MSU Extension MI Ag Ideas to Grow With

 - Hi, my name's Lyndon Kelley, I work for Michigan State University and Purdue extension. I do mostly irrigation work, but today, we're helping work with the group of MSU field crop agents, putting together today's Michigan ag ideas to grow with session. Before we get started, I'd like to thank the sponsors that we have Greenstone Farm Credit, North Central Sarah program, that'd be the sustainable ag and research and education program. And then from the Alaska group here, the great thing about doing webinars, you have the power to get along ways from home. Mark Wolbers and the Alaska Pioneer Fruit Growers Association have all helped us put together today's programming. So our first speaker today is Dr. Scott Shearer, he is a longtime member and the chair of the Ohio state, agricultural sorry, Food Agricultural and Biosystems Engineering program. He's going to be talking about some specific information as far as advancements in control of farm equipment. And we bring Scott up. - Thank you very much Lyndon for the introduction. Make sure I get my... There we are. I will go ahead and share my screen and we'll get started. I wanna begin by recognizing this title indicates what we're gonna see in the next 10 to 20 years. I think it might be appropriate to reframe this a little bit in that it's what we're seeing today that will probably be adopted over the next 10 to 20 years. And so it's an interesting time right now. If I think of back over my entire career as an agricultural engineer, and Lyndon indicated that I'm now chair of a similar academic unit here at Ohio state to agricultural engineering, everybody's changed their name across the country. But when I go back in time and I begin thinking about my work in terms of automation, automating agricultural field machinery, a couple things come to mind, and that is early on in my career, technology was expensive. Today, the cost of that technology has come down substantially and I'll give you some examples as I go through the slide presentation here of things that are pretty common today. The reality of it is that we're really only constrained today by our imagination in terms of how we're gonna utilize technology and I think when I get done with the presentation today, you'll have a better feel for where we are headed with technology and agriculture. And we're on the cusp of some pretty interesting times right now because of the evolution of some, some recent technologies, one of those being artificial intelligence. So with that, I'll go ahead and get started. My intent is to put things within context. So overarching questions, I'm gonna attempt to answer these. I may not do a perfect job, but this is where I'm headed, where my thought processes are today. What will drive automation adoption? And there could be the regulatory landscape, there could be profitability. I'm suggesting that both the regulatory landscape and profitability are gonna drive what we're doing in agriculture. What will limit automation adoption? I think there are some red herrings that we have to be aware of and there's one in particular that concerns me greatly, I'm gonna address towards the end of this presentation. How will equipment ownership change? I think historically, one of the things we've always thought about is farmers own the equipment and they might do custom work certainly, they might lease equipment, but generally it's the farm owner operator that's in control of this equipment. And I'm suggesting that may change. How will business models change? And as we move in towards automation, there are gonna be those providing the automation and the question is, is how are they gonna be rewarded in terms of this automation? And I got some examples that I'll talk about that are very specific. Can farmers afford to control all aspects of their operations? Farmers historically have been independent business people, I appreciate that. I also appreciate the fact that they, they have to be well schooled in about 10 to 12 different areas to be profitable. They're capital asset managers, they're HR experts, they're agronomic experts that they're experts at managing federal programs, they're experts at managing machinery and technology and so again, I see things that have evolved in agriculture and what I'm suggesting is, in the future, it may be difficult to remain and have that expertise in all areas so that you can take advantage of things. And I know that one of the things that often trips up farmers are marketing the products that they do produce. Okay. And then this last one, what disruptive technologies are on the horizon. And I think that is perhaps one of the most important questions that I'm gonna try to answer today. Here's a graph. This is of ballasted tractor mass for tractors tested at the University of Nebraska. These happen to be John Deere tractors and the only reason why I chose John Deere and I noticed I got red tractor on there and I'll come back and talk about that. But when I look at these tractors tested at the University of Nebraska, there's really two, two time periods that I think are noteworthy. The first one is 1920 to 1960, and we noticed that we had a modest increase on an annual basis in terms of the heaviest tractors tested. These are all internal combustion engines. I wanna make certain that everybody understands that we go back to the days of steam traction engines, obviously gross vehicle weights were a lot higher, but starting in 1920, everything kind of switched from steam to internal combustion engine. You'll notice that modest increase in ballasted tractor mass was really kind of governed by the spark ignition engine, whether it was run on kerosene or gasoline. In 1960, though, the diesel engine becomes popular. And I know there's a transition period that lasted over several years, but you'll notice that once we got to compression ignition engines, the increase in ballasted tractor mass really takes off. And the point I wanna make here is if we look back over the last 60 years for the biggest John Deere tractors tested at the University of Nebraska, that increased tractor mask again, largest tractors tested, increased about 870 pounds of force per year, year in, year out for 60 years, which tells me currently, I don't know anything that's gonna reverse that trend. And the other thing that I wanna point out is while these are plotted for John Deere tractors, any manufacturer would fit into this category similarly and that's the reason why I put the red tractor on here. You'll notice the 620 on the hood. Well, that's 620 horse four wheel drive tractor. The other thing is, and we can talk about averages and numbers and things like that but generally, if you get the operators manual out, they're gonna recommend a certain ballasted tractor weight and that's usually gonna be per engine horsepower, okay. Depends