John Deere See & Spray Technology: Insights from Industry Experts

Follow the Agriculture Technology Podcast on Apple Podcasts, Spotify and YouTube.

You can find past podcast episodes and view show notes by visiting our podcast website.    

Tony Kramer: Hi. I'm Tony Kramer, your host of the Agriculture Technology Podcast. I'm sitting down with agriculture technology and equipment experts to help you enhance your operation for today, tomorrow, and into the future. In this episode, we talk everything See & Spray at the Future of Farming event live from the Grand Farm Innovation Campus. We put together a See & Spray industry panel featuring Lee Johnson, See & Spray architecture and integration lead at John Deere.

We had Emily Paris, product manager at Blue River Technology. Nick Mazouch, agronomy manager at Mid-Kansas Co-op. We had Jared Bilodeau, a grower out of North Dakota. We also had on a couple of RDO employees, Mark Dorn, our corporate business sales manager, as well as Katy Hill, product manager for application technologies here at RDO Equipment Company. What really made this panel so informative and unique was the variety of experience on it.

Lee and Emily gave us a huge breadth of knowledge in See & Spray technology from their respected backgrounds at Blue River and John Deere. Nick and Jared both gave us great insight from a grower's perspective. Nick helped launch the first three See & Spray rigs in Kansas, and Jared has his own See & Spray Ultimate. Rounding out the panel with Mark and Katy of RDO, Katy supported solutions like See & Spray across the RDO footprint, and she also farms with her husband in North Central Oregon.

Mark has over 25 years of experience in commercial application equipment sales, focusing on the efficiency and profitability for RDO's customers. From diving into the software and camera detection technology, to going over results from 2024 trials, to even honing in on using the technology from growers' points of views, let's learn more about See & Spray technology live from the Grand Farm Innovation Campus.

(Grand Farm Innovation Campus)

Tony Kramer: To get started, I want to direct the first question at Lee and Emily. Let's go back to the beginning. Let's talk a little bit about how did See & Spray become to be? What drove Blue River to talk about or to get into this business and this industry? Then what eventually led to John Deere acquiring Blue River? Let's start there.

Emily Paris: I'll give a little bit of history for Blue River. That image on the left, that was one of our early prototypes. We were working in lettuce and lettuce thinning. We are based in Santa Clara's-- Right next to Salinas Valley, there's a lot of lettuce. That's where we started. Over the next 10 years, we got from that push cart mounted on bicycles to a See & Spray sprayer. This started in 2012. Again, we had a prototype in lettuce thinning. Over the next couple of years, we went from 2 seedings to 18 seed lines.

I think it was right in the end of 2015, we realized we need to get out of specialty crops, and we need to get into row crops. That's when we transitioned to row crops. You'll see, one of our first feasibility tests there in 2018, it's a small shrouded, pool behind, works on one row. Over the next couple of years, got to the point in 2017 where we were acquired by Deere. That really kicked off the transition to joining the John Deere platform.

2019, that's when we started doing our boom feasibility work. We had a half boom in the field in 2020 and a full boom in 2020 as well. Yes. From winter to summer. Then, first machines out there running with customers, operated by customers in 2021. Then we launched Ultimate in '22, followed by Premium the next year in '23. It's been a fast journey, I would say, from that bicycle mounted push cart to where we are today.

In 2024, so we hit a really cool mark. We sprayed over 1 million acres with See & Spray in 2024. We were seeing 59% average savings on those machines. From 2022 to 2024, we've seen a 20X adoption in machines and a 20X growth. We're in 15 new states using See & Spray. By not having to spray that chemical on the plant, the beans weren't having to metabolize it. They saw a three to four bushel per acre increase in yield in those crops as well. Exciting stuff coming out.

Tony: Anything to add on that, Lee?

Lee Johnson: Ironically, I got to be in the room at Blue River in 2018. We sat around, we looked at the technology that Blue River had. They were going one mile per hour in only lettuce, 10 feet wide, with a completely shrouded hood to control lighting and wind. We said, well, for this to really make a mark on the industry, we probably need to go 10, 12 miles per hour minimum. 120 feet wide is our most common spray boom. We surely can't control the light on that big of a machine, so it's going to have to be in natural light.

Certainly we're going to have to start in corn, soy, cotton arenas. What do you think about five years from now that we do all of that and bring it to market? That was in 2018. In 2022, in the fall, was when the first machines rolled out and got shipped to customers.

