Trascrizione completa di Episode 12 - Precision Ag Research in Pecan

00:03:04.479 --> 00:03:13.360 I mean they need to fall, they need to drop, I mean they need to dry a little bit, a couple days, but but we talking about how quickly are we getting them up? 00:03:13.520 --> 00:03:15.680 Is it anything different, you know? 00:03:16.319 --> 00:03:20.800 Oh yeah, I mean with any variety, it's depends somewhat somewhat on the weather you've got. 00:03:20.879 --> 00:03:26.719 Uh right now, it seems that most things that I'm seeing that have been shook, they're coming off pretty dry. 00:03:26.960 --> 00:03:31.919 And uh you know, so you don't have to leave them on there on the ground too long right now. 00:03:32.080 --> 00:03:36.560 That may change as we start to get later into the harvest and get more rain, that kind of thing. 00:03:36.719 --> 00:03:46.400 But with Lakota, you know, the biggest knock on Lakota is that it's you know, the meat turns dark if it if it uh if you wait too long to harvest. 00:03:46.639 --> 00:03:47.520 So that's one thing. 00:03:47.599 --> 00:03:50.000 So you don't want to leave them laying on the ground too long. 00:03:50.159 --> 00:03:59.680 And it's early enough in the harvest season that prices are are usually pretty good at the time that that pawnees or or the Lakota start to get ready. 00:04:00.080 --> 00:04:05.120 So I think it's probably worth going out there and and getting Lakota. 00:04:05.199 --> 00:04:09.759 You don't need to leave them laying out there too long because it will turn dark, that'll that'll hurt you on price. 00:04:10.159 --> 00:04:15.759 Um so you wanna you wanna get them up, you don't want to drag your feet too much on Lakota. 00:04:16.720 --> 00:04:20.160 The last question on all this to me is is kind of market related. 00:04:20.639 --> 00:04:36.560 But let's say you unless you got like a custom sheller and selling into the meat market, unless of that situation, does it matter you know, if we wait on everything to come in to get that single shake? 00:04:37.040 --> 00:04:40.800 In other words, yeah, does it not need to rush into the early and shell markets what I'm getting? 00:04:41.199 --> 00:04:44.079 It it depends on what you what you're growing, what you have out there. 00:04:44.240 --> 00:05:01.040 I mean, if you have something that's gonna go on the gift pack market mostly, so something like Pawnee, uh Caddo, uh probably Creek to some extent, Oconey, um you know, any of that, that's really good gift pack stuff. 00:05:01.199 --> 00:05:16.800 Uh it's your higher quality uh, you know, kernels that you probably want to go ahead and get that, even if you have to shake twice to get them all, because um the peak price is going to be you know that gift pack market. 00:05:16.879 --> 00:05:23.120 And once that gift pack market's full, you know, that's tends to be when we see prices start to drop a little bit. 00:05:23.920 --> 00:05:28.879 Um but you've also got a weigh, you know, what kind of crop you got out there. 00:05:29.040 --> 00:05:51.279 I know these high diesel fuel prices are are making a lot of people think about shaking just once, and there's a lot of situations that's gonna be the thing to do because especially for some varieties, possibly true for all most varieties, but um certainly for things like Stuart, Sly, Moneymaker, you know, desirable probably to some extent. 00:05:52.000 --> 00:06:01.360 It's probably gonna pay to wait and just get those once because you're not gonna make up for that fuel cost on the prices you get for those nuts, probably. 00:06:01.519 --> 00:06:09.759 But um, so yeah, you just gotta kind of weigh what you've got out there, how much of it you have, um and how much you got in it. 00:06:10.399 --> 00:06:13.759 You know, we're here now getting ready for harvest of 2026. 00:06:15.199 --> 00:06:16.639 Yeah, harvest is here. 00:06:17.279 --> 00:06:27.519 And in some attempt to segue, I'm still talking to people in East Georgia, they talk about, you know, hey, I've I've just twit switched over some of my fungicides and you know, I like this program, I like that program. 00:06:27.680 --> 00:06:32.560 However, I'm hearing that West Georgia seemed to have much more scab. 00:06:33.040 --> 00:06:47.600 And in the conversation, I was like, you know, we had such low pressure for scab this year that yeah, I wouldn't say that much of our changes of of fungicide couldn't be compared to the difference because of we had no rain. 00:06:47.759 --> 00:06:50.480 We were so dry and all the pawnee over here looked good. 00:06:51.199 --> 00:06:55.199 Some growers are starting to spray with a drone. 00:06:55.759 --> 00:06:57.279 And this is my segue. 00:06:57.680 --> 00:07:02.480 Um uh we got today a special guest, Dr. 00:07:02.639 --> 00:07:10.639 Luan Oliveira, and uh he is a specialty for working with drones and these uh specialized equipment. 00:07:11.040 --> 00:07:23.680 And one of uh my people over in East Georgia, you know, they they tested their drone on about 25 foot tall trees, had a windy day, couldn't get a lot of the spray to go through. 00:07:23.839 --> 00:07:27.920 They said, well, we were a little nervous, but they went on spray through the year. 00:07:28.160 --> 00:07:31.519 But at the end of the season, they said, Scab looks good. 00:07:31.759 --> 00:07:34.079 Again, we've had such a dry year. 00:07:34.240 --> 00:07:37.600 It may be hard to say how much it worked this year. 00:07:37.839 --> 00:07:42.800 So in that segue, that's my attempt to segue, I want to introduce Dr. 00:07:42.959 --> 00:07:50.959 Luan Oliveira from um I might better let you say it because I won't say the You're fine. 00:07:53.360 --> 00:07:54.720 Yeah, yeah, yeah. 00:07:54.800 --> 00:07:55.680 You're on. 00:07:56.160 --> 00:08:00.000 So uh thanks for hosting, thanks for inviting uh Lenny and Andrew. 00:08:00.160 --> 00:08:01.600 I'm Luan Oliveira. 00:08:01.839 --> 00:08:05.360 I'm a precision ag extension specialist at UGA. 00:08:05.759 --> 00:08:08.560 So I'm with the Department of Horticulture.