upon whether you're pulling high draft tools or low draft tools, but those numbers could range up to 130 pounds of ballasted tractor mass per engine horsepower. So necessarily as engine power goes up to harness that engine power, if you're gonna turn it into youthful work at the draw bar, you're gonna have to have ballasted tractor mass. And I think it kind of goes to this issue today and what I'm concerned about, and that is what's going on with soil health as increasingly we see larger and larger equipment on north American farms today. This size issue is not strictly limited to tractors. The Fendt Ideal 10 recent released combine from AGCO, 790 horse. And so I don't know that I could have ever imagined in my career as an agricultural engineer ever seeing an 800 horse combine, but this certainly gets people thinking about where things are headed. Brent, which is a product manufactured by Unverferth, here in the state of Ohio, now is offering 2,500 bushel grain carts. So think about this, when you take the base weight of the grain cart, 40,000 pounds, 38,000, and you add to that 2,500 bushels of grain be it weever or soybeans, we have a gross vehicle weight just for the grain cart itself, approaching 95 tons. And I keep urging people to think about what's happening, not to the top 10 to 15 inches of soil, but what's going on below. And I'm suggesting, and we have some evidence to suggest or justify the fact that when we get these large axle loads, and again, I realize this is on tracks and I get the contact pressure is a little bit lower, but when we get these high axle loads, that is what's really contributing to what we consider to be deep compaction. In other words, below the top soil layer and in some cases up to 40 inches deep into the soil, or about a meter into the soil. So as gross vehicle weights go up necessarily, I think we're beginning to see more long term concerns relative to soil compaction, especially at deeper depths within the soil profile. Why do farmers till? And I think we're kind of in a bit of a vicious circle here. We know that at least here in the state of Ohio, we have a lot of farmers that perform deep tillage, or I'll say run a disc ripper about every two to three years because they see the yield response. I also appreciate the fact that we're advocating for no till in conservation tillage farming. But when I come back to this question, why do farmers till, we can talk about aerating soil, we can talk about mitigating compaction, we can talk about improved infiltration. At the end of the day, the reason why farmers till is because they see a yield response the year after they till. And so what's gonna break that cycle, I don't know, but as we increase equipment size, this may become more prevalent. This is a, a 3D depiction. These are CT scans and they come from an institution in Europe, but I think they're very noteworthy and that the one on the left, where you see the green distribution of dots is the void space within the soil and this is pre traffic. Okay, this is before we have an axle load going over. And where you see the greater collection of green dots in kind of an aggregate form, those are macropores. Those historically, what governs in some respects, infiltration into the soil profile during rain events. When we look at the... And that's the image on the left hand side pre traffic. When we look at the image on the right hand side, this is after a 35 PSI load has been applied, I wanna remind everybody that generally, grain carts on tires have pretty high inflation pressures. And again, you cannot avoid the manufacturers warranty and expect them to stand behind the tire. But if we look at the one on the right hand side, after a traffic event at 35 PSI, we'll notice that at least in the upper part of the soil profile, most of the macropores have collapsed. Okay. We begin thinking about the ideal soil condition, which is about 50% soil particles, 25% void space for air and 25% void space for water and you can see that that's been compromised in this soil because of the traffic event. I wanna remind people too, that generally we're gonna see the collapse of macropore space in soils that are wetter, especially around harvest time and or spring planting time. So again, dry soils are not gonna experience this sort of things because the soil strength is higher and they will support the rates. We get into wetter soils though, we're gonna see this type of activity. We have been working with CAMSO, a track manufacturer. One of the things we're trying to do is, is get a handle on what's going on within the fields, in terms of compaction. One of the primary ways we do this is with the Soil Cone Penetrometer and I wanna, I guess, note at this point in time that Soil Cone Penetrometer readings are a proxy and somewhat respects for soil bulk density. So I wanna be careful because, what we're measuring is soil strength. It is related to bulk density, but I wanna be very careful to acknowledge the two are substantially different. The other comment that I wanna make is in general, when you're collecting Soil Cone Penetrometer readings, you wanna certain that the soil is near field capacity. I think a lot of people make the mistake of collecting Cone Penetrometer readings when the soil profile has not been rewet in the fall, and they run the risk of not being able to do fair comparisons between what's going on in the field. So we have a lot of data that we've been collecting, but I also wanna get people thinking about traffic events that occur in the field. And we have a group here in the US and a large group in Australia talking about controlled traffic. And I'm a fan of controlled traffic, don't misunderstand me. But on the left hand image, I have here, I have a four wheel drive tractor with 800 millimeter section with tires on it, pulling a grain cart. And so I just kinda laid out the wheel spacing and I realized that that farmers could do a better job of spacing tires and choosing tire configurations. But in this case, when I look at the amount of soil that's trafficked, it's about 48% of the field and that is if the grain cart is chasing a combine with a 12 row head. The other thing I wanna get people thinking about is, is as grain cart sizes increase and by the way, this was only a 1500 bushel grain cart, but as grain cart sizes increase, we're actually trafficking more of the field with greater loads than for smaller grain carts. And I always question farmers about the size of grain carts, because I'm wondering to myself if they may be substituting grain cart capacity for trucking capacity. By the way, when I begin looking at some of the combines today, class 10 machines, my guess is in good corn, they can probably shell, I'm gonna say seven to 10,000 bushels an hour and again, it's gonna depend upon how the farm's set up. When I look at that ideal ten