B (3:27)

Cody grew up on the farm, and during that time, he witnessed how some of the most pivotal technologies started changing ranching. It wasn't just things like hydraulic working chutes. Over time, the introduction of automated milking machines enabled producers to start focusing on other tasks on the farm. Most of them are made of a laser guided robot arm that cleans and then attaches itself to a cow's teat for milking. Traditional dairy farms mostly rely on human labor for daily tasks like milking, feeding, and manure management, among other things. But this can eat up almost a third of a farmer's workday, leaving little time for health inspections and maintenance. With automated milking machines, in less than 10 minutes, a cow is milked. And most of the time, you can simultaneously milk multiple cows. These systems aren't really new. They first emerged in the early 1990s, just as Cody was beginning to take more responsibility on his farm. And like countless other farmers across America, the advantages of automated milking meant more time for farming, less reliability on labor, schedule flexibility, and much more. Still, that was 30 years ago. And as with most technological trends, there's evolution. If automated milking systems really shaped the industry a few decades ago, what does that relationship with technology look like today? What's changing the game now?

D (4:48)

Our focus over the, and certainly over my time here in company has been around increasing the level of automation that's there.

E (4:56)

It's been around sustainability. So whether that's from an environmental point of view in terms of producing energy usage, water usage, chemical usage, and also in the animal welfare side of it. And then more recently, we're looking at.

D (5:09)

How we can combine all of the.

E (5:11)

Information that the equipment is gathering and analyze that and using artificial intelligence to.

D (5:18)

Help think better decisions back to the operator or to the business manager.

B (5:25)

John Daly is an engineer and a researcher in Ireland. He works for a company called Dairymaster, which focuses on building automation systems for dairy farming. He leads their research department. And like Cody, John grew up on a farm. And even though he went on to pursue engineering, he was aware of the Challenges the dairy industry faces things like labor shortages, which can lead to low productivity, with which in turn leads to tighter profits. So he took all of these concerns and as part of the research and development team at Dairymaster, worked to produce practical and innovative solutions. John's work is symbolic of how new technologies like AI can really enhance the work of dairy farming, an industry left relatively untouched by advanced technology. It broadens the conversation beyond automated milking systems, and it addresses bigger concerns and needs, like ensuring production remains sustainable and reliable and which decisions could maximize output.

D (6:21)

With the automation. What we want to get to is basically make that work easier, so take out the physical and manual labor for the operators. We're not saying that we're going to completely replace people in the parlor, but what we do see is that we'll move to a situation that the people that are working in the parlor are doing more of a management type role. So they're overseeing the equipment, they're overseeing the cows, they're able to do the animal husbandry, and they're able to deal with a cow. I mean, we see that in our own production environment here, where we have a lot of very automated equipment, but we still will have an operator. The operator might be managing two or three or four machines, and those machines are all running autonomously, but the operator is still there overseeing those machines.

B (7:08)

Dairymaster's work in automation actually unfolds in three major parts. Personalized milking, robotic teat spraying, and a rotary parlor that milks, monitors and inspects nearly 20,000 cows a week. Swiftflo is a personalized milking system that identifies and applies unique settings for each cow, in some cases reducing milking time to about 60 or 70 seconds a day.

D (7:31)

And that's when every cow comes onto the platform and that cow is identified. Using the COFID technology, we can personalize the milking settings to that individual cow. So we call it individual milking for individual cows, because different cows, depending on.

E (7:46)

Their breed, the parity, their stage of.

D (7:48)

Lactation, she may have different requirements also.

E (7:51)

Depending on the health stays, she may have different requirements for the milking equipment. So instead of having a one size fits all, we can actually tailor those settings to the individual cow.

D (8:00)

And that's using factors like pre milking.

E (8:03)

Stimulation, so pulsatious stimulation, where we can really help increase the letdown for the cow and that shortens the milking time. We can also do flow dependent pulsation, flow dependent milking settings where we can change the rate and ratio of the milking equipment, depulsation settings for that individual cow. And then finally we can have variable takeoff where we can change at what point we determine that that cow is milked out enough, and then we can.