00:08:09.040 --> 00:08:20.720 And uh by saying that, but all my precision ag studies are focused on specialty crops, and one of the number ones are the pecans, right? 00:08:21.439 --> 00:08:25.920 So um since I started back in 23, Lenny brought this to me. 00:08:26.079 --> 00:08:30.399 He said, Hey, uh growers are starting to look into spraying technologies. 00:08:31.120 --> 00:08:41.679 And um first one was Smart Apply, we can touch on that, and the second one uh was uh drones and also some robotics. 00:08:42.159 --> 00:08:54.960 So um I'll be open to any questions you might have, Andrew, or I can just bring in all these studies we've done so far on the Pecan side for spray drones and smart apply as well. 00:08:55.840 --> 00:09:04.960 Hey uh Luan, before we go into all that, can you kind of give us a little bit of your background, where you're from, you know, where you school, what what you're doing? 00:09:05.039 --> 00:09:05.519 Yeah, yeah. 00:09:05.759 --> 00:09:18.000 So um I'm originally from Brazil, and um there we do have a degree is called agronomic engineering, so which is equivalent to agricultural engineering here. 00:09:18.159 --> 00:09:26.879 Uh with we have more of the plant physiology side, so that's my bachelor's, and then I went to school there as well, graduate school. 00:09:26.960 --> 00:09:32.399 So my master's and PhD were in agronomy, but focused on agricultural engineering. 00:09:32.639 --> 00:09:53.840 Back in the days, I was uh working on sugarcane harvesting for my master's, and for my PhD, I've done, I actually came uh went from Brazil to Alabama, so spent two years there as a uh research scholar working on precision planting technologies. 00:09:54.159 --> 00:10:00.960 So I shifted from sugarcane to soybeans, uh, corn and cotton in Alabama. 00:10:01.200 --> 00:10:16.080 And then um after that, after finishing my PhD, I went to Nebraska uh to work as a county extension agent for about two years and a half as well, before joining UGA in 23 as an assistant professor. 00:10:16.320 --> 00:10:22.159 So that's my I'll say trajectory from bachelor's until I got here. 00:10:22.320 --> 00:10:29.519 So I was uh used to um broke crops a lot, like when I was in Nebraska just working on soybeans and corn. 00:10:29.679 --> 00:10:35.840 Again, everything on precision ag, precision planting technologies, and uh drones. 00:10:36.480 --> 00:10:46.240 So that's uh all the I would say um uh knowledge I brought with me to apply on the specialty crops here in Georgia. 00:10:47.440 --> 00:10:48.080 Great. 00:10:49.279 --> 00:11:02.080 Um so yeah, you mentioned you work on all specialty crops, so so that includes like all the vegetables, uh blueberries, I guess peaches, pecans, a little bit of all of it. 00:11:02.159 --> 00:11:04.639 So you're a you're a pretty busy fella. 00:11:06.960 --> 00:11:07.919 Try to, right? 00:11:10.720 --> 00:11:14.480 So uh all the vegetables as well, onions but peppers, tomatoes. 00:11:14.639 --> 00:11:26.720 We've

done some studies on carrots, and all of them on not on the pecan side, but on the blueberries, onions, tomatoes, and peppers. 00:11:26.960 --> 00:11:32.480 They're more instead of like just spraying technologies, they're focused on labor savings technologies. 00:11:32.639 --> 00:11:43.279 So that's when we bring in the robots, that's when we bring in um spot spraying technologies to the field and evaluating them in Georgia grounds. 00:11:44.080 --> 00:11:51.360 You mentioned earlier, you know, one of the things you've worked on with pecans is a smartest smart apply spray system. 00:11:51.919 --> 00:11:56.320 And for those who don't know, uh I guess we'll let you describe that. 00:11:56.399 --> 00:12:00.320 You can describe it better than me a little but in just a second. 00:12:00.480 --> 00:12:05.120 But we've seen pretty good adoption of that by a lot of growers, it seems. 00:12:05.519 --> 00:12:11.679 And uh I won't go too much into what they've been seeing because you can tell us a lot a lot better than I can. 00:12:11.759 --> 00:12:22.559 But um just tell us a little bit about you know what that smarterply system is exactly and how it works, and then some of the results you're seeing with the work that you've done with it. 00:12:23.120 --> 00:12:23.840 That's great. 00:12:24.000 --> 00:12:35.360 Um, so a smart apply is if I just describe it how it is right now, is a AI-driven LIDAR-based spot screen system. 00:12:35.919 --> 00:12:48.080 Um, by saying that, like when I say AI, we're bringing to artificial intelligence, which is uh the capacity of a computer to mimic human intelligence. 00:12:48.320 --> 00:12:51.440 That's that's what AI uh stands for. 00:12:51.759 --> 00:13:04.559 Um imagine that you are on a uh you're in a tractor and you're spraying your field, and you just want to spray the trees, you like the the canopy, not the trees by itself, right? 00:13:04.639 --> 00:13:06.159 You want to spray the canopy. 00:13:06.399 --> 00:13:15.759 And with your eyes, you can tell where that's that branch starts and where that canopy ends after that first branch started. 00:13:16.159 --> 00:13:20.720 And uh what we currently have are I would say hydraulic actuators, right? 00:13:20.799 --> 00:13:36.320 You come in, you turn off uh that uh uh you turn off that hose where the chemical is gone, the pump is still on, uh, or you'll turn off the hole sprayer when you are in in that gap in between the trees, so you don't spray a whole tree. 00:13:36.799 --> 00:13:52.799 When you bring in AI and uh a lighter, which uh with which stands for light detection and ranging, it's basically a laser shooting 180 degrees, um looking for that canopy. 00:13:53.120 --> 00:13:58.720 So it's it's it's going back and forth, so it's it's sending and receiving all the time. 00:13:58.879 --> 00:14:13.039 So when that laser sees that little branch, the AI, which is mimicking human intelligence, say, oops, there's a branch right there, and the nozzles are already filled with