seven hundred ninety horse, I'm certain that you can probably get pretty close to the 10,000 bushel an hour. What I begin asking farmers is, is how do you move 10,000 bushel an hour out of the field? If it's only 7,000, that's fine. Several years ago, we were shelling 6,000 bushel an hour with a class seven machine. So when I give you some of these numbers, I realize that depending upon the field efficiency and number of other things, they may not be there, but under ideal conditions, dry corn, certainly 10,000 bushels per hour is within reach with some of today's equipment. On the right hand side for reference, this is a 10,000 pound machine. Yes, it's only gonna be 10 foot wide, and we're gonna, again, traffic a fair amount of area, but I wanna get you thinking about the depth of influence because of gross vehicle weights and axle size or axle loads. Again, we see compaction influences down 41 deep, 41 inches deep on the left hand side, we look at a 10,000 pound with, I'm gonna say moderate tire section widths. We always see an influence going about 13 inches into the soil and again, we're only traffic and about 33 of the area. So I just wanna get people thinking about controlled traffic and tire size and selection, and some of the challenges that come along with attempting controlled traffic. By the way, with AB lines today, we are controlling our traffic. And I think some of our pinch rope planter compaction could be the result of multiple or repeated passes year after year so just throwing that out there is something to think about. So why do farmers till? Left hand side, what we see is these are multiple, I forget there was something, five or 600 Cone Penetrometer readings here, but if you look at that blue line, that's kind of in the middle of the range that we saw at different depths. When I get over 200 PSI, I begin getting concerned about crop growth. We know that usually the influences are pretty substantial when you get to 300 PSI but the other thing is, yield loss due to compaction usually manifests itself in terms of notable reductions in yield during dry years, which begins to make sense. So anyways, we look at that blue line in the middle of the figure on the left hand side for conventional tillage, we see what we call a compaction nose that extends from probably about three inches deep in the soil, maybe down to about six or seven. A lot of people talk about that being the tillage pan and the tillage pan might well be, something to be concerned about, but I think most of that nose is occurring because of traffic events. Okay. On the right hand side, the field has been deep tilled, and you'll notice that compaction nose in the three inch to seven or eight inch depth has been relaxed substantially and I think this is the reason why we see a lot of farmers doing deep tillage. Please note that that tillage relaxation occurred down to what, about eight inches, maybe even 10 or 12 inches deep into the soil with that deep ripping before we get below where we could effectively reach with a tillage tool in terms of alleviating the compaction. I'm not a big fan of deep tillage, don't misunderstand me, but you begin to, I guess, see it from a farmer's perspective when that compaction nose relates to a reduction in crop yield. And so I get it. So what's the solution. Well, every few years, I change this slide or whatever. I use this in multiple presentations. What I'm trying to get people to think about is there's a lot going on with respect to automation and agriculture. And we see a lot of autonomous vehicles coming to the forefront. Some of these are constant vehicles, they may never see the marketing light of day, others are real, and they are available today. I'm gonna point out a couple of these. This image up here of the Kubota M5 with a toolbar on the back with row units, this is Sabanto Ag and I'm gonna come back and talk about Sabanto Ag because I think they're in a very interesting position in terms of, what they're attempting to do. Center left is a Yanmar tractor, this is actually a fully autonomous tractor being marketed in Japan. Okay. Not to be outdone Kubota, the bottom row, kind of the modern looking tractor with tracks. They introduced this concept vehicle, I'm not certain where that's at right now. Kind of an interesting one. Bear Flag Robotics is modifying existing tractors in this case, looks like a John Deere 6R is what's been my modified. Deere had this sprayer, this high clearance sprayer here, small high clearance sprayer, introduced it a couple years ago at Ag Technica. Over here on the right hand side, this orange sprayer is SwarmFarm out of Australia and I'm gonna come back and talk about that a little bit more. And then we also have DOT out of Canada, and they're been some significant changes in terms of which companies are acquiring, which companies here as well. This is kind of going the other extreme in terms of small. This is Xaver, which is AGCO's concept, a fully autonomous tractor, but this is seeding one row at a time. You can see the GPS receiver on top, relatively small machine and most of you are beginning to think to yourself, well, if I got a 16 row or a 24 row planter, I'm gonna need a whole bunch of these and yes, you're right, you're gonna need a swarm of these vehicles. But we also need to begin thinking about, the bearing goes out on a 24 or a 36 row plant or the entire planter's down, a bearing goes out on one of these and you have one of multiple machines that is down for service. So again, just thinking a little bit about this in terms of where things might be headed. A couple years ago, I visited SwarmFarm in Australia, they have been building autonomous sprayers, the year that I was down, they had built 20 of these and I asked the company owners if they had sold all of 'em, they looked at me, no, they haven't sold any of 'em. I'm thinking to myself, what kind of business model do you have where you don't sell anything? And they said, well, we don't plan on selling these, we lease 'em to farmers. Okay. So I asked about the lease cost and what came along with the lease, a technician comes out, sets up all the fields, the machine set up it's operational in the field. The farmer has to pay for fuel and lubrication and the farmer has to obviously pay for the chemicals, going into the sprayer tank and then reload the sprayer. But when I got done, the cost of spray application with these machines was, and again, if I'd done my conversions correctly, this is fallow, so a bit of a different agriculture. They're attempting to kill green vegetative material in the field during the fallow part of the season to conserve moisture and certainly you get that in terms of crop production, but their cost of spray application was about a buck, an acre when you looked at the least value of the machine. Now, obviously chemicals are above and beyond that. But