The Imprecision of Using Drones for Precision Pesticide Applications

University of Nebraska – Lincoln Author

08/31/2020 Added

82 Plays

Description

Dr. Greg Kruger Research update on the use of drones for application of pesticides, looking at coverage uniformity.
Feedback form

Searchable Transcript

Search:

[00:00:06.560]Hello, today I'm gonna talk a little bit
[00:00:08.750]about the imprecision of using drones
[00:00:10.900]for precision pesticide applications tool.
[00:00:13.170]What I'm gonna do is talk a little bit
[00:00:14.590]about using drones for pesticide applications
[00:00:16.681]and a little bit of the research that we've done.
[00:00:19.635]As you can see,
[00:00:20.520]I've got a couple of coauthors down there.
[00:00:22.356]Dr. Brad Fritz, USDA, ARS scientist,
[00:00:25.640]and one of my graduate students Trenton Houston,
[00:00:27.670]these two both have been integral in terms
[00:00:30.490]of helping develop the material for this section.
[00:00:33.930]Without them, this really wouldn't have been possible.
[00:00:36.840]Using the drones for pesticide applications,
[00:00:39.380]it's not necessarily a new concept,
[00:00:41.160]but it's a concept that's really slowly starting
[00:00:44.180]to take off in the US.
[00:00:45.470]Here you can see a couple pictures of drones
[00:00:47.940]that we have in our lab.
[00:00:49.660]These are both set up to make pesticide applications.
[00:00:52.160]As you can see there, there's some more,
[00:00:54.150]but yet there are some differences
[00:00:56.550]between the both of those.
[00:00:58.220]For example, the one in the top left hand corner,
[00:01:01.020]six propellers,
[00:01:01.860]the one on the bottom right hand corner has got four.
[00:01:04.800]The two drones have a different tank, carrying a tank sizes
[00:01:08.890]and different lift capacities and things like that.
[00:01:10.810]So as you, as we go through this, you'll see that,
[00:01:14.350]there's really a key thing there that every one
[00:01:17.240]of these drones seems to be a little bit different
[00:01:19.200]in terms of their function, in terms of what they do
[00:01:21.370]and how they work.
[00:01:22.460]So when we think about the benefits of drones
[00:01:25.170]for pesticide applications, the true benefit
[00:01:27.470]is that we can reach areas that may not be accessible
[00:01:31.070]by traditional application methods,
[00:01:33.080]areas where airplanes and helicopters
[00:01:35.370]might have a hard time getting in,
[00:01:36.650]but yet for one reason or another,
[00:01:38.180]we can't get a ground application right there,
[00:01:40.340]you know areas.
[00:01:41.173]So along riparian areas,
[00:01:42.730]sometimes fruit and vegetable production,
[00:01:44.510]high value crops that may not be growing
[00:01:46.780]on the perfectly well flat line areas, things like that.
[00:01:50.290]Or maybe make an applications in enclosed airspace's
[00:01:53.570]moreso inside the orange building sort of things like that.
[00:01:57.470]Really makes us a benefit if it's fully autonomous.
[00:01:59.960]So we have the technology today that a drone
[00:02:02.740]could land on our refueling tank.
[00:02:05.050]And really the only thing they operator
[00:02:06.700]would be doing in terms of the actual operation
[00:02:09.970]would be changing the batteries out.
[00:02:11.640]The rest of it would be autonomous flight
[00:02:13.770]off of pre-scripted flight plans.
[00:02:17.090]I put them the next bullet point on here,
[00:02:19.790]and then they can't be compared to aerial
[00:02:21.430]or ground applications.
[00:02:22.500]We'll going into some detail about that,
[00:02:24.130]but they really are truly different
[00:02:26.280]than our traditional area or ground applications
[00:02:29.450]in a number of different ways.
[00:02:31.610]One of the things that you must recognize is,
[00:02:34.690]in the US that we have to have a drone operator license
[00:02:38.570]through the FAA but it also requires a special permit.