chemicals just waiting to be open. 00:14:13.279 --> 00:14:18.240 So though those are hydraulic, electrohydraulic actuators, uh solenoids. 00:14:18.399 --> 00:14:24.720 So those solenoids they open and then that chemical goes straight where it should be going, right? 00:14:24.879 --> 00:14:34.799 Without me trying to take that decision on closing or uh opening uh manually the that chemical line. 00:14:35.360 --> 00:14:39.039 So if if that that's it, it's it's it's not that complicated. 00:14:39.200 --> 00:14:47.039 Basically, AI is taking the decision on opening and closing that nozzle based on that sensor, which is a laser. 00:14:48.399 --> 00:14:48.879 Yeah. 00:14:49.120 --> 00:14:56.399 So what tell us what what kind of results you've been seeing on I mean any trials you've had with that or working with growers using that? 00:14:56.720 --> 00:15:02.159 Yeah, so um I've like we've done a trial last year on it. 00:15:02.320 --> 00:15:05.279 This year we focused more on the drone side. 00:15:05.519 --> 00:15:21.600 Uh, we wanted to understand how much is the chemical savings if you if you have like the I'm not gonna say a worst-case scenario, but I would say like uh an older orchard, which is the most variable scenario scenario you have, right? 00:15:21.919 --> 00:15:24.159 Yes, different branches everywhere. 00:15:24.559 --> 00:15:31.759 Um, so what different size trees, yeah, yeah, different size trees, yeah, big gaps, exactly 40 by 40. 00:15:31.919 --> 00:15:35.279 Um, so what we've done, we installed a flow meter. 00:15:35.360 --> 00:15:37.759 That's that's that's the the that was the first step. 00:15:37.840 --> 00:15:55.039 We installed a flow meter on uh that chemical line, both sides of the sprayer, and we wanted to know how much was the chemical savings using Smart Apply versus a conventional broadcast application at 75 gallons per acre. 00:15:56.000 --> 00:15:58.480 So same uh number of trees. 00:15:58.559 --> 00:16:02.399 We had seven trees per plot, uh, four applications. 00:16:02.559 --> 00:16:09.840 So that was about uh 28 trees sprayed um multiple times during the season. 00:16:10.000 --> 00:16:39.279 So all of those 13 uh uh spray activities we've done, and then we compiled all of that chemical uh numbers, like uh chemical uh I would say uh expenditure uh number, and then uh at the end of the season we compiled everything to know like how much smart apply spent and how much the conventional also done, and uh we came back with an average of 60 percent of chemical savings. 00:16:39.840 --> 00:16:42.559 That's a lot of money, yeah. 00:16:42.720 --> 00:16:52.720 So chemical savings was not the only um uh I'll say factor there, the variability was also a factor playing, right? 00:16:52.879 --> 00:17:05.279 Um when your going like is is balancing a lot and it's shaking a lot, so uh sometimes you have air gaps on those hoses. 00:17:05.519 --> 00:17:10.880 So the variability of the spray output was also higher on the conventional. 00:17:11.279 --> 00:17:17.839 And on pecan sprayers, uh, I just had a pecan grower, I asked him this question last week, literally. 00:17:18.079 --> 00:17:21.359 Said, hey, um, how much is how big is your tank? 00:17:21.519 --> 00:17:26.400 And um how much can you spray uh basically with that tank? 00:17:26.559 --> 00:17:40.240 And then just doing that quick math uh of of a um a hundred uh a thousand gallon tank, if you're spraying hundred, uh if you're spraying 10 acres, that's uh at 100 gallons per acre, you can spray those 10 acres only, right? 00:17:40.319 --> 00:17:41.920 Then you come back, you refill it. 00:17:42.160 --> 00:17:51.200 So if you're spraying those same um 100 acres using this multiply, you could do 60 more. 00:17:51.359 --> 00:17:54.079 So you could do instead of doing 10, you could do 16. 00:17:54.559 --> 00:17:58.720 That's that's the I'll say another value of the system. 00:17:59.359 --> 00:18:19.519 And um I've seen a lot of adoption, uh Lenny, as you said, from growers, but I've also seen some growers, some growers um complaining about calibration, uh early season, uh, which is something I haven't I'll say faced yet because we've used the system for a year only. 00:18:19.680 --> 00:18:21.519 And I'm curious about that. 00:18:21.680 --> 00:18:28.720 That's that's that's the next step of research we're gonna be actually doing here uh for next year. 00:18:29.039 --> 00:18:36.640 Yeah, another big complaint that I hear from growers is that um they have to clean those sensors off. 00:18:36.720 --> 00:18:44.720 That's one of the I think one of the big keys to this, because you know, when you're spraying a lot of that stuff, well, you get the residue on the sensors and all that. 00:18:44.799 --> 00:18:47.279 So you have to clean those things pretty frequently. 00:18:47.359 --> 00:18:54.720 I think at least some what I've heard is like at least once a day, but probably more than that if you're doing a lot of acres, I imagine. 00:18:54.799 --> 00:18:55.039 Uh-huh. 00:18:55.200 --> 00:18:57.359 Um, you have to clean them up. 00:18:57.440 --> 00:18:59.839 Um, it's an optical sensor, right? 00:18:59.920 --> 00:19:11.839 Uh there's there, it's literally a as I told you, that is 360, but that that laser shooting 180 or like 200 degrees across the whole canopy. 00:19:12.079 --> 00:19:23.519 And even though they installed a blower to blow that dust, you know, because there is something on there all the time trying to blow that dust, it doesn't work uh every time. 00:19:23.680 --> 00:19:30.880 So you have to get, you know, your uh uh little uh uh they they give you a cleaner. 00:19:31.039 --> 00:19:39.119 Uh I don't know what that made, and uh a piece of cloth, go there, clean everything, and then go back to work. 00:19:39.359 --> 00:19:40.000 Yeah, yeah. 00:19:40.160 --> 00:19:45.119 So if you if you don't clean those sensors, it's not going to work right, and you're not gonna be happy with it. 00:19:45.200 --> 00:19:47.599 So that's