I begin to compare and contrast custom application rates in the Midwestern part of the US with high clearance sprayers with what's going on with SwarmFarm. So again, I think there's something there, it's gonna be interesting to see how this evolves, a bit of a different business model. Again, they're leasing machines. By the way, there's probably a term here I need to talk about a little bit. You'll notice that FaaS the two lowercase, a's, this is kind of a takeoff on software as a service, as we see in the IT sector. Well, think about farming as a service. In other words, are we gonna contract a contract with companies if you will, to perform specific operations on various farms. I like this one here. And I talked a little bit about Sabonto AG before. I think Craig Rupp was running eight of these. There were four higher horsepower M5s, and then five or four lower horsepower M5s, but these are fully autonomous. If you follow Craig Rupp on Twitter or follow Sabanto on Twitter, they've had some interesting occurrences. The sheriff has visited their fields on several occasions. People in the area are concerned that a farmer might have fallen off the tractor and the tractor was out roaming the fields on its own, so to speak. But Craig Rupp has been doing this for the last two or three years, and oddly enough, he's been seeding soybeans and so... And I kind of asked myself, what farmer's gonna pay to have their soybeans seeded. We know that one of the things going on in the Midwest, it's going on here in Ohio now, farmers are finding if they plant their soybeans first as opposed to corn, that they're much more profitable. And so one of the things that I want to get people thinking about is the fact that it used to be, we planted corn for about a month, starting what, April 15th through May 15th, and then May 15th, we began planting soybeans until about June 15th. What I would suggest is given the crop response to environmental factors that really, if we've seen this window shrink to probably planting corn and soybeans in the April 15th to May 15th timeframe. Aaron Wilson here at Ohio State, who is a meteorologist talks a lot about shifting weather patterns and one of the things that, that he has been showing individuals is at least in the state of Ohio over the last 30 years, we've lost five working days during the spring time period and five working days during the fall harvest period, spring planning versus fall harvest and so, again, as we see shifting weather patterns, the question is, is how do we become more time with our operations? Craig Rupp, an interesting concept, he's thinking about 24 by seven operation. One of the other things I wanna get you thinking about though, is you have to tend this equipment. When I say tend it, you gotta fill it with seed, fuel and the like to keep it running in the field. Craig was tweeting several times in the past that these tractors were running 30 to 40 hour shifts. I wanna remind you, they're running 30 to 40 hour shifts. They don't require vacation. There's no medical benefits that have to be supplied. There's no retirement programs. So, again, begin thinking about where things might be headed in the future as we move towards automation. By the way, seeding operations and soybeans, if you're a little off that's okay, soybeans are pretty responsive and they will fix some of these problems that do occur. So I get it where farmers would be a little less concerned about the quality of seeding operations when you begin to automate those with soybeans in comparison, or contrasting that with corn. Here's a statement that was made by Craig Rupp in an interview recently, he said, "as a general rule, machinery cost for a planting operation are 25 cents per horsepower hour." This would be with large fixed frame tractor. This would be a Deere 8R or a Magnum tractor from CASE IH pulling a decent size planter, probably a 16 or 31 road, depending upon whether you have splitters or not. His statement is, "ours are down to 7 cents per horsepower hour." So in other words, you take the engine horsepower, you multiply it by these numbers and that is what the cost of these seeding operations are gonna be at least on the machinery side. So please note that's a, that's about a three or 3.5 to one ratio in terms of moving from a human operating a tractor to it becoming fully autonomous. So Farming as a Service. And what I ask is, is Farming as a Service the future of the equipment industry? When you ask Craig Rupp what he sells early on, he said, I do not sell hardware, I do not sell software, I do not sell technology, but rather I install seed. And so from Craig's mindset, at least going back to this, he's very much focused on Farming as a Service. And so is that the new norm in agriculture with automation? I think it could get interesting. I think this quote is very appropriate and at random, at this point in time, but Henry Ford, "if I'd asked people what they wanted, they would've said faster horses." And so today when we ask farmers what they want, they wanna be more timely but the primary way of doing that is through bigger equipment, not automating equipment, removing the operator from the tractor and going small. So anyways, stay tuned, we'll see what wins out. Personally, I'm thinking smaller autonomous equipment, but lots of equipment may rule the day in the future. Here's an interesting one. This is a Ravens DOT. I'll say Ravens DOT, Raven purchased DOT. But recently at the end of 2021, CNH Industrial bought Raven, I think it was about a 2 billion dollar acquisition. When they bought Raven, they ended up getting two fully autonomous products. I'll say fully autonomous, but two autonomous products. One being DOT, which is a set of power wheels if you will, that can be adapted to various different equipment compliments in terms of being able to put sprayers on here, seeding equipment, fertilizer application equipment. My understanding is in Western Ohio, we have one of our retailers in Ohio that's using DOT now for fertilizer application. So this technology's beginning to emerge, there are some people that see a cost effective component to it, but I wanna remind you that as some of these startup companies begin to mature and they're kind of ripe for takeover by larger manufacturers and I get it. Larger manufacturers are very concerned with their existing product lines and in general, I don't know if they have the stomach for venture capital applied to some of these new concept, but certainly when these concepts mature to a point where the major companies think that there is a market there, or it will begin to affect their market share, they begin to transition. I think that's one of the things we've seen in the machinery industry most recently is this thing about automation while there were a lot of concept vehicles out there are really becoming part of the mainstay