Tony: Yes, quite the timeline there, obviously, from that start, the conversations with Blue River and then the John Deere acquisition. Really cool to hear that story on how quickly it adapted and where we got to where we are today. You see the sprayer that's in the room, 120-foot boom, with that premium kit on there. Really cool to see that. Now the next question, I want to come over to Mark and Katy. From the dealership side, so we had our first See & Spray users in 2024, Jared being one of them with his Ultimate machine. Let's talk a little bit about what we learned as a dealership, the capabilities of See & Spray, and where we can take it into the future.

Mark Dorn: We learned that two-tank system was really nice. It afforded us the opportunity to put a residual product down, and then in a lot of cases, change up some herbicide chemistry, perhaps use a chemistry that would maybe be cost prohibitive to do in a See & Spray mode and really clean up some fields, take care of some challenging weeds, and again, to do that very affordably.

Katy Hill: Our RDO customers were able to cover about over 35,000 acres in 2024. We saw, between our Midwest and our Northwest territories, around a 67% savings in chemical reduction. One of the coolest things was it was a big learning experience for our customers and for our teams. We got to see how even with the business model, the way it is, how they were able to put money back into their pockets.

It was also really neat to see how our growers are stewards of the land, they care about their soil health, that's their land, and they want to take care of it. Any time we can implement a technology that allows for that while giving them a very good ROI, that's something that you get excited to represent. We also did some of our own ag technology field trials. Like I said, Midwest and Northwest combined, that in the Northwest we're running in more of a fallow application, and our Midwest growers are able to do more in crops.

We did see that 67% chemical reduction overall. We had five of our Midwest growers, we saw 56% savings, 43% savings, 67% savings, 31% and 44%. Everybody had a really good showing. It was a very wet season last year. Even when guys were not able to be in the field as often as they wanted before those canopies closed, it still penciled out that it was very worth that savings.

Tony: Everything that we saw, our customers utilizing it in the field, what were some of the things that we learned in 2024 when it comes to supporting the See & Spray technology?

Katy: I would say we learned how to communicate with Deere on our needs. We learned a lot for our own support systems. Our product specialists are amazing because they're the boots on the ground, and they're working with our growers directly. That's the real feedback that we were able to get from them. It's helped us ramp up a more robust support system, especially we're going from about 7 machines in 2024 to nearly 60 in 2025. It allowed us to really see what systems we need to put in place, what questions got asked, what were some of the heartaches that people had, and how we can find those solutions.

Tony: Great. Now I want to direct the questions to the end users. Jared and Nick, you got to utilize the technology in 2024. Jared, let's start with you. Let's talk a little bit about, before actually getting the technology in the field, was there skepticism? Were you nervous? Now, obviously, you purchased the sprayer, so you were ready to use it. What were some of the thoughts going through your head ahead of the 2024 season?

Jared Billadeau: There was some skepticism. One story I told earlier was when we went down to the product launch in 2023, or whenever it was, the dealer-customer product launch, I remember flying back thinking, "Boy, that's going to be a tough sell in western North Dakota with no-till and everything we do." A year later, I ended up buying one. Yes, that skepticism died pretty quick. I was excited for it. There was a learning curve.

Sprayer tips are extremely important. For those of you who have bought See & Spray, definitely talk to your dealer on that. We're going to get to that later on. I think my biggest take-home message from 2025 is when you're running See & Spray, it takes a long time to empty a tank of chemicals. You might not need that person running your water trailer or anything like that, which is a good thing in John Deere's eyes.

Another thing I really learned in 2024 is your fields are very predictable. It's crazy. After you look at a See & Spray coverage map, you can see a tree row that was removed 10 years ago, or you're going to see the perimeters of your fields are heavier in weed density as one would expect.

Tony: Nick, what about you?

Nick Mazouch: Oh, definitely. Being from a cooperative, you want to do a good job. We didn't want to have claims. We didn't want to have mistakes. We didn't know the technology well. We definitely found out quick it was not a beta version by any means. The weed maps were very interesting. Like you said, you can see things. You learn from those. You learn real quick where your problem areas are. Went very well, that's for sure, and brought a lot of interest.

Tony: Great. Let's go back to Lee and Emily. Now that we got an idea of what we saw in 2024, what the customers experienced in 2024, let's dive into 2025. How do anybody in the crowd that maybe has purchased See & Spray for the 2025 season, or maybe they've been thinking about it, how can we best set ourselves up for success? Part of that comes down to the operations that we're running this technology in. Lee and Emily, why don't you guys talk a little bit about limitations of the machine? We talk about weed sizes, maybe canopy considerations. Tell us a little bit about that.