[00:02:42.040]So there's a number of exemptions and things like that,
[00:02:45.160]that you have to file with the FAA in order
[00:02:48.090]to actually release pesticides from the drone application.
[00:02:51.217]And that process is continually evolving
[00:02:54.130]and quite cumbersome.
[00:02:56.770]Now, however, even though there's not a lot of folks
[00:03:00.000]in the US that are flying drones
[00:03:01.660]for pesticide applications,
[00:03:03.470]globally there's a lot of of work that's already been done.
[00:03:08.990]And a lot of applicators
[00:03:10.160]that are out there making applications.
[00:03:11.987]And it's estimated that there's approximately 55,000 drones
[00:03:15.750]in China that are making applications.
[00:03:17.730]And that number maybe as much as a 200,000 or more today,
[00:03:23.446]the 55,000 as a reliable number that we've got.
[00:03:26.200]But we suspect that there may be quite a few more than that.
[00:03:29.140]And globally, there may be as many as 10 to 15,000
[00:03:32.590]other drones making applications,
[00:03:34.660]where in the US we're certainly a less than 50 today.
[00:03:38.860]So when we think about manned versus unmanned applications,
[00:03:44.400]there's really three areas that I think deserve a little bit
[00:03:48.720]of attention when we talk about, be a different,
[00:03:51.945]the speed, capacity and scale.
[00:03:54.740]So if we think about the speed of the application,
[00:03:57.230]a typical air tractor, five or two today,
[00:04:00.540]is gonna have a working speed of 120 to 160 mile an hour.
[00:04:05.340]And that may be a big conservative
[00:04:07.460]that maybe even a little bit faster than that.
[00:04:09.810]So absolutely high-speed, environment,
[00:04:13.960]we've got a high shear environment on that liquid coming
[00:04:16.380]out of that also as a quite different environment.
[00:04:19.660]When we think about the drone, typical working speeds,
[00:04:23.020]I have 90 miles an hour here.
[00:04:25.327]The 18th is probably pushing what the drone can do.
[00:04:29.150]And the nine, maybe a little bit high,
[00:04:31.990]even though some of the research
[00:04:34.480]that we're seeing may suggest
[00:04:35.730]that even in that six to eight mile an hour,
[00:04:38.120]it might be a better working speed.
[00:04:39.480]So, we're talking about speeds
[00:04:41.250]that are much just more similar to ground applications
[00:04:43.930]than actual aerial applications.
[00:04:46.630]When we talk about capacity,
[00:04:49.360]particularly for us here in Western Nebraska,
[00:04:51.130]as you guys all know large fields,
[00:04:54.290]this is not the type of application that we're looking
[00:04:57.100]to replace an airplane with.
[00:04:59.400]An air tractor five or two has got a 500 gallon capacity.
[00:05:03.930]Our air tractors range anywhere from a 400 or three
[00:05:07.220]or 400 gallons on up to as much as 800 gallons,
[00:05:12.460]the high end capacity of an unmanned aircraft
[00:05:17.300]that we'd be using to make a drone application today
[00:05:19.890]would be five gallons.
[00:05:21.134]Most of them are smaller than obvious.
[00:05:25.474]And then the last one scale,
[00:05:26.369]and this is what really limits us in terms
[00:05:29.970]of scaling this thing up into something
[00:05:33.060]that could be used for a row crop.
[00:05:36.530]The limitation really becomes the size of the aircraft.
[00:05:39.500]So, if we go back to that air tractor five or two,
[00:05:42.360]we're looking at a 4,500 pound aircraft,
[00:05:44.833]a 52 foot wingspan and a working range of 620 miles,
[00:05:50.490]so round trip.
[00:05:51.470]So, we look at that compared to that drone.
[00:05:54.160]Now we're gonna be a 40 pounds
[00:05:55.960]without the batteries on there.
[00:05:58.060]FAA is gonna restrict us to a 55 pound gross weight takeoff,
[00:06:02.700]and we could have a wingspan eight rotor wingspan
[00:06:06.980]that maybe a few foot across,
[00:06:10.143]on a range of anywhere from eight to 15 minutes loaded.