product offering for some of these companies. Going back to Henry Ford's quote, everybody begins thinking about machinery in the field today, and it kind of taking the form that has in the past, well, I want want you to look at this. This is a sprayer that's a bit different. This is ecoRobotix, and this is a (indistinct) vaultek, in other words, solar cells, please note there's two robotic arms and they're applying micro sprays to weeds. So in other words, they're only spraying the weeds and please note the positioning of the spray nozzle on those end defectors. And again, what I go back to is thinking about agriculture in a bit of a different mindset and us getting outside of what we believe to be traditionally equipment is, is the end all to what's going on. So Unmanned Aerial Systems, I need to make a comment about this because as the FAA, at least here in the United States becomes more comfortable with integration of Unmanned Aerial Systems into the aerospace and I talk about small Unmanned Aerial Systems. They're becoming much more widely accepted in agriculture, especially for sensing, okay. And they have a role to play there, there's no question about that. But I also begin to think about what about spray application with Unmanned Aerial Systems? We're still subject to the 55 pound limitation, which if you look at the drone on the right hand side with the spray boom on it, you're limited about two and a half gallons of spray capacity. And if you think about that and you think about labeling requirements, really, we can't spray too many acres necessarily. Now there are companies that are looking at swarms of drones, which expands that, in other words, putting three or four of these in a field at a time. This came... The image in the middle VoloDrone came from Techcrunch, this is a company that I think Deere has partnered with in Europe. But if you look, they have about 60 gallons of spray capacity on this. And this drone is obviously above the 55 pound limitation we currently have here in the US. But please note that if you're lifting 60 gallons of spray material, it has quite a bit of lift capacity. That ring around there is about six and a half foot off the ground, just slightly taller than human being but you also see the spray boom with either 15 or 20 inch nozzles. So again, let's begin thinking about how Unmanned Aerial systems might influence what's going on. Here's another one and I've had an opportunity to talk to a lot of farm groups and I talk about robotic irrigation. I had a farmer raise his hand and goes, we've had that for a long time, it's called center pivots. And while I agree with that, conceptually, this is a bit different. This is a 360 yield centers, rain, and it's a new product that's coming out. I think they're gonna have a limited release this year, maybe five or six machines out. But the point that I wanna make is they're going to 120 foot boom on this with effectively, there are wide drops for applying water. This is a low volume machine, about 200 gallons, a minute. It is independent or operates independent of field shape. It is independent of the direction that the crop was planted. Please notice it is a high clearance machine as it unreels and places the hose between two rows of crop going through the field, it irrigates and then it turns around and comes back, picks that hose up and irrigates on its way back to recover the hose. My understanding is a central connection point in the middle of 160 acre field and this thing will handle about 160 acres a year making about 12 passes over the field. And I begin thinking about if we're making 12 passes over the field in this case is to apply water, maybe we might need to be thinking about how we also use this to apply nutrients as well as crop protectants. The other thing I think that is very noteworthy is, if this machine is proven, what value does it bring to animal producers? And the thought here is for dairy and swine farmers, it's really opened up nine months of the year for animal nutrient source application. Standing crop, it doesn't make any difference what the height of the crop is. But I also begin thinking about how we use our nutrient sources more effectively by applying all during the growing season and meeting the crop needs and hopefully avoiding some of those large rainfall events where we see a lot of offsite movement of these nutrients. So again, robotic irrigation, I think it's here, I think it's gonna at least be part of a solution to a bigger problem in terms of managing nutrients. We see a proliferation of In-Situ sensors. Those are sensors that are installed in the field. Teralytics has been advertising their soil fertility sensor that's placed in the field and obviously internet connected. CropX has been doing a lot of great work in terms of soil moisture. We also see a densification of weather stations on farms throughout the US. Today, it's my thought that farmers probably know more about the weather and the distribution of it than the national weather service does. When you think about the densification of weather station, certainly in the Midwest, farmers are one of the primary sources of weather data to day. Maybe they're not certified in accordance with national weather service standards, but nonetheless, a lot of good information coming off these stations. Central Ohio's kind of become home to data warehouses. All the major data warehouse groups are in central Ohio now. Amazon, Facebook being two of the bigger ones or AWS, Amazon Web Services, but was talking to one of the database companies Oracle, couple years back. And they were suggesting that up to one third of the data stored in these data repositories potentially will be coming from food systems. And when we think about food systems, we think from seed to table, if you wanna think about it in that context, but certainly we know that most of our machinery now is internet connected. We know that when we buy tractors from John Deere or from a CNH Industrial, they automatically come connected to the internet and then farmers that are upgrading existing equipment, have a multitude of options. And I've listed a number of those here, Farmobile, Climate FieldView, Raven Slingshot, AgFiniti from AG Leader and the Trimble ConnectedFarm. This is in addition to what AGCO CNH and John Deere offer. I don't know whether everybody realizes, but the minute you turn the key on modern farm machinery, it's transmitting data from the controller area network or the canvas, if you will, to the cloud for archival purposes. Farmers have to intentionally turn this off in the case of some of these companies. Deere several weeks ago, introduced the 8R autonomous tractor, okay. And a video clip here that was provided by Deere. But if you'll notice there's not an operator in the tractor, in this case, the tractor is pulling a chisel plow. And we begin