that maintenance is the key to it, I think. 00:19:47.839 --> 00:19:52.880 Yeah, that's the that's the price you pay for like 60% at least like off-chemical savings, right? 00:19:53.039 --> 00:19:53.599 Yeah, yeah. 00:19:53.680 --> 00:20:10.559 You gotta weigh, is it is it worth that little bit of time you spend doing that, is it worth the time you save both in in mixing that having to mix, you know, more more tanks if you're not using Smart Apply, but also um the savings that you the the financial savings you have there. 00:20:10.799 --> 00:20:28.960 Um and you know, the other thing you touched on this is that it's gonna be it's really better suited, I think, to um the orchards like you described with that have a lot more variability, um, different size trees, different, you know, maybe some skips in the orchard, missing trees. 00:20:29.200 --> 00:20:38.480 It's gonna save you a lot more in that situation than it would on a more uniform orchard where they're all the same age and same variety with all that kind of stuff. 00:20:38.880 --> 00:20:55.200 Uh yeah, yes, but like for example, exactly, because um, if you're doing a broadcast application on younger trees, like uh I'd say like less than uh 15, 20 feet tall, they're not touching each other yet. 00:20:55.279 --> 00:20:57.359 There's a huge gap between them. 00:20:57.680 --> 00:21:13.759 Um we haven't done the math or we haven't done the trials, but that might uh also represent a big chemical savings, even more when you have um older orchards with a more very variable environment there. 00:21:13.920 --> 00:21:18.799 Yeah, that's something we also have on the pipeline for for testing, just with water. 00:21:18.960 --> 00:21:30.480 Like what we've done this year was uh we calibrated for the chemical density with the flow meter and everything, but like this this one should be just with water and say, hey, what how much is the savings? 00:21:30.640 --> 00:21:35.759 We're we were not interested in um uh efficacy, right? 00:21:35.920 --> 00:21:47.839 Because this study that we've done with smart apply was hey, this does this 60% chemical savings also represents the same control of nut scab and also represents the same yield. 00:21:48.079 --> 00:21:58.720 Unfortunately, like the scab pressure on this trial was so high that we couldn't get to yield, but the control was the same in between smart apply and conventional application. 00:21:59.039 --> 00:21:59.359 Yeah. 00:22:00.000 --> 00:22:16.400 And I think, you know, as far as the efficacy goes, I think in in what I've seen in growers, orchards that are using it, most orchards, you know, it's it's working just as good as when there's if they're spraying it, you know, a conventional sprayer without the smart apply on there, as far as the efficacy goes. 00:22:16.640 --> 00:22:29.359 Um one thing I do hear from some growers is, you know, we uh you know, we've had three wet uh wet summers in a row, which is gonna drive a lot of pecan scab, especially over

here in the western part of the state. 00:22:29.440 --> 00:22:31.119 We had a lot of scab this year. 00:22:31.759 --> 00:22:42.640 And I hear some growers that were using smarterply, maybe say for the first time or whatever, during this stretch of of wet weather that we've had over the last three summers. 00:22:43.279 --> 00:22:51.599 And so a lot of them, you know, uh are blaming the smarter apply for for their scab issues. 00:22:51.839 --> 00:22:56.319 But I don't think it's I don't think it's the smarter apply that that they're getting that from. 00:22:56.400 --> 00:22:59.440 I think it's just you know, the conditions that we've had. 00:22:59.519 --> 00:23:04.559 They're their scab, we see scab in a a lot of orchards, whether they use smarter apply or not. 00:23:04.720 --> 00:23:16.319 So I don't think uh I think growers that are uh making that judgment are kind of it's kind of a knee-jerk reaction, and they might need to slow down a little bit and and think about it. 00:23:16.400 --> 00:23:20.880 Um, because I don't think it's solely the smarter apply that's driving that that scab. 00:23:21.279 --> 00:23:34.960 I think you're still your results and other other results I've seen in the past are showing that uh side by side it it works just as good as the conventional as far as uh you know efficacy goes. 00:23:35.359 --> 00:23:37.759 Um so yeah. 00:23:38.559 --> 00:23:43.359 But I guess that about covers uh smart apply. 00:23:43.759 --> 00:23:50.000 Uh and I know you're doing a lot of other stuff that we want to get to, so tell us what else you got going on. 00:23:50.720 --> 00:23:57.279 So um and on on still on the spraying sides, uh, we've done spray drones, right? 00:23:58.079 --> 00:24:13.119 So uh we've done a I'd say a wider trial trying to computate how much is the coverage on the different parts of the canopy using spray drones. 00:24:13.599 --> 00:24:29.359 So um and and that goes back to label a little bit because uh first, like when you're like I I have to tell this every time I talk about spray drones, like there are lots of regulations around uh those machines. 00:24:29.839 --> 00:24:52.720 And um first one is that you have to be um compliant with FAA on everything, you have to pursue all the licenses, which means part 107, part 137 uh waivers from FAA to should to uh spray chemicals from to dispense chemicals from an unmanned aircraft. 00:24:52.960 --> 00:24:59.039 You have to have your pesticide license from the state, so there's a lot of uh of uh regulations around it. 00:24:59.279 --> 00:25:02.480 And uh the second regulation is about the label. 00:25:02.640 --> 00:25:15.119 Uh, that product that you're spraying must be registered for aerial application, and then like just by looking at some of the uh pecan labels, uh pretty much all of them are. 00:25:15.519 --> 00:25:25.839 I do have uh, let's see here, I'm pulling my presentation from um this year's Pecan conference, but I'm I'm pretty sure I didn't put that table the table