D (8:29)

Remove the milking equipment.

B (8:31)

RFID stands for radio frequency identification. It's a type of wireless technology that automates data. What you do is attach a tag or a chip to a person or animal, in this case a cow. The microchip then identifies data and sends that information out through radio waves. Think contactless payment or even library book. Tracking it directly feeds valuable information into a device like a computer or a mobile phone. Just like humans, animals have their own sets of behaviors and health conditions. They can be stubborn or incredibly cooperative, aggressive or yielding. But understanding who they are individually helps inform decisions about how to manage them.

D (9:14)

Other types of technology that we've introduced recently, for example, is our robospray. So that's where we use robotics to actually interact with the cow on the platform. So we do robotic teat spraying of the cow. So this is where the cow can be disinfected. And we can do that either before the cow is milked, so when she enters into the platform, or after the cow has finished the milking process and before she goes back to her pen. And we can do both or we can do either, or we're using machine vision there to really get accurate spraying of the teeth and accurate coverage of the teats, because having that process performed consistently and repeatedly on every cow is really key to making sure that you have good other health and you reduce your incidence of mastitis. And of course it's also a labour saving, so it feels frees up labour time for other tasks on the farm or for more management type roles in the parlour. Because again, if you're a producer where you're milking 24 7, you can't afford to have the parlour break down, you can't really afford downtime. So we want to make sure that when we green underneath the cow and we spray the cow's ulur, that we're targeting the teats directly. And by having a special type of camera on the end of the robot arm, we can guide that robot to guide where it's going.

B (10:30)

But perhaps one of the most informative pieces of Dairymasters technology is their health and fertility monitoring system, which transmits important data in real time.

D (10:39)

So what we're monitoring there is really is the cow's movement. And from that then our algorithms can determine very, very accurately how much time per day the cow is spent ruminating, how much Time she spent resting, how much time she spent feeding. It's also monitoring her level of activity. And we can use that to determine things like is the cow in heat, you know, so looking at her fertility cycle and when's the optimum time to breed that cow? When's the optimum time to breed that cow? Depending on whether it's artificial insemination with traditional semen or whether it's artificial insemination with sex semen, which is where the semen has been targeted to select only for female embryos for. For heifer to produce heifer caps. So our mood monitor system is doing that on the activity level. And then when we look at the cow coming into the parlour, so we're monitoring the movement of the cow in the parlour, we're monitoring her milk flow. So how often every three seconds. We're monitoring the flow rate and the yield from that cow while the milk machine has been attached. And with that, we can determine things like her letdown, how quickly is she letting the meltdown, how much of yield is she given in the first 60 seconds, in the first two minutes, what percentage of yield is she given at that point? Also, is there bimodal flow? So bimodal flow is where we get an initial let down, and then the milk flow might stop before it starts again, and that will have an impact on both the productivity from the parlour, but it can also have an impact on the other health because the milk machine is on the cow for longer and so on. So we're also monitoring as well as flow rate, we can monitor temperature, we can monitor half the milk, and we can monitor all of these real time and produce metrics, real time based on this, and feed them back then through the Dairy Vue360 system to the owner or to the farm manager.

B (12:30)

John's work with Dairymaster is pushing the frontier of not only the dairy industry, but also our understanding of animals, especially those that contribute to the wider food industry. Monitoring systems can unlock a deeper understanding of cattle breeding, not just in terms of production and profits, but in closing gaps when a veterinarian is unavailable or when a cow seems tired, moody, or simply down.

F (12:53)

I'm Julio Giordano and I'm a professor of animal science in the Department of Animal Science in the College of Agriculture and Life Sciences at Cornell University.

B (13:09)

Julio knows a lot about dairy. He grew up in rural Argentina in the province of Cordoba. Cattle, beef and dairy farming are in his family.