thinking about, gosh, when John Deere announced this product offering, they did this at the Consumer Electronics show, they did not do it at the National Farm Machinery show. Well, not specifically, or exclusively there in terms of the first introduction. They didn't do it at Agro Technica, they did it at the Consumer Electronics show in California. But as we move forward and we begin thinking about when CNH Industrial buys Raven, which had two autonomous concepts, they had Autocar, which was an autonomous grain cart and they also had DOT. Deere now announces an autonomous 8R, there's a question about how they're gonna market this and what connectivity is needed, but this becomes real now. It's not high in the sky, it's not five or 10 years down the road, these products are available today. I do want to note that one of the, I think primary technologies that Deere added to this tractor was the stereo vision system. There's actually six of these on the tractor. If you look up front on top of this set of weights, there are three stereo vision systems, one looking at either side, one looking ahead. When I look at the cab, I begin seeing this stereo vision system and it looks like there's one over this side, one over this side as well. My understanding is six total. These are collision avoidance systems. And when I say stereo vision, they're using stereo vision for ranging to get the distance to these objects that are detected and then they're using artificial intelligence to analyze the images coming from these cameras. The thing that I think that's interesting is in one of these advertisements, they made the statement that they used 50 million images for training the artificial intelligence in the StereoVision system. And so this is the technology that is changing the necessity for the operator to remain in the cab of the tractor. Once we're able to detect and respond to potential collisions within the field, the question really becomes is, do we need an operator sitting in the seat anymore? And for some operations, Deere is suggesting with tillage that it is likely to be one of the first operations that'll be automated from the sense of removing the human from the tractor cab itself. I would argue that today, the human operator and the tractor in many cases is really kind of babysitting technology so to speak. In other words, most of these tractors have auto steer, most of 'em have end of return automation. This person is really there to monitor the performance of the tractor implement combination and to respond to things that might occur that would cause these operations to stop. So let's talk a little bit about artificial intelligence. I got four pictures here, there's a reason for that. The young man on the left hand side is learning to pour milk and sure enough, he has a glass of milk, but I'm not certain about his efficiency. Okay, in other words, he got it in the glass but he also probably ended up with a lot more of it on the floor. The young lady in the pink t-shirt or whiteish pink t-shirt is pouring glass of milk and sure enough, she has the, the milk jug over the target. She's figured out, you gotta turn it up to pour the milk, but she doesn't quite know when to stop. Young lady in the orange t-shirt has mastered pouring milk from a milk jug, she's starting to turn it back up as the glass gets full and then the young lady all the way to the right has learned to pour from a different milk cart. My point is, as humans, we learn by making mistakes, and we remember these mistakes that we've made, and it helps us develop our motor control skills. It helps us develop our thought processes. It helps us attach meaning to a lot of the stimulus that we're receiving through our eyes and ears. We have a number of companies today attempting to do the same thing. We know the Googles, the Facebooks, the Amazons, most of you're aware of at least some of the artificial intelligence they're using. But suffice it to say there's about 900 other companies in the US now, a lot of them startups figuring out how artificial intelligence is gonna be incorporated into our everyday lives. My understanding is Google can tell whether or not you're sitting at your keyboard, regardless of whether it was you that logged onto their website or somebody else. They do that through the rate at which your keystrokes occur and the mistakes that you make, okay. Facebook is using what facial recognition in some cases there's a lot of thought about whether or not you can determine the moods of humans based on their facial features and the relationship with those. And then obviously we know what Amazon has been doing, and if anybody's been, well, Google or Amazon has looked at webpage advertising, they know you pretty well as an individual. So let's turn to agriculture and think about how we might use this artificial intelligence. I had a PhD student who just finished looking at artificial intelligence for detecting corn crop stress. Left hand side, you see nitrogen deficiency, yellowing of the mid rib of the corn leaves at the lower portion of the plant. By the way, this is not the top of the plant. By the time you see nitrogen efficiency at the top of the corn plant, it's a postmortem. At the bottom, we begin to see yellowing and leaf margins, which would be indicative of a phosphorus deficiency. And then we see a couple different corn diseases here, gray leaf spot, Northern corn leaf blight and we look at the distribution of lesions and locations on leaves and then as one might guess corn borer damage as the leaf is coming outta the world of the plant, we see the insect feeding damage. But these are all visual signs of different stresses that occur on a corn crop. By the way, we have similar stresses going on with soybean crops as well but think about when we have the ability to put cameras on the sprayer, going through the field and make threshold decisions about apply, or do not apply what it might mean in terms of the effectiveness of the inputs to the cropping system, in terms of crop protectants, as well as our cost of doing business. Most of you probably experience this. You're logging onto a website, there's a lot of these websites now using something called recapture. And recapture, you're trying to prove that you're not a computer, how do you do that? Well, you got a three by three grid of images here, and you might be asked to identify which images have crosswalks. Okay. Have you ever noticed that recapture has several other types of images? There's stoplights, there's bicycles, there's buses, a whole host of things. What I'm trying to get you to think about. If I go back to this slide up here, what I really need are subject matter experts to identify what the nutrient deficiencies are, and perhaps what the crop diseases are from these symptoms that appear visually. I begin thinking about endurance for soybean diseases here at Ohio State University, as well as Mark Lux for weed infestations. But in this case, what Google is doing is they're using humans to identify or annotate their images. So certainly you're going to demonstrate that you're not a robot on the first set of images, but if most of you are like me, I usually have to go two or three sets of images deep before I log on. Essentially what's going on is, is you might be annotating images for Google, for Google car and again, if you go back and look at the nature of these images, they all pretty much align with what Google is expecting in terms of a artificial intelligence to operate Google cars. So let's fast forward to what's going on today in agriculture. 