here. 00:25:26.079 --> 00:25:36.799 So uh most of those fungicides are registered for aerial, so uh which are from the uh uh UGA recommendations, right? 00:25:37.519 --> 00:25:45.200 But most of those fungicides are also they also requires maximum output applications. 00:25:45.440 --> 00:26:00.000 So if if we spray 75 gallons per acre or 50 or 100 with an air blast, uh that concentration of the product is gonna be higher with the drone because the output that you're spraying, the GPA is lower. 00:26:00.319 --> 00:26:06.000 So uh most of those products they're gonna be at five or ten gallons per acre maximum. 00:26:06.160 --> 00:26:07.839 So you can spray less, right? 00:26:07.920 --> 00:26:18.160 You can spray like if if the max is 10 gallons per acre using an aircraft, you can spray nine, seven, five, but you cannot spray 11 or 12 according to you. 00:26:18.240 --> 00:26:19.680 You can you cannot go up. 00:26:20.000 --> 00:26:43.839 So uh the first trial we've done was testing those two different outputs, like the two different amounts of chemicals, which was so we found actually within 13 gallons per acre and uh five gallons per acre, because at the time we had a spray drone, 13 gallons uh was the the tank size, so within 13 and five. 00:26:44.079 --> 00:26:47.920 So at five, no good coverage at all. 00:26:48.079 --> 00:26:58.319 Um, and I'm not talking about coverage and percentage, I'm I'm talking about droplets per square centimeter, which are tied up to fungicide efficacy. 00:26:58.640 --> 00:27:06.880 So uh at the top, we've reached that uh threshold of 70 droplets per square centimeter. 00:27:07.200 --> 00:27:20.079 Um at the middle of and in the middle of the canopy there, so the way we've done that, just going back a little bit, um we installed water-sensitive paper cards across the whole canopy of the tree. 00:27:20.400 --> 00:27:25.440 So uh top, like on very tall trees, I would say. 00:27:25.680 --> 00:27:36.880 Um and then we've done uh trees over 50 feet, trees in between 25 and 50 feet, and trees below 50 uh 25 feet. 00:27:37.039 --> 00:27:41.200 So we call them mature, mid-age, and young uh trees. 00:27:41.279 --> 00:27:46.880 Okay, so at the top, all of those trees were covered. 00:27:47.119 --> 00:27:58.559 So the drone was doing the job of covering that that canopy top, flying at 10 feet above that canopy, okay, with a 16 uh feet swath, effective swath. 00:27:59.200 --> 00:28:02.880 And that's on like a say a big 60, 78. 00:28:03.599 --> 00:28:03.920 Exactly. 00:28:04.000 --> 00:28:04.720 Yes, yes. 00:28:04.799 --> 00:28:20.160 Uh the big trees or um uh a mature, or we call it mature, like which are the big, uh, mid-size 50 to 75 feet, and then also um uh actually 25 to 50 feet, and then below 25 feet as well. 00:28:20.559 --> 00:28:23.119 So that that top was covered. 00:28:23.279 --> 00:28:38.240 However, when you get to uh that midterm on those trees, um none of those treatments, either five or 13 gallons per acre, uh, could actually uh bring that fungicide down. 00:28:38.559 --> 00:28:42.640 Um how far down you say how far down you say penetrate? 00:28:43.519 --> 00:28:56.240 It on I that study was hey, if the tree is 75 feet uh or 60 feet, you we fraction that into three pieces, like so that's 20, 20, and 20. 00:28:56.480 --> 00:28:56.799 Okay. 00:28:57.039 --> 00:29:09.039 So uh and then the uh the cars were placed, I would say, very at the top of the tree, like when when getting that that that that sprayed, right? 00:29:09.359 --> 00:29:14.240 On the middle was like if it's 60 feet, 30 feet down. 00:29:15.119 --> 00:29:20.240 So um I I I'm pretty sure you've seen this graph before when I showed this on conferences. 00:29:20.480 --> 00:29:40.480 We still got some some of that um output there, but it wasn't enough uh to justify uh those 70 droplets per square centimeter on younger trees below 25 feet, uh actually the ones we tested, they were pretty young, they were like actually below 15 feet. 00:29:40.799 --> 00:29:48.640 So we could instead of getting uh 70 droplets per square centimeter, we got 50. 00:29:48.880 --> 00:30:08.000 So uh, which when we go back to the literature, there's a study from uh Singenta that says that in between 50 or 70 droplets per square centimeter, you could get you could get good efficacy on uh systemic fungicides, uh systemic fungicides. 00:30:08.400 --> 00:30:13.680 So um, but that was just on the young ones, on the big ones and mid-sized ones. 00:30:13.839 --> 00:30:21.599 No, like you cannot get that output all the way down to that tree, unless uh, and and we're only talking about 13 gallons per acre, okay? 00:30:21.680 --> 00:30:40.240 We're not talking about like 20, because uh if you go up to 20, there's only one fungicide, and I don't record which label is for pecans according to the uh UGA recommendations, that you can go to 20, but is inefficient for drone applications. 00:30:40.400 --> 00:30:44.240 Like uh the tank is right now the the biggest tank. 00:30:44.400 --> 00:30:53.920 Uh there is a drone that is 50 gallons, uh very massive, serious, uh, serious air, uh, very expensive as well. 00:30:54.160 --> 00:31:07.839 But the ones that we have, like the T100 from DJI or XAGP 150 max, I think they're like 25 gallon tanks, which uh decreases a lot the efficiency. 00:31:08.000 --> 00:31:14.720 So it's basically one tank goes sprays, comes back, so one tank at a time per acre. 00:31:15.039 --> 00:31:16.240 Well, that's a lot. 00:31:16.480 --> 00:31:17.680 Yeah, it's a lot. 00:31:18.559 --> 00:31:29.759 So in most, you know, a lot of people hedging now, so you know, I think kind of a you know, peak height that we want to hedge at to me is about 40 feet. 00:31:30.079 --> 00:31:36.880 So with at that height, it would probably still just cover the top, but not get much from the on the on the bottom, right? 00:31:37.119 --> 00:31:47.039 It would cover that top, it would cover that top, and if you're talking about uh I have here