F (13:18)

When cows are healthy, they produce more milk. When cows are well fed, they produce more milk when cows feel good, when they have good well being, they produce more milk. For a lot of technologies, it's no longer a question of if they will be adopted or should be adopted by farm, but it's more of a question of when. And the answer to when has to do with multiple factors such as the needs, the specific challenges of farms, the ability of farms to afford, buying some of these technologies. But there are some technologies that are transformational and that probably be adopted by a majority of farms. I mean, mechanized milking is an example.

B (14:04)

I asked him about sensors, kind of like the ones Dairymaster uses with RFID technology, and he told me they work by identifying anomalies, changes in cow behavior and things like body temperature of the cows.

F (14:17)

Typically when cows are not feeling well, when they have a disease, rumination time drops substantially. So the minutes of rumination go in a healthy cow from 550, 570 minutes per day to as low as 200 minutes or 250 minutes. And that's typically an indication that something is not going well for that individual. And that is what the software from the technology, the software associated with the sensor will identify that changing pattern, will identify that anomaly, that deviation from the normal baseline for the cow, and then it will flag, it will say this cow should be looked at because there might be something going on. Body temperature. Typically we see spikes in body temperature when cows undergo an infectious health disorder, and you see increases of 1 degree Celsius or more in cows that have a disease. So for the most part, the majority of these automated health monitoring systems or reproductive status monitoring systems, work by detecting deviations from baseline levels.

B (15:42)

Dairymaster is now working to enhance how AI can provide a more reliable feedback loop for farmers who aren't on the field. It helps set expectations for each week, like how how much milk will be produced. But if a pivotal step in dairy farming involves cow monitoring and reliable production, how do you know the milk is ready to ship? How do you know the milk is ready for cheese, yogurt and butter production? And what determines how much farmers get paid. All of this is part of a symbiotic relationship in dairy production called metering and sampling. It basically measures how much milk is produced, test the quality, and uses that information to determine how much a farmer gets paid. Metering measures the milk, sampling tests the milk. But in the industry, there's been a lack of confidence in how traditional forms of calibrated farm tanks, which is a fancier way of saying huge milk measuring tanks are reliably precise. It's been a concern among farmers who need to constantly update their tools.

G (16:40)

The old fashioned way to measure milk on the farm was to get a calibrated farm tank. So the farm tank itself is the measurement instrument for the milk. So you can imagine every few years that farm tank will be calibrated and they'll try to understand how accurate is that farm tank. So when our technology was being tested to see, okay, well, is this new technology accurate, obviously we had to get a good baseline to measure against. None of these farm tanks were due to be calibrated. They're only due to be calibrated once every five years. But they calibrated them all in advance of the testing of the Piper system. And what was interesting was that 5 out of 13 bulk tanks needed to be calibrated. So obviously there was a hidden issue there in measurement in the traditional way, that tanks could be not due to be recalibrated, but could still be producing inaccurate results. And the importance of that to a farmer can't be overestimated, because really, the two things that govern the paycheck of every farmer out there is how much milk do they produce and what components does that milk have?

B (17:55)

Lee Hamilton is a lawyer turned dairy business leader. Her family operated a company that specialized in metering and sampling, and they would often come to her for legal advice on scaling in the U.S. now she's the CEO of that company, Piper Systems, and this is what they do.

G (18:11)

So in Ireland, where obviously the business started, the traditional thing is to sample automatically using a meter and a sampler, like the Piper system. But in the US they were reliant upon different sampling methodologies, and those sampling methodologies varied wildly. But the most common one would have been that you agitated the milk tank for a long period of time and then dipped a sample out of the top using a dipper or a ladle, and then would pour that into a sample vial. And obviously that's very dependent on the agitation process and the milk being totally homogenized by the time you dip that sample. The issue with that, of course, is that milk is a stratifying liquid. So all the time you're fighting against physics, because what the milk wants to do is have skim at the bottom and all of the fat at the top. So what the Piper system does, and one of the reasons we were driven to go into the US market, was it helps producers to get an accurate sample so that they can really get the benefit of all the work that they do, nourishing the cows properly. Making sure their herd is healthy, getting their production up high, nice and high, so that they're producing lots and lots of milk, but also capturing the value of those components like butter, fats, like proteins that are in the milk, so that the farmer can be fairly paid.