2017, John Deere bought Blue River. Blue River was a company using AI and some image processing techniques to thin lettuce crops. Well, Deere has come out with a couple products now that basically uses this camera technology to direct spray material. In the case of fallow, you can spray only the green weeds, but more recently, what they're able to do is discriminate between noxious weeds and the cotton crops. And they're suggesting that we could see herbicide reductions at least follow up to 77%. And I've even seen some reports of herbicide reductions in crops like cotton of up to 90%. So think about artificial intelligence, being a replacement for some of our traditional herbicide or crop protective packages. Again, if we extend this to crop disease, we extend this to nutrient, mitigation of nutrient stresses in crops, it changes how we view agriculture in the future in terms of the effective utilization of some of these. By the way, same thing's going on in Europe, you have AGCO Bosch and BASF with Raven developing similar technology. I also want to note that AGCO recently invested in a startup company called Greeneye Technology, which is out of Israel, but using the same type of camera artificial intelligence combination for spray application. Here at Ohio State University, we recently received two NSF artificial intelligence centers. One of those is focusing on agriculture as use case. What I'm suggesting in the future, everything's gonna be internet connected, internet connected, including sprayer. Today, before you go into plant a field or apply nutrients with any kind of application equipment, you probably have to download an application file. That would be the prescription. In the future, what I'm suggesting is that same sprayer will have to download the intelligence for the activities that day at that location for that specific crop. And so, as we begin thinking about these artificial neural network classifier models that will be prevalent, there's gonna be a whole host of those and they're all gonna be based on use case. So I'm suggesting and John Deere has shown a couple of use cases to begin with, but I'm suggesting that for Ohio farmers, we might have up to 300 use cases which will require different intelligence pay loads. And so the question becomes, how do we generate those models if you will, those classification models. Where are they stored and then how do we get to those to the machine when they're needed in the former fashion that they're needed. So again, this is where things are headed in terms of artificial intelligence. Let me talk about one of the red herrings and that is broadband connectivity. USDA currently defines broadband internet connectivity is 25 megabits per second, download and three megabits per second upload. I want to caution everybody that download speeds are not necessarily what you wanna focus on as we begin to automate a lot of things going on in our farms. The upload speeds will be critical as well. My understanding is talking recently with one of our farmers who's using robotic milkers, there's new generation of robotic milkers coming out in the story that I understand is they're suggesting that farmers need a hundred megabits per second download and 50 megabits per second upload to make that robotic milker work correctly. And so, as we begin thinking about that, and we begin thinking about rural parts of America, especially in the mid west, the question is will broadband internet connectivity constrain technology deployment in agriculture, and ultimately delay profitability for some of our farmers. I'm suggesting our future needs are probably gonna be 300 megabits per second download and 300 megabits per second upload. And that's probably gonna be something that's kind of middle of the road for a lot of that autonomous equipment that I've been talking about before. In fact, John Deere was talking about, and I think I have this correct in terms of that autonomous 8R tractor needing 5G connectivity. There's gonna be a couple different ways they get that. In terms of talking about 5G connectivity and I realize I'm talking to a group of farmers a little bit farther north here, or a group of agricultural professionals. But what I wanted to do, what I wanted to do was show 5G connectivity here in the state of Ohio and obviously it's gonna be long interstate highways and major more populated areas and what's driving a lot of this is subscription rates. What does that mean? Well, if there aren't enough customers subscribing to 5G, the cell phone providers are gonna be less inclined to building additional towers to provide that 5G connectivity. By the way, there's 5G ultra now, which is coming online as well. So again, just kinda showing you the state of Ohio, we have 4G LTE over most of the state, although I'll be the first person to admit, when I drive in Northwest Ohio with my cell phone, I drop calls all the time. Well, one of the things you have to remember is voice has priority on most of these cellular networks. So again, think about the evolution of agriculture, think about broadband internet connectivity, and think about what's gonna constrain us as we adopt a lot of this automation. These are subscriber rates for broadband internet access, a lot of this is gonna be through fiber and cable. A bit of a different map, but again, we begin thinking about how do I get broadband internet access to mobile machinery in rural parts of Ohio? I think what's going on today. We see a lot of our county commissioners in the state of Ohio who have access to four funds to expand broadband, rural broadband. We see a lot of fiber going in in a lot of locations. The question becomes is how do we connect those farms to that fiber and most specifically, the automated equipment. So a recent study, and this is kind of an interesting one. Katherine LoPiccalo of the Office of Economics and Analytics for the FCC looked at the NAS database, National Agricultural Statistics service database. They also took their subscriber database for broadband internet access and she came out with these two conclusions, which I think are pretty interesting. Doubling the number of 