in front of me, you could re actually reach a little bit of the middle, right? 00:31:47.440 --> 00:31:52.480 Uh, but the bottom, uh that would actually be a good trial. 00:31:52.640 --> 00:31:57.759 Yeah, it's a hedge and non-hadget to see the coverage inside of that tree. 00:31:58.000 --> 00:31:59.039 Yeah, we can do that. 00:31:59.119 --> 00:32:00.319 Let me know if you want to try it. 00:32:00.640 --> 00:32:01.039 Awesome. 00:32:01.200 --> 00:32:01.680 Yeah. 00:32:02.079 --> 00:32:05.119 Um one thing I wanted to ask about this. 00:32:05.279 --> 00:32:09.039 So you know, it works great for young trees. 00:32:09.599 --> 00:32:15.200 Um it does great on the top for medium size and then tall, big old trees. 00:32:15.519 --> 00:32:25.359 Um a lot of the issues we have with scab control, uh especially or we see a lot of it on these big trees where coverage is an issue. 00:32:25.519 --> 00:32:34.880 Where with the it with the air blast sprayer, doesn't matter which one it is, you're not getting much coverage above about 40 feet, which is why, you know, that to me that needs to be kind of the peak. 00:32:35.039 --> 00:32:37.200 If we're hedging, that needs to be the peak height. 00:32:37.440 --> 00:32:40.480 Um because you're not getting much coverage above that. 00:32:40.799 --> 00:32:50.880 But um where I see a big value for this is in these big old trees where there's no chance of getting good coverage up in the tops of those trees. 00:32:51.119 --> 00:33:26.720 And we see this some with uh aerial applications with airplane too, that if you target those sprays for the top of that tree at at critical times, usually that's going to be sometime from June to August, you know, when you got a lot of pressure for nuts gab, um you can use the aerial applications to kind of supplement what you're already doing down below that can can help you a lot with with uh getting coverage in the top and protecting some of those those nuts in the top, which a lot of times on those big old trees, that's where most of your crop is. 00:33:26.960 --> 00:33:28.240 It's the top of that tree. 00:33:28.400 --> 00:33:32.480 So I do see some value with it, yeah, especially for that. 00:33:32.799 --> 00:33:36.319 Yeah, I have I have some mixed mixed feelings about it, Lenny. 00:33:36.400 --> 00:33:52.640 And and that's because um if we just follow the label, right, and then we spray um um down below 75 gallons per acre, um, and then you come like from the top below uh uh down up, right? 00:33:52.720 --> 00:34:00.319 Like you go down up at 75 gallons per acre, and then you come from the uh with the drone and then spray 10 more. 00:34:00.480 --> 00:34:02.160 Are you like double spraying? 00:34:02.559 --> 00:34:03.279 Yeah, exactly. 00:34:03.359 --> 00:34:18.400 That that's that that's that's those are the mixing mixed feelings I have, even though like we have results proving that you're getting very, very amount of that chemical um from the from the top to the bottom. 00:34:18.639

--> 00:34:21.039 Yeah, but you're still only allowed so much per acre. 00:34:21.360 --> 00:34:22.320 Exactly, exactly. 00:34:22.400 --> 00:34:24.320 That's per acre, not per tree, right? 00:34:24.400 --> 00:34:24.639 Right. 00:34:24.880 --> 00:34:27.360 So I have uh some mixed feelings about it. 00:34:27.599 --> 00:34:29.199 Um that's a very good thing. 00:34:29.440 --> 00:34:44.559 But like speaking about it, um, I know that that has been a dry year, but um my studies like because we've done this, like actually uh the Pecan uh commission has been so supportive of these trials with the drones. 00:34:44.719 --> 00:34:49.599 Um and the uh proposal last year was hey, how about with if we supplement? 00:34:49.679 --> 00:34:51.119 We're yes, we're off label. 00:34:51.199 --> 00:34:54.239 To me, we're off label, but how about if we supplement? 00:34:54.400 --> 00:34:59.440 How how does um how how how's that gonna be, right? 00:35:00.000 --> 00:35:02.559 And uh I have the results in front of me here. 00:35:02.639 --> 00:35:04.400 Those are the mid season results. 00:35:04.480 --> 00:35:05.119 Uh Dr. 00:35:05.280 --> 00:35:12.639 Uh uh Clement Oliveira, he has been doing the scab grading, uh, and this is not scab severity. 00:35:13.039 --> 00:35:23.039 It was the same, like uh like we got the same results actually doing just the standard program, um, and then doing the air blast plus a drone. 00:35:23.519 --> 00:35:25.119 Yeah, when was that study done? 00:35:25.199 --> 00:35:26.960 You remember it was here uh at Ponder. 00:35:27.840 --> 00:35:29.039 Yeah, it was here in Tifton. 00:35:29.199 --> 00:35:31.920 Uh we had the name of the orchard here. 00:35:32.079 --> 00:35:45.440 So yeah, we have pressure here, but like uh when we look at uh and then we went beyond because we wanted to understand also uh what's the problem if we just spray with the drone. 00:35:45.679 --> 00:35:49.199 But for doing that, I need to be even more off-label. 00:35:49.360 --> 00:36:00.000 Like uh um we what we've done, we we got uh the air blast sprayer versus the drone by itself, but we only sprayed mirror stop. 00:36:00.239 --> 00:36:15.360 So we've done like 10 applications of mirrorvis with each one of those technologies because the intention is not to uh assess like um uh chemical efficacy, but uh to assess the technology efficacy. 00:36:16.400 --> 00:36:22.480 And uh so it was mirror stop over mirrorvis, and that's those results were also the same. 00:36:22.719 --> 00:36:47.519 Like uh the drone uh on that mid-size orchard got the same actually, uh the severity using mirror over mirrorvis with the air blast sprayer at 50 gallons per acre was 21 percent, and uh with the drone this cap severity on on the nuts was about like five, three to five percent. 00:36:47.840 --> 00:36:54.719 So no statistical differences, and the entrea check entreated check was about 45 to 50 percent. 00:36:54.880 --> 00:36:58.400 Yeah, so it was doing the job, it was doing the job, exactly. 00:36:58.559 --> 00:37:03.280 But again, it goes back to what Andrew was mentioning at the uh early on. 00:37:03.519 --> 00:37:05.360 It's been so dry. 00:37:05.599 --> 00:37:14.239 Um that that might be uh why the drone is working the same as uh uh the air blast in this case. 00:37:14.480 --> 00:37:22.079 Well, I think if that was yeah, if that's if that was done in East Georgia, it's it was dry in East Georgia, so they that's why you don't see much scab over there. 00:37:22.159 --> 00:37:35.440 They didn't get the rain and the pressure we had, but I would say from certainly here in Tifton, uh we had we had enough pressure here to drive quite a bit of scab pressure or a lot of scab uh incidence. 00:37:35.679 --> 00:37:40.960 Uh so yeah, I think you had enough pressure in that test to see if it worked or not. 00:37:41.119 --> 00:37:43.679 Yeah, the non-treated check was about 45%. 00:37:44.320 --> 00:37:54.400 Um and then one thing to point is that like uh this is the mid-season that was done mid-June uh through early July, this analysis. 