Emily: Weed sizes, it's pretty wild. My first time out to the field seeing this machine work, we're hitting quarter inch by quarter inch weeds. It's thumbnail size. It takes you a little while to see what you actually hit if the machine hit it going 12 miles per hour. Weed size, quarter inch and up. We operate right now in fallow, corn, soy, and cotton. All of that is up to a 16-inch crop height. We are working in 30-inch rows or more is going with the row. In soy, you can get down to 15-inch spacing, and you can run at up to a 30-degree angle that you're running in. The main goal there is to make sure that we can see the weeds within those crop rows.

Lee: I think if you think about how we take on the challenge of there's 70 different crops out there at least, all these different row spacings, all these conditions, soils, all the different weeds. We broke it down into what are we going to start with. What Emily walks through, that's our first viable product. That's what's out on the market. That's successful. Is it the only thing that the machine can do? No. It's what we've verified. It's what we've rigorously tested.

I know there's going to be questions about what if it's 22-inch corn and things like that. The machine is very capable. What we've basically put out is that's what we've verified. That's what we have tested. Of course, Emily and I, it's our job to now quickly expand on that. Now, what is the next row spacing? What is the next crop that we need to do? I'm sure if you've thought about cameras going through your field, you've thought of dozens of different things, and you've told us dozens of different things you can work on. We're working on most of it. Yes, that's our starting point. We'll go from there.

Tony: To follow up on that question, I know there's a handful of people in this crowd that different crop types, different row spacings, what does that timeline look like? What does it take to adopt or to release a new row spacing or a new crop type?

Emily: Thinking about releasing a crop type, for instance, so the models that we built, some of you might have been participants in our data collection efforts where we actually had people out wearing, we call them DCMs, data collection modules, they're backpacks with cameras on them, collecting tons of images throughout a lot of fields. Diversity is really important for us. We're taking those images, we're getting them labeled, we're saying what's crop, what's weed, what's ground, and then we're feeding that into our models.

Our models get better with the more imagery that they have and the more conditions that they're seeing; soil types, crop varieties, weed types. The more data we get, the faster we are at building those models and the better those models will be in new situations that they haven't necessarily been trained on. They'll be generalized to work well in those situations as well. With the first three large crops, it's a year or two of data collection and testing and refining those models to get where we wanted to get to where it's equivalent to broadcast or better within your field.

I think as we continue to collect more data, we're getting better at building those models, and we're getting faster, and we're building a bank of imagery that we can use to more quickly enter new crops, new regions, et cetera.

Tony: I know a lot of people probably ask the questions of when is this crop coming, when is this crop coming, but obviously you guys have a process in place, and it takes time. Full well knowing that you guys are working on more crop spacings and more crop types. I want to go back a little bit. We talked about the weed size. You talk about thumbnail. This question, I want to go back to Jared and Nick, is every field a good candidate for See & Spray? How did you guys navigate that question yourselves when you were utilizing the machine in 2024?

Nick: I can say for 2024, that was a difficult decision. I shouldn't say difficult. There's more question behind it. 2025, the way the pricing, the way it's set up, you just utilize it. It's very, very simple. I'm sure we'll get into there somewhere where they'll talk about how it works. 2025, either you're spraying or you're utilizing the camera. Last year where you paid for the whole field, no matter how much of it you sprayed, if you had See & Spray on, you paid for that whole 80 if you were spraying an 80.

For us, as a business, that became a difficult guess of how much we'll spray. We shot for 35%, and it worked out luckily. Overall, there was fields that were 11, there was fields that were 50%. There'd be some, in a fallow situation, like wheat stubble, Kansas, a lot of wheat. Volunteer wheat comes in after a rain, the field is covered. You could tell pretty quick that wasn't a situation, but there was a difficult gap when you got in that 50% range, but now it's for all of it.

Tony: Jared.

Jared: I would echo what Nick said there. Last year, it was challenging to determine where you wanted to turn it on and off, but this year it's pretty simple, especially when you're in the fallow setting at a dollar an acre or whatever it costs. Doesn't take long to pay for your See & Spray savings there. In crop, that's going to be something, after you do a few fields and see what the cameras are capable of picking up on and everything there, after a while, you have a gut feeling once you pull on a field. I would say most fields are a good candidate, but there are a few where you're just going to turn the broadcast spraying on and go.