25 by three megabit per second, per second connections per thousand households is associated with a 3.6% increase in corn yield. Now you gotta remember what's driving this increase is access to information. So it's farmer access to information. She also made the statement in this analysis that doubling the number of 10 by one megabit per sec, megabits per second connections, trying to get that out for 1000 households associated with the 2.4% decrease and operate expenses per farm operation. So if you think about that, and this is before we have connected machines, there's a lot of potential with this internet connectivity, just from the perspective of being able to supply information to the people who need it in a timely manner. Okay. So, building on this case, and one of the things I'm trying to get people to think about is, today, we talk about subscribers in rural America and whether or not there are a sufficient number of subscribers to justify build out a broadband connectivity. What I'm suggesting is we reframe that question and the question that I began asking farmers is what are you willing to pay per acre in the way of increased rent if you have broadband internet connectivity for that field? Okay. Well, if we do get that 3.6% increase in yield, and again, that's a big if, but if we do get that and hopefully we would expect more from these internet connected machines. Is it unreasonable with an additional income of $36 per acre using today's numbers 200 bushel per acre and $5 per bushel, maybe not everybody gets that. But are farmers willing to pay $10 an acre to get access to that $36 per acre of additional income. I think putting it in context, there'll be approximately a hundred to 200,000 available every five years for a 2000 acre farm to secure broadband internet access for multiple machines on the farm. And so, again, I'm talking about reframing the question and looking at it slightly differently. Another issue that's gonna be important is, is we begin to automate, we're talking 24 by seven operations, the question is, is what are our retailers gonna do to resupply that equipment? In other words, are they gonna be there to tender that or tend that equipment in the field? Are they gonna be there to refill it with seed fertilizer, fuel, herbicides, whatever the case may be. We know that a lot of our retailers in Ohio are gearing up to turn around and refill trucks in a very short order but the question is, will they be interested in tending these machines in the field in the same manner, in other words, delivering those inputs to the machines themselves. The other question I ask is what's gonna be the role of FBN, Amazon and companies like UPS in terms of supplying these inputs to farmers. We understand there are a lot of rural retailers that are doing the application in terms of spray application and that certainly that business is gonna continue to evolve. But I'm also thinking about the logistics of getting the right inputs to the right producer at the right time in the field so that it does not interrupt seeding operations. Okay. I throw this one out there just to get people thinking about things. I mentioned that we have five days fewer in the spring to plant, five days fewer in the fall to harvest. When we get those big rainfall events, how quickly can we get back into the fields with large equipment versus small equipment. And obviously a small equipment, we get the top two inches dried out and it will support traffic, it's a little bit easier to get back in the field. I'm gonna kind of bring this to a close with the Gartner Hype Curve, the Gartner Hype Curve. What is it? You got technologies on the left hand side, they come on the scene and the press picks 'em up, the farm press and other organizations. And they get a lot of people excited about this. So we go from the innovation trigger up to the peak of inflated expectations. Farmers begin buying some of those products, and we go quickly into the trough dissolution. In other words, the product did not quite perform as it was billed to perform. The next period is the slope of enlightenment and that is when farmers begin adopting these technologies and they begin making money with them because they figured out how to use 'em and effectively integrate 'em into their operations. And then we get out on the plateau productivity. So why do I show this? So this was by the way, it was put together by Ryan Rakestraw a number of years ago and it's probably changed a little bit, but I wanna remind you yield monitors and auto steer are out there on the plateau of productivity. What does that mean? I buy a John Deere 8R today, it comes with auto steer, I don't have a choice. It's baked into the price. There are no unlocks to buy nothing of that sort. So it's a technology that's mature. Yield monitors are essential for combines to perform their function. Again, I don't have an option. What I'm suggesting that farmers should be thinking about is adopting technologies on the slope of enlightenment. That is where the opportunity exists to make money with the technology. If you wait until the plateau of productivity, you're gonna be forced to adopt it, to remain competitive. And so places for our universities to be, is gonna be on evaluating these technologies and how they maybe applied on crop production within the respective states. And I think there's still a role for the cooperative extension service there again, because we end up being neutral third parties. We're not selling anybody, anything. So some take home messages. My take on this is digital automation, technology and agriculture is still evolving, but likely here to stay. I don't think anybody's gonna argue with that. Will likely be delivered via Farming as a Service. FaaS again, remember that Farming as a Service. Will expand equipment utilization, capital recovery in shorter periods of time. I wanna remind everybody when you buy a John Deere 8R tractor, a CASE IH Magnum or a Fendt Vario, those tractors are generally gonna have a 20,000 hour life, but farmers are only gonna use 'em about 500 hours a year. So you're gonna have a capital asset sitting in a machinery shed that is really not doing a whole lot for you much of the year. If we can begin using this equipment 24 7, it changes that cost structure in terms of capital recovery. We'll address skilled labor shortages. I think a lot of farmers this past year, experience this. Having people in the field and harvest time, reshape how farmers utilize crop production inputs. My example with Deere and an 80% reduction in herbicide application, that's a game changer. Will address soil compaction, multiple smaller machines, expanding conservation and no-till production practices. The one red herring in all this is, the thing I caution people about is, will broadband internet access limit ag tech adoption in rural North America?