00:37:54.719 --> 00:38:01.119 But like for the next one that we've done in mid-August, we've also got a lift. 00:38:01.360 --> 00:38:08.159 We went all the way to the top, and then we measured scab severity at the top, at the middle, and at the bottom of that tree. 00:38:08.320 --> 00:38:14.400 So we could actually compare both of those technologies uh to see if the drone is doing like a better job. 00:38:14.719 --> 00:38:16.719 Those results are being processed right now. 00:38:17.039 --> 00:38:23.119 Uh, if the drone does a better job on the top versus that um air blast on the bottom. 00:38:23.440 --> 00:38:24.159 Well, that's great. 00:38:24.480 --> 00:38:26.159 I'll be looking forward to seeing that. 00:38:26.239 --> 00:38:27.920 So that's that's good good work. 00:38:28.159 --> 00:38:32.159 And I know you've only got probably about 15 minutes or so left with us. 00:38:32.639 --> 00:38:37.760 Is there are there any other studies on pecans before we go into this next little question? 00:38:37.840 --> 00:38:39.679 Uh before there are. 00:38:39.920 --> 00:38:40.239 Okay. 00:38:40.559 --> 00:38:41.119 Go ahead. 00:38:42.000 --> 00:38:51.280 Um, so the next one is about drones, like how we could use those little things to help the growers to be more efficient. 00:38:51.599 --> 00:38:59.440 Indeed, I have a grower coming in this afternoon to look at some of the results uh we uh were having with that. 00:38:59.760 --> 00:39:14.960 So the idea again is to use um artificial intelligence to do the job for us with the drones um on measuring tree size, counting how many trees you have um in that orchard. 00:39:15.119 --> 00:39:16.639 I know you know it, right? 00:39:16.960 --> 00:39:30.239 But uh if you have tree by tree um I'd I would say data or those measurements tree by tree, you can actually uh be more efficient on determining how much you're gonna spray. 00:39:30.559 --> 00:39:36.559 So on those studies, we flew

uh a Mavic 3 multispectral at 200 feet. 00:39:36.719 --> 00:39:41.440 So the drone was capturing uh uh a very good pixel size. 00:39:41.519 --> 00:39:45.679 So we're not talking about pixel, every picture is full of pixels, right? 00:39:46.000 --> 00:39:49.599 But they are like uh very, very tiny, so you can barely see it. 00:39:49.840 --> 00:40:01.920 But on old on those back old days on those like uh digital cameras, when it when you've taken a picture, maybe like 10 years ago, if you zoom in, you see a lot of those little squares, yeah. 00:40:02.159 --> 00:40:03.920 So that's what we've done. 00:40:04.000 --> 00:40:13.760 Uh we flew, and the pixels were about like half an inch, which is uh smaller still than a pecan leaf, like a pecan tree leaf. 00:40:14.079 --> 00:40:19.119 So um by zooming in, we could see individual leaves still. 00:40:19.199 --> 00:40:20.880 So that's what we want. 00:40:21.280 --> 00:40:40.159 So we flew that drone uh 200 feet above the canopy, and we wanted to know how accurate those uh we call them uh digital wag platforms because you get you get you you fly the drone, you take like several pictures, and then you send that to a platform, you pay for it, of course. 00:40:40.400 --> 00:40:49.199 Um you pay for it, and then at that image is stitched, so they create a map for you automatically. 00:40:49.360 --> 00:40:51.360 So a map of your whole field. 00:40:51.440 --> 00:40:56.079 So let's say it's it's 25 acres, you cannot cover 25 acres with one shot. 00:40:56.239 --> 00:40:58.880 You have to take like several different shots, right? 00:40:59.199 --> 00:41:32.800 So it processes, it gives you a map, but uh uh those new platforms they they they uh bring more services, so you can train AI models inside of those platforms just by uh drawing, uh teaching the system on hey, this is a pecan tree, or uh and then you circle those trees, not all of them, some of them in that software, and then uh the software learns what the tree is and then it extracts the data from it. 00:41:33.039 --> 00:41:48.719 So, for example, in that study, we've done mature, again, mid-size and very, very young trees, because we wanted to know how accurate those softwares are on measuring uh canopy size and plant height. 00:41:49.440 --> 00:41:51.519 Um, why we've done that? 00:41:51.679 --> 00:41:57.679 Because again, we wanted to be more efficient on determining how much output I want to put on those trees. 00:41:57.840 --> 00:42:05.039 According to UGA, if the tree is over uh 50 or 75 feet, I apply 100 gallons per acre, right? 00:42:05.199 --> 00:42:06.880 And that's a flat application. 00:42:07.039 --> 00:42:11.840 If the tree is below 50, uh, I apply 75 gallons per acre. 00:42:12.000 --> 00:42:24.079 And if the tree is in between 25 and 50 feet, I believe, or or no, 75 to 50 is 75 GPA, and then 50 to 25 is 50 gallons per acre. 00:42:24.800 --> 00:42:27.519 Um, orchards are not the same. 00:42:27.679 --> 00:42:35.840 Uh way, like uh across the time,

people had transplanted, so you had younger, you have larger trees. 00:42:36.079 --> 00:42:43.920 I uh so uh in our eyes, they're not really good on telling us how tall things are. 00:42:44.159 --> 00:42:50.239 So, like sometimes we go to an orchard, we say, Hey, that tree is over 75 feet, so I'm gonna just put 100 gallons per acre. 00:42:50.400 --> 00:42:52.880 But indeed, that tree is actually 50. 00:42:53.119 --> 00:42:54.079 A little bit higher than that, yeah. 00:42:54.239 --> 00:42:54.639 Yeah, yeah. 00:42:54.719 --> 00:43:02.320 We missed uh we missed by uh uh 25 feet, and we are spraying 25 gallons more just because of that. 00:43:02.719 --> 00:43:06.719 So, bottom line, those results came in 99% accurate. 00:43:06.960 --> 00:43:24.639 So uh, and the uh I would say conclusion is yes, you can measure trees with drones, yeah, and not only the height, but also the size of the volume of not the volume, the size of that canopy in uh square inches, for example. 00:43:24.960 --> 00:43:36.159 And um uh with that, we've done, hey, can we actually identify down the trees after hurricane severe with severe weather events? 00:43:36.400 --> 00:43:46.880 Yes, we could uh add we we could do that as well with uh more than 90% accuracy, so that saves on um time, right? 