Tony: Let's roll that into a question on the use-based license models. Katy and Mark, let's talk a little bit about what that license model looks like, what a customer will pay for, what they won't pay for when utilizing the See & Spray technology.

Mark: For 2025, I want to commend the folks at John Deere, they made a great decision in my mind and simplified this a lot. You'll pay $5 per acre in corn and $5 per acre in soybeans when the technology is turned on, and it's not spraying. You're not spraying; you're not paying. It makes it really easy for-- Nick, he always said, especially more so on the commercial side, where there was challenges knowing when you go in a field, should I spray this whole thing, should I disable the technology? Assuming your mix that you're spraying is $5 per acre or more, you would want to leave the technology turned on.

Katy: It was really cool to see what those savings looked like for our customers in 2024 because even with the model, how it was in '24, where you were getting charged for every acre, it's still penciled out to make a lot of sense. Now with this only pay when you don't spray model, you don't have to sit and think about if you should turn it on or off. You can just turn it on and go.

There's also really cool support things. You can set your system into, it will go into fallback mode if you get into a situation where it's too dusty or your boom is too high or your operator is going too fast. It'll give you that alert where you can go all into total broadcast mode or shut off. Now you don't have to worry about getting charged for those acres when you are getting into broadcast mode because it's not going to charge you.

Tony: Now I want to go back to Jared and Nick. Use-based model, obviously, you guys bought the technology in 2024. You were going to use it regardless. What were your first thoughts, your initial thoughts getting into it and having to pay per acre license fees? What was going through your mind with that?

Nick: Just knowing, I guess, as a business, we invested in this, it has to pay, we're cooperative. We didn't want to have issues later. We didn't. That was probably the biggest. When we first started, you're going to do a little bit, and then we go out, and be like, "All right. We're in two fields. How many days do we need to wait to make sure it works?" Across the territory like that, after a month, we felt definitely confident.

Tony: You mentioned nozzles, Jared. Talking about the specifics. I have a sprayer on my farm. I want to get into See & Spray, the premium precision upgrade kit. I already have nozzles on the sprayer. Can I just utilize those nozzles going forward? Katy and Mark, I guess that one is directed to you guys.

Katy: My biggest disclaimer, I'll put out there for everyone in the room, I'm not an agronomist. I will never put my neck on the line to tell you what to nozzle. I'm married to a farmer. I'm learning, but we will always lean on what our growers are experiencing and what's working best for them. They're the biggest wealth of knowledge. They'll hopefully be consulting their agronomist. John Deere has a great nozzling guide. That's going to be how we guide people this year. There'll be nozzle charts and things like that available. We have a wealth of resources that everyone can consult from different places.

Mark: Maybe to add to that. As we were developing the system, as you can imagine, we're going through the field at 12 or 15 miles per hour. That's 18, 20 feet per second. That all of these cameras are trying to facial recognize every leaf, make a decision on is it crop or a weed, send the signal to launch the spray, and then make sure that the spray hits the leaf before it's out of bounds from us.

We spent a lot of time working with various nozzles, characterizing how long it takes to get up to a full fan, to get up to the full dosage rate, to fly the particles of liquid through the air until it gets to the plant. There is a guide that we put together that lists out the different nozzles that we've characterized that we know work well. Those range from fine particle size up to the ultra low drift.

There's a couple of new nozzles that we worked with our nozzle supplier of choice, and they developed a couple of target spray low drift nozzles for us that we feel very confident in. It's all of our data is wrapped around those. John Deere is very careful, like Katy mentioned, to not tell you how to farm. We know farming. We know farmers. We've done everything we can to do the testing. We've put out what we've learned, what we recommend. Then, yes, go from there, and we'll learn with you. There's a lot to the nozzles, for sure.

Nick: It's not, I don't want to say difficult. You go through that guide. With having custom applicators, other guys running the rig, we're using a room that have applicators. Brand and styles of nozzles is like brands of pickups and tools. They are diehard P-Jet, AIXR. That's all I'm going to use. What did that nozzle make for droplets for what you're doing? You needed that droplet size for these herbicides. You can go through there and say, "Now I'm going to end up using a ULD, or the ultra-low drift max." All the droplet sizes that are needed are there. It's just a different style. Really what it comes down to is the droplet size, and they're out there.