[00:06:14.720]So we're really not going to go much further
[00:06:18.040]than line of sight even if we were allowed to.
[00:06:21.840]However FAA restrictions that instructs us to define it,
[00:06:24.810]the line of sight,
[00:06:25.800]so this thing's not going to go 600 miles.
[00:06:28.930]It's gonna only operate in the area
[00:06:30.720]where we're making that application.
[00:06:34.580]Now, the big issue when we start talking about drones
[00:06:39.400]or unmanned aircraft, is there's a big data gap in terms
[00:06:45.001]of both information available for applicators
[00:06:49.770]as well as information available for the FAA
[00:06:53.070]and particularly the EPA to make policy
[00:06:57.290]and regulatory decisions.
[00:07:00.460]The gaps, and you'll see in some of our data user,
[00:07:04.837]these are real first one, at least height.
[00:07:08.430]So if we think about a ground application,
[00:07:10.603]we're ideal, boom height is gonna be anywhere
[00:07:13.870]from 15 to 30 inches,
[00:07:17.647]forwarding an air tractor five or two or another airplane.
[00:07:21.740]We're looking at the ideal boom height of 10 to 12 feet.
[00:07:25.550]The drones are probably gonna be somewhere
[00:07:27.030]in between those two, but exactly where that false,
[00:07:31.240]we don't know.
[00:07:32.530]A flight speed effects.
[00:07:33.630]So we talked a little bit about that range of flight speed,
[00:07:36.340]but we don't know exactly what flight speed effects
[00:07:39.120]are gonna have on the distribution deposition
[00:07:43.730]of that pesticide application
[00:07:45.670]and the efficacy of that application.
[00:07:48.130]So there's a big gap there.
[00:07:51.100]The next one on my list
[00:07:52.130]was nozzle boom placement impacts of,
[00:07:54.365]we think about it,
[00:07:56.670]the droplet size coming out of a drone is gonna
[00:07:59.210]be very similar to a ground rig if we're using
[00:08:03.170]the same nozzles, because we don't have that sheer.
[00:08:06.190]But if we use the same application parameters
[00:08:10.650]that we have, that boom height higher
[00:08:13.600]than what we have for ground application,
[00:08:15.650]we have the airflow effects coming off of that aircraft.
[00:08:20.690]As you see in our next bullet point,
[00:08:22.080]the multi-rotor wash effects
[00:08:25.710]that the boom placement nozzle spacing
[00:08:29.020]and also placement on the booms become very very different
[00:08:33.660]in terms of how we're gonna try to set that up.
[00:08:36.500]Also for unmanned or drone deer aircraft applications.
[00:08:40.738]We're generally looking at very low volumes compared
[00:08:43.110]to what we would a typical ground application.
[00:08:45.860]Which have a ground application,
[00:08:46.830]we might be looking at anywhere from a five to 20
[00:08:50.050]or 25 gallon per acre or row crops.
[00:08:53.520]For an unmanned drone applications,
[00:08:55.900]we're gonna be looking at,
[00:08:56.790]is somewhere from a quarter up to maybe a gallon per acre
[00:09:01.210]at the most.
[00:09:02.043]So very very different in terms of volumes there.
[00:09:06.270]And that's gonna affect that the droplet size
[00:09:09.130]and distribution pattern.
[00:09:11.580]Effective swath width versus the uniformity and rate,
[00:09:14.430]it we'll get into this.
[00:09:15.500]But today with the knowledge we have the equipment
[00:09:20.050]that we have, this is a very imprecise science compared
[00:09:23.500]to what a ground or aerial application might be.
[00:09:27.540]I say all of the above for different UAS or unmanned drones,
[00:09:34.020]because every single drone is gonna be set up differently,
[00:09:36.640]different miles, different booms, a different number
[00:09:39.670]of rotors, different application speeds, boom heights.
[00:09:43.290]And so there's a lot of different combinations
[00:09:45.810]that we really don't understand what those interactions
[00:09:48.150]are between those different combinations.
[00:09:50.340]Now, I'm gonna spend the