00:43:47.199 --> 00:43:52.159 For uh uh I'll say responding to those severe weather events. 00:43:52.320 --> 00:43:59.199 Like after hurricane, fly a drone, count how many downy trees you have for insurance claims or ASA. 00:44:00.079 --> 00:44:07.360 uh also helps for good help for government incentives afterwards doing that inventory. 00:44:07.519 --> 00:44:11.519 So yeah, so those were I would say three major studies. 00:44:11.679 --> 00:44:25.360 The first one is Mart Applied, second one spray drone, and third one uh using the uh scouting drones for inventory management I would say on on Pecan Orchards. 00:44:27.760 --> 00:44:58.639 I wanted to have this discussion a little bit I'm I try to be open-minded um but I'm also I'm a little bit old school too and so um there's part of me that is a little skeptical of not not your work or anything but just like the whole AI and and super high tech in general. 00:44:58.960 --> 00:45:07.599 And one of the issues I have at this point with it is that there's so much variability in pecans. 00:45:08.079 --> 00:46:33.679 You know, within one orchard from one orchard to the next you know you got you got different varieties, different age trees different soil types just you know different spacings different uh management practices there's so much variability in it that I've always been a little concerned that you know any model whether you're talking to AI model or any type of model is only as good as the information that that you're putting into it and one of my concerns has always been you know with all this variability um and pecans are a very young crop um you know they've only been grown let's say commercially for uh say a little over a hundred years um and so they don't I don't the the information out there on pecans and growing pecans and how they respond to different things and we don't have the database that things like corn and cotton and all those things have so that's always been a concern of mine but you know we're always learning more just the stuff you've talked about today you know has shown that yeah there there's places this definitely fits and can work. 00:46:34.239 --> 00:46:43.039 So you've helped to calm my my concerns a little bit um with with all the technology out there now. 00:46:43.840 --> 00:51:22.800 I still feel like I'm I'm kind of being left behind a little bit because I'm you know and and then you were right Lenny um like the data you collect they are uh for I would say estimation right like for example what you mentioned like those models they are for disease prediction or yield prediction they are as good as the information you collect and they are right now uh on the pecan side right they are right now applied to that orchard where you collected that data yeah but um when when you're talking about like plant height or uh these biometrics studies like uh it's uh you you don't need that two years of data because that's not uh crop um focused that's object focused so you can measure the height of pretty much anything um using uh a difference in between uh the RTK GPS uh coordinates from the top of that tree to the soil so that's uh that's it that's that's where that ai comes in but when you're talking you you're you're your concern is fully valid and you're completely right like when we're talking about like uh modeling for uh disease prediction right uh you need a lot of environmental data you need a lot of inputs from uh the region not only from the field uh specifics to determine hey scam pressure in that field is gonna be very very high this year so you need historical data you need you need a lot of information in order to uh be very certain about something when you're talking about modeling you're you're completely right then your concerns are are uh fully valid well I mean I think this is just I mean it's just a new new we're getting into an age where it's we just have new forms of technology available to us and and those technologies just like all technologies in the past they're they're a tool you know they're they're a form of a tool that that growers can use and and probably should use in a lot of cases but I think the one thing I want growers that I always want to emphasize with this stuff is that these are tools that that have certain have specific uses that that can help us but there's no one institute for growers themselves being out in that field though those orchards and those fields and knowing what's going on out exactly exactly exactly um you you know about the pecan shield project right that we are developing we just sent this proposal and we want you build those models but those models are going to be built within five years as say um we're gonna collect data from 15 years ago up to uh 20 years right because with within five years we want to have a model for scab prediction and we're gonna use a a different tool like because like for us to estimate or determine how much scab we're gonna have we need to know how much scab we currently have and that's the uh AI driven uh spore trap pretty sure you've seen those right like uh it's it's it's a scanner is a microscope that collects the dust from the air and then there's a microscope there with an AI that says hey that's a spore like a pecan's uh scab spore and then depending on how many you have inside of the microscope which is automatic uh you can determine using environmental data like temperature if that's haggard non-haggit if it has cover crops or non-cover crops in the field uh that scab severity button should could be higher or lower lower so yeah we're we're we're we're getting there but it's gonna take some time for sure Lenny yeah well it's certainly an exciting time for for someone in your position to be be working on all this stuff and I'm glad we have you here to do that because I certainly couldn't do it okay so I appreciate you guys guys it was a very very nice conversation I enjoyed a lot yeah we sure appreciate you coming on and speaking with us Luan thank you Luan look forward to working with you yes sir yes gentlemen stay tuned for the next episode you can get more information on the University of Georgia Pecan blog in a nutshell the Pecan podcast