B (19:41)

Unlike most industries, dairy farmers in the US Actually have very little control over how much they get paid for their products. They can't set a firm price. They are given a price. Farmers sell their milk to processors like Land O Lakes or Nestle, and those companies then pay dairy farmers. And it's not the same for farmers across the country. The U.S. department of Agriculture sets base prices for milk based on where you live, like your home state. So a farmer operating in Florida might not make the same as a farmer based in Iowa. And those prices don't take into account farming expenses. So companies like Land O Lakes and Nestle not only process dairy, they also produce their own dairy products like butter or powdered milk. So they tell the USDA how much they made from butter and powdered milk products, and that determines how much a dairy farmer is paid. Farmers get paid relative to how well wholesale milk products do on the market week to week. Of course, there are more layers to the economics of US Milk production, and that's why metering and sampling are so important. Processors also pay farmers based on milk components like butter, fat, protein and other dairy solids. And the higher, more accurate these numbers are, the better farmers are paid.

G (20:53)

So in the case of Piper's product, the producer will get an accurate weight and sample for payment. The hauler will get efficiency in terms of the turnaround time to pick up that farm, so they can make a significant saving there. And the processor will get okay, you know, efficiency of the entire supply chain, yes, but also reliability of the weight and the test that they are going to then deal with their customer, the cheesemaker or whoever it might be at the far end. So reliability of those insights around weights and tests and also the integrity of that sample. So to be able to know that that sample has been hygienically taken, they don't have to worry about that. And then visibility of their entire supply, supply chain. So again, that idea of visibility all the way back to the farm. So it's interesting, you know, looked at one way, we're selling a piece of equipment that lives on the farm and belongs to the farmer. But looked at in another way, we're selling a piece of equipment that has the ability to deliver value for the producer, for the hauler, for the co op, and even ultimately for the processing. Let me tell you a story about dairy sampling. There's one farm in, in the Northeast and you might or might know this, but co ops pay a bonus according to how much protein and how much butter fat you can get into your sample. This one farmer in the Northeast kept missing out on his quality bonus. Month after month after month. He wasn't hitting the quality bonus, but he was feeding his cows perfectly. He was managing their nutrition. He couldn't understand, okay, why am I still missing out in this quality bonus? And eventually he came to us and we were able to put in a system for him on his farm. And suddenly he was able to achieve the quality bonus. Now he was delighted because it meant $10,000 to him. So it was, it was worth doing. And it probably paid for the system in a couple of months. But why it happened was because he was a direct load farm. And on direct load farms, they, as most farms, they milk the cows in groups. What that means is that the highest producing cows are all milked together and the lowest producing cows are all milked together and all the way through like that. So you go from the highest to the lowest, or vice versa. So with a Piper system, you're sampling in tandem with the flow of milk through the line. So we get a representative sample by making sure when the high producing groups are going through, we're sampling much faster along with the flow. When the low producing groups are going through, we're sampling much, much slower. And when no milk is going through the line, so when everybody's on a coffee break, we stop sampling. Now with old fashioned sampling, they might be sampling the same cow over and over and over and over again. And that's what meant this farmer was missing out on his quality bonus. But when he put in the Piper system, suddenly overnight, he was getting full representation of all those groups that he was milking. And it turns into a weighted average sample rather than a simple average sample. And that was why he was able to achieve that quality bonus.

B (24:14)

For most dairy farmers, financial stability is a goal. Not really. A matter of fact, selling to processors isn't always reliable. And tech innovations like those of Dairy Masters and Piper systems aren't easily accessible to the market. Sometimes they just don't fit into the budget. And because milk is a commodity, if those prices on the market go down, farmers tend to be the ones who feel it. Most things like labor shortages, droughts, disease, regulations, and dire economic conditions can completely offset a farm's production. Technology on the outside is a practical asset, but on the inside, it can be an elusive miracle. Still, for some innovators, there's plenty to address. Optihal is an Irish software company that works specifically with dairy cooperatives to optimize the transportation of milk from farm to processing plant. It's an example of a less invasive technology that can lessen obstacles and reduce emissions in the industry. Here's Gary Gallagher, the company's CEO and founder.