Mark: We also did a lot with spray angle. The computers on board know what nozzles you have on there. They're taking all of those aspects into account for the timing of spraying, they turn on and off, but it also affects how many nozzles get turned on. We would like to gravitate toward more 60 and 80 degree nozzles, especially on a 15-inch space machine because it maximizes the uniform coverage, and it eliminates extra nozzles from having to be turned on.

All those extra nozzles with the wide sprays, we're only turning them on to get that little slice that would come over to the weed zone. Certainly there's recommendations that are out there. The savings is calculated on how many nozzles we turn on and off. If it's a wide pattern, we're going to turn on enough nozzles to make sure that we're not creating resistant weeds.

Tony: Let's dive into that a little bit deeper. You talk about the savings and the settings of the machine. First let's start with Katy or Mark from the dealer's side, and then we'll go on to Nick and Jared from the grower, the user's side. What was the experience or what did we learn about savings with See & Spray when it comes to adjusting settings or the types of nozzles we're using? What did we see out there for savings in 2024?

Nick: For us as co-operative, having the three, just because we ran these numbers before we came up here, we save our customers $112,000 across 42,000 acres. That basically shoots to about $3 an acre with the fees or subscriptions of last year. That number will definitely increase this year on savings.

Katy: As far as our users went, we're doing our best to build out guidance for best practices going forward. We're so grateful for our early adopters, those growers who helped us learn so much, but we're coming from an angle of you can change those sensitivity settings so that you can get greater savings or less, but we like to set them right about the middle. Then our farmers have gone out and seen-- They're making sure they get that weed kill.

They could really tighten those margins down even more. I know we have growers in the Northwest who are up at 96%, 98% savings because they have gotten so good at those sensitivity settings. It's a learning process, but we are able to capitalize off the things that you guys have learned to help you guys moving forward.

Tony: Jared, do you have anything to add?

Jared: Yes. When we started, I might've been a little more skeptical on this, so I headed at the most aggressive settings you could; so spraying the most it could. We were at about a 50-50 savings. Then as I got familiar more with the technology and the savings and stuff like that, I throttled it back to the default factory settings. It was about a 60% savings. I think I got as high as 63% once though. Just the default settings alone are going to save you a lot. There's not going to be any skips out there. The technology and the cameras on that machine are very good.

Mark: There's tools you can use really built into the machine. If you drive it out into the field and you walk out in front and you say, "There's a small weed, or there's a small weed that's in the row," lay your pocket knife next to it, lay something next to it, drive the machine up, turn that camera on the display, and you'll be able to see, of course, your pocket knife will be glowing. It's going to tell you, with color coordinating, if it's green, it's a good plant. If it's purple, it's a weed.

If it's not showing the weed that you want to kill, you can right there change the sensitivities and see what those do. To their point, the medium setting is the right setting by far the majority of the times. There should be a really good reason why you're leaving that medium setting, that number three. I think in the future, as we get better and better, there probably won't even be this setting. For sure, we wanted some flexibility in the system.

Tony: Lee, maybe you can dive in a little bit deeper. There's a couple of different settings on that. The levels of settings, whether it's the high sensitivity or low sensitivity, and how that makes adjustment. Do you spray more, do you spray less, or what those settings are?

Lee: There's a one to five setting, five being ultra sensitive. If you have any doubt, just spray down to one. Hey, be super certain that it's a weed before you spray that kind of range. It's basically confidence built into the models. You think about See & Spray is like this facial recognition program. You see somebody walking at you hundreds of yards away, the first thing you're able to tell is if it's an adult or a child. Then they get closer, and you start to pick up on other cues. It's like that.

The confidence of the model, it's using color, but it's not using color very much. It's using the shape of the leaf, the growth pattern of the plant, the reflections of different light sources we can't even see ourselves. It's doing all of these in milliseconds. It's classifying, hey, I've seen that before. That's water hemp next to a soybean, whatever. It's that confidence level. There are times where the leaf is partially behind a good leaf. It's seeing a portion sticking out, and it's saying, "I'm pretty sure that's a weed, but I'm not completely certain." That's really where the sensitivity, but three, that middle setting, it will be very good.

Emily: Then there's one other setting that I think directly correlated to the savings that you're going to see is those buffers. What we recommend is running in small, we're pretty confident you're going to hit that weed. As the wind starts to pick up, that's when we start adjusting those buffers either laterally or longitudinally to make sure you're hitting that weed, and it's in the center of the spray development. Typically run at a small, that's where you're going to get your most savings. As the wind picks up, and you want to make sure you're hitting it, start extending those.

