wealth of data can enhance management decisions regarding feeding and health, which in the long run will increase milk production in the herd.

How a nutritionist usesrobotic milking system data

Halahan has been working with automatic milking herds since 2012. He currently works with four robot herds in Pennsylvania and Maryland.

He says he typically starts analyzing the data by looking at the dashboard for an overview. Here, Halahan can see what is happening in the herd over a 24-hour or seven-day time period. He checks milk production, cow visits and refusals.

“I used to put a lot of stock in refusals, but I don’t look at it too much any more,” Halahan says. “It can be heavily abused by several cows that crave pellets.”

He also doesn’t put a lot of weight on milk component values, as he has found a lot of variability here depending on how well the calibrations are kept up in the robots.

Halahan does look at the herd overview, which lists cow numbers, production data, days in milk and other information on the cows. It is easily accessible, he says, and he is able to sort the list by days in milk, for example, or click on individual cows to see the trend for that particular animal. He can see if fresh cows are making a nice uphill climb or identify a certain day where an issue has occurred.

This nutritionist also likes to look back at data from three, six or 12 months ago. “It’s pretty fascinating to go back a full year and compare this day to one year ago,” he says. That way he can eliminate some of the seasonal variation that can occur.

By looking back in time, Halahan will check to see if the herd is making progress, whether it is at the same stage of lactation or has the same percentage of heifers.

He is able to break down the data for a more detailed analysis on segmented groups. For example, a fresh cow report for day 10 to day 40 can show him how well the dry cow ration is performing.

Looking at peak milk, Halahan says he is able to see if first-calf heifers and mature cows are peaking in the appropriate range.

“I can pinpoint certain groups. I don’t just have to look at the general herd, I can see heifers, mature cows, special needs,” he says.

Rumination monitoring is fairly new and not all of Halahan’s clients are using it, although he says it will likely continue to gain more traction.

One aspect he particularly likes about robot herds is his ability to change the feeding program while on the farm. He is able to review feeding rate charts and implement new information on robot feeding rates before he leaves the driveway.

Using production data from the farm that day, he can adjust the feeding rate for fresh cows that need more, or with bodyweights, he can scale back what is being fed to the pregnant cows so they don’t get too fat before dry-off.

The change in feeding amounts can be done immediately at the dairy after he sees the cows, their body condition and manure. “There’s almost no lag time at all on robotic dairies,” Halahan says.

The daily accessibility of charts, graphs and data allows Halahan and the dairy producers he works with the ability to see the information clearly. Being able to analyze data instantly, they can make more accurate management decisions easier and faster. PD

Photo by Mike Dixon.

karen lee

Karen Lee
Progressive Dairyman

Producers comment on data use in robots

As a part of the Penn State Extension Technology Tuesday webinar on “Collection and Use of Data in Robotic Milking Systems,” Mat Haan, Penn State Extension dairy educator, shared video clips of dairy producers that use robotic milking. Here are some excerpts of producer comments on how they utilize the data provided to them.

Blair Courter
Mill Hall, Pennsylvania
“The best feature I like about the robots is the milk quality graph. It tells you if she has an udder health problem. You verify it and then you treat it. Then you can see it every time she milks; you can see that treatment, if that’s the right treatment, right away.

You don’t have to wait 12 hours or the next day, guessing on your side. The robots show you on that graph that the treatment is working; she’s coming back down. I really like that feature, being able to see that visual sign whether it’s getting better or it isn’t working and you need to switch to something else.”

Amy Martin
Gingrich Meadows Dairy
LeRoy, Michigan
“The greatest challenge was actually retraining us and the employees. This is a totally different way of managing cows. It’s different than just a parlor situation. Just the whole breeding, mastitis treatment, monitoring cows is just totally different.

The computer is not that challenging, I mean, it is workable for any employee, they can understand how to use it; our kids can understand the computers and the technology.

It’s more that in a parlor you are physically milking that cow and looking at that milk and that cow every day. Instead, in a robot, you are looking at reports and you are walking through the barn looking at that cow, looking for a different sign.”

Margie Weiss
Weiss Centennial Farms
Frankenmuth, Michigan
“When the vet comes, we know right away that this cow is down X amount of pounds in milk or that she’s running a fever. So it kind of gives them a heads-up on more of the actual indicators of cow health.

“Our nutritionist definitely uses it. We have TeamViewer on our computer. She has the log-in information and she goes on our computer, I think almost daily, and checks out overall herd performance and individual fresh cows; she can monitor those cows.

I think she’s really getting a lot out of the program. She can see the butterfat and protein levels. She can use those to adjust the ration a little bit.”

Lewis Horning
Ephrata, Pennsylvania
“I try to look at the computer at least a couple of times a day. I often like to look at cows that haven’t milked since I last looked at the computer and also, of course, cows that are late.

A cow that consistently comes through at seven or eight hours and she’s at 12 hours, for me that’s a red flag to go see what’s wrong with her. She may be in heat or sick. There have been ways – just looking at the reports – to indicate what’s going on.”

Olivia Platt
New Columbia, Pennsylvania
“I use it primarily for two things – to detect sick cows and to detect cows that are in heat. I personally still go out in the barn and watch my cows. I like cows. I’m a cow person. I want to be in my barn. I go out; I do watch them. Sometimes, as a result, I find cows in heat before my computer does.

That being said, when the computer finds a cow in heat based on rumination and activity, it’s, in my opinion, 98 percent accurate. If I come in in the morning, the first thing I check usually is my heat probability list, and if there’s a cow on there that is 80 percent likely to be in heat or more, I don’t even go out in the barn, I just call the breeder.”

Ken Shaffer
Port Trevorton, Pennsylvania
“The nutritionist, he uses it quite a bit for balancing out the ration. Going by the rumination, as far as putting a little more fiber into the diet or something like that if the rumination is too low or too high. It’s been a learning experience for him. We were the first robot that he’s been working with. So he’s been kind of learning with us.”

Howard Straub
Triple H Farm
St. Johns, Michigan
“We do not use the rumination tags, but we do have activity, so we have our breeding schedule set by that. I don’t even have to see a cow in heat; if I look at the computer and it says she’s in heat, she’s probably in heat. So we can breed off that.”