H (25:12)

Ultimately, our software helps people like transport planners, like operations managers, like logistics managers. It helps them to save time and to find sort of new savings around reduce mileage and reduce emissions and reduce costs. So instead of me as a logistics manager having to spend hours and hours trying to work out how I might make some minor changes to, like, a fleet of 100 trucks. Our system does it in a matter of minutes. You know, it understands milk, it understands what happens on a daily basis. So it saves our customers, you know, hours and hours daily in terms of calculating the most efficient route plans for this part of the supply chain. There are people typically in an office, like a truck depot, and their task is, okay, next week I have to pick up milk from these 2,000 farms, and I have to deliver to, let's say, up to about 50 or 60 different plants. Okay, so I've got all this supply coming from farms, and I've got all this demand coming from the processing plants. So it's up. And in the middle, I've got, let's say, a fleet of 100, 150 trucks that I have available to move the milk from one to the other. So it's my responsibility then to plan out the most efficient way to complete that operation. And, of course, I've got so many constraints to think about. There's only specific times I can visit each farm because of milking times. So across 2,000 farms, I have different collection times at each of the farms. So I've got to think of all of that. Some farms can be accessed by big trucks, some farms cannot. I've got different milk types, I've got different size trucks. I've got different plants who want different delivery times. So there's a huge amount of constraints on that planning. So what I do is, what our system allows the user to do is to ingest all of the available information about the milk at the farm, the trucks that are available, the plants where it's going to. It ingests all that information and it automatically, using a mathematical algorithm, it automatically calculates the most efficient way to do that. So it will give back the user a full route plan with a list of maybe 100 different routes. That is the lowest cost solution for them to achieve that.

B (27:45)

Ensuring that milk is properly distributed is as important as the quality of the milk. It prevents slumps in the supply chain and promotes a good relationship between farmers, processors and distributors. And in one other crucial aspect, it reduces carbon emissions. About five years ago, it was reported that the dairy industry contributed to about 1.5% of total US greenhouse gas emissions. That includes methane and nitrous oxide emissions by cattle, and, of course, milk transportation. But by 2017, the industry reduced nearly 20% of its carbon footprint from the 10 years prior. And new technologies include including Optihull software, can lessen the impact.

F (28:24)

Technology is shaping the dairy industry and will continue to shape the future of the dairy industry. There are again, a number of challenges and a number of opportunities for the dairy industry for which technology or some form of technology is at least part of the answer is at least part of the solution. Technology will probably never completely solve all the problems, will never help farms take advantage of all the opportunities, but it has a major role to play.

B (29:04)

Julio, like other experts in the field, is confident that technology will continue to expand the growth of the dairy industry. It's actively contributing to a major part of the global economy that's long been overlooked as too physically demanding. But recent trends in automation and data analytics are shepherding in a new movement that allows for greater agency, independence and growth for everyday farms, both big and small. Because at the end of the day, milk is a commodity not just on the market, but also at home. It's the milk millions of people across the world pour over their children's cereal or use for baking or blend in smoothies. It's what keeps many of us alive.

G (29:43)

You know when you see a truck going down the road and it has milk in it, you know that it's going to feed a family and you know that it will be probably the best nutrition they can get. And being involved in that, I think it's one of the things that makes me feel really good about being involved in the dairy industry. I think there's a wholesomeness and a value to it and that we often forget.

A (30:13)

Thanks for listening to the next Innovation. This series was produced by Situation Room.

B (30:17)

Studios and Powered by Enterprise Ireland. Investing in the next wave of innovation. Our executive producer is Christine Barata and our senior producer is Sharon Barreiro. Emily Beeman and Leila Sharowi are the associate producers. Additional production assistance by Global Situation Room and a Special thanks to Dr. Julio Giordano. I'm your host Jennifer Strong. Until next time.