Lee: On the engineering side, we put blue pond dye in the first tank and we go out and spray areas. We go look to see if the non-dyed areas, if we can find a weed, or if in the blue, and several of us have been out in the field. There'll be sometimes eight people looking at a blue dye patch, and no one can find a weed. Then finally someone sees this quarter-inch weed at the center, that it's sprayed. That's the kind of things that make you say, "This thing knows what it's doing."

Tony: Yes. Obviously, a lot of different settings that can come into play when we talk about the savings, whether it's the cost of our chemistry, the tank mixes that we're running, or it's the sensitivities being adjusted on the machine. All that's going to play a factor in the savings that we're seeing. Mark and Katy and, Lee and Emily, feel free to chime in here too. Models of machines, how can I get this type of technology? Do I have to go to the dealership and order a brand new sprayer in order to get this stuff?

Mark: Good news. You got a 2018 or newer machine with exact apply. You could have installed a bucket, installed a-- you'll need a SF-RTK receiver boom track pro to a Gen 5 monitor. That's all stuff that your account manager can bundle into the package. In 2018 or newer machine, an exact player of the two big requirements.

Tony: Is there anything we can't get in a precision upgrade 2018 or newer sprayer?

Lee: Of course, we focused on the heart of the market, the 120-foot boom, there's pockets being developed for other models. I know the 100 foot is on the way. We'll tick those off, as we can develop them. The pay as you go type, the really strong part of that for us is it encourages us to keep these older machines, developing new technology for them, keep them in the field, spraying acres versus a sell it and then forget it and go on.

How do we sell the next new machine? From an engineering standpoint, we talk daily about what's the fleet size for this model. We need to bring new features to that model. As we're developing for the model year '26, '27, '28, this has really caused an engineering communication change of let's always be thinking about that fleet of machines that's in the field. How do we keep getting those better?

There's a bunch of machines out here that are going to be running for the first time. As we're developing that next feature, we're thinking about, that architecture that's in the field, can that support it? Things like, 2018 wasn't just randomly picked. It's an electrical architecture on that machine that was capable of doing it. The engineering side, we had to do some things to make that machine capable. You step back, the architecture before that was a 2014 to 2017.

It just became, all right, so there's a lot more controllers we would have to update on the machine. Really going forward, this model, it really helps us focus on the machines that are in the field and keep getting those better.

Katy: One of the things that I really come to appreciate is this relationship between RDO and John Deere. We really get to communicate what our customers need, and whether it's fixed right away or not. There's that process, we're aware of that. What crops do you guys need to see? What spacing do you need to see? What boom length do you need to see? We do feel, for the most part, pretty heard on that. That allows us to get you guys better solutions more quickly. That's a perk.

Tony: Looking forward into 2025, so Jared and Nick, you got to run the machine for one season. Now we're coming up on the running it for the second season. What are some of the changes you guys saw happen in 2024 that you're going to make in 2025 to make your See & Spray experience even more successful than what it was?

Nick: We definitely did some learning sometimes with hard knocks. Do not underestimate the power residual of Dicamba and Banvel products. Where we are running with the ultimates, let's say for a corn post application, we're going to run half the Dicamba in your, some would call residual tank, the tank that sprays all the time, and the other half in the post See & Spray tank. We definitely seen some effects to running a high rated Dicamba on corn, and then it not raining or just a little bit of rain.

That Dicamba was a residual for that small window before the main, I guess, you would say, residual was activated. We definitely had a learning course there, but that's really about it.

Jared: I'll use See & Spray on every acre pre-emerge. It's, like Mark said, with the new pricing structure, that's a no brainer. By doing that, to comment or to piggyback on Nick, you can up your chemical rates a little bit higher there for a more effective weed control, especially in kosher and crops like that, where we're starting to see some resistance. That's one thing I'm definitely going to build upon in 2025, better sprayer tip selection. Now that all of us are educated a little bit more on that, that's another one for 2025 to learn upon. Those are the two main items, I suppose.

Tony: I want to go to Emily and Lee. What does the future of See & Spray look like?

Lee: It's pretty good. These guys are big fans of the technology. The way it sits today, I jokingly will say it's never going to be this bad again. We have some awesome things in the works. What is in the works? You've all thought through maybe the shortcomings of where we are with our first viable product. More row spacings, more crops, more insights. While these cameras are going through the field, what else would you want to know?

You've got this 120-foot wide scout going through your fields multiple times. We're very focused on-- Let's not forget about those insights. Maybe you want to get insights on how well your planting did, the population you chose. Those are certainly things. We have such a long list of things that we're prioritizing and trying to bring to market. What's exciting is so many of them are software things. The great thing about the software is we can get those to go on machines that have already been sold or, include them in pockets for older machines.

Emily: Yes. I would say just think about your production system. We want the sprayer to be able to run in every pass and either save inputs or improve outcomes. That's where we want to get. I think another thing is, this is why we love to come to events like this and come talk to farmers. I think all of our best ideas have come from you guys. We want to hear what your pain points are. We want to hear what excites you. Yes, after this, come find us and let us know.

Tony: Jared and Nick, then what are some of the things you guys, as growers, as users, would like to see come to See & Spray in the future?

Jared: I would like to be able to use the technology more as a pre-harvest burndown aid. If it's sensing green in your crops, the spray there, there's a lot of times hilltops and side hills and stuff like that. You're applying a glyphosate, and whatever else your tank mix may be, in areas where it doesn't need to be applied. We could see a major savings there on our farm. That's one thing I'd like to see moving forward.

Tony: Katy and Mark, what about you guys from the dealership side?

Mark: I don't know. I'm easily mesmerized by the weed maps. I could study these See & Spray weed maps for hours. It's neat now that we have 2024 behind us. It was fun to overlay a weed map with a yield map and see, "Hey, did that weed pressure translate into a lower income on that specific piece of ground or that spot in the field?" Is there anything I can do from a management standpoint? Is it a spot that can be drained in the field to reduce weed pressure?

Again, I thought that was neat. For me, being the weed nerd here, apparently, I would like to see more identification of weeds and some segregation, so to speak, on what type of weed it is.

Katy: Yes, and just piggybacking on what Mark said, being able to maybe identify how a field should be mixed because since we are spot spraying, we don't know exactly what to mix. Then the classic that I know a lot of our customers, I hear all the time, spraying in low light, especially in our fallow users, that's a pretty big one for us. What does the process for that look like for development?

Mark: To go back to what I said earlier, you eat the elephant one bite at a time. We're certainly looking at most, if not all, of those areas. Part of this, and because it's software that can get out to the masses pretty quickly, pretty easily, as an engineering team, as a product management team at John Deere, we wrestle with how perfect should it be before it gets to your farm. With this first release, we put it through all of the rigors.

When a startup company comes out with a spot spraying solution, they just need revenue. They need to survive, and they're super scrappy. We're John Deere, we've got this brand, you can tear the brand down in a heartbeat by screwing up. That drives us to test to be super rigorous in verifying that this works, that it does exactly what we say it does. We'd like to get to the point where we are more scrappy, and that we do get--

If customers are screaming, saying, "I just want to variable-rate my desiccant on these crops based on color," that we can develop that new tool, get it out there. Maybe it's a beta release. Maybe there's some sort of program to that. The most frustrating part to us is how long it takes to verify in all these different regions and all these different conditions. Because when we put it out there, you all have this expectation.

We are trying to figure out how can we be more scrappy, get these features out because we could roll it out with a software update in a heartbeat and get it to you. It's really throttled by the expectation of it's got to be right, it's got to be perfect. We're trying to balance that and get new features out quicker.

Tony: Now, we are going to open it up to the floor here. Don't everyone raise it at once now. There's wheels turning.

Participant 1: How do you guys know how much chemical to mix up when you're going out to spray, if you're only spraying the weeds that are out there and not waste that chemical that's in the tank, obviously, as a co-op, and then also as a grower?

Nick: As a co-op, one thing we definitely seen more, after we built the central dispatch, which helps a lot, is running the same tank mixes. For us as a company, we're going to have like a Bear mix and a BASF mix that are our priority go-tos. As of now, we have been keeping the See & Spray rigs on those. There's one-off mixes. We try to utilize that through another rig. Then just trying to do as much as you can. Day's work, it lines out. It has come to the point where you get to the last field, and you just run it out conventionally.

Mark: We do the same thing. That is one nice thing about the ultimate sprayer with the two-tank configuration. You can cheat on your sizes there. The last field, you just try to get it as close as you can without having too much. Then if there's 10 acres left, we just broadcast it to be done with.

Lee: We do hear a little bit from those growers that have been doing it for a couple of years that they do get much better at that predicting, but we're also working on features that would maybe allow you to get-- The system would be able to tell you, all right, now is the best time to turn to broadcast, and we can empty this tank, if it's the last part of the job. Maybe it's go back and spray the inverse of what you sprayed on the outside round, or something like that, something to make those more usable. Then also just technologies of, how do we get rid of that problem altogether? We're working on some solutions in that area as well.

Tony: That is definitely one of the burning questions that we get on the dealership side. It's really awesome to hear that you guys are working on tools to help the customers with that.

Participant 2: First of all, thank you for your time. I know everybody is busy at these days. I appreciate your time and, of course, your insights. I'm a student at NDSU, as well working on a target spring robot. It's more of a technical engineer related question. I would say for technology right now, it's advancing very quickly. My question is what are the limitations, technological limitation, for today's in order to get more speed wise on detection, as well as going on the road? Would it be like more into software or maybe into hardware, such as computations, AI?

Lee: You're talking about spray speed?

Participant 2: Yes. Spray speed. Yes. Detection.

Lee: There's a lot of things in play that set that speed, the angles of the camera. You could look out farther, but then you're looking more at a skewed angle. You can see a little bit less of-- They're at an angle where you try to get out as far as you can, but you're also getting a top-down view on that. That sets the distance. The computation speed, we're talking about milliseconds. The time that it takes for us to see something-- we try to see it from--

Every plant on the ground isn't just being looked at by one camera. It can be seen by at least two or three cameras at the same time. We break down the time it takes to see it to when it will be gone. We need to start spraying before we get to the weeds so that we can get up to full rate. We have that broken down by millisecond. As new graphics processing power comes out, that gets faster. The offset to that is we want to know more about what we see in the future.

Right now, as you go through the field, if you set it to corn, it's basically saying, "Are you corn? Yes or no." If the answer is no, we spray. In the future, we want to do more than that. You don't want to be going through a soybean field, spraying a broadleaf herbicide, and spray all your volunteer corn. That's two questions. Are you a soybean? No. Are you corn? Yes. Then spray. The computational power will definitely also help the amount of things.

Do you have insect damage? Do you have nutrient deficiencies? All of those things, all of those take compute time. Yes, technical wise, there's a lot of parameters that we're battling with.

Emily: One other thing that we talked about a lot earlier with the nozzle tips, so there's the physics of spraying. You have to start that spray. It has to reach the reach the weed, and it has to be at that right pattern by the time it reaches the weed. That's a big part of it as well.

Participant 3: Hi, this is more directed to the John Deere engineers. Along with See & Spray two of the other technologies I see talked a lot about in weed control is application via drones and also using lasers to eradicate weeds. With the image classifier model that's built for See & Spray, does John Deere have a vision for the future with applying it to different vehicles like drones for fields that have more adverse terrain and greater side hills where spraying with a large row crop sprayer is more difficult, or applications like lasers where there's areas that people are concerned about runoff or organic farming practices?

Lee: Those are two very different categories. On the drones, drones are interesting. They have gone from something that's the maybe a novelty to very capable both camera wise, payload wise. They're getting more capable. There aren't herbicides yet that have labels for aerial applications. Right now they're being used for insecticides, fungicides, things like that. For herbicides, that has to go through the whole label change process. Maybe that's further out there.

They do make good scouts. Part of that challenge though has been the amount of turnaround time for it to go out and scout your field, do the compute, get back to you. Then, maybe it rained in that amount of time, and so the data is going to be different by the time your sprayer can get out there to do it. On the drone side, yes, that's an area that we monitor quite closely. A number of our dealers are drone dealers and developing expertise in that area.

Certainly something that we're not ignoring or just assuming it doesn't have an effect on us. The other ones of robotically electrocuting weeds and things like that, we're aware. We go to the farm shows and nerd out a little as we figure out what they're doing. Some of that seems to be pretty niche. Productivity is something that-- When we dabble in something that's specialized, a lot of times it comes up to just productivity. You guys want to cover hundreds of acres a day. Maybe that's some of the limitation both to the drones and to the different ways of killing weed. We try to stay as connected to those things as we can, and monitor. We've got folks that focus on making sure we're up to speed.

Tony: I think that's a great question to end this on. Thank you guys very much. Please take a moment to subscribe to this podcast if you haven't already. You can subscribe to this show on the many different podcasting apps that we're streaming this out to, such as Apple, Google, Spotify, as well as many others. While you're out there, drop us a review. We'd love to hear what you think about the show. Lastly, make sure to follow RDO Equipment Company on Facebook, Instagram, and X, and also catch our latest videos on YouTube. You can also follow me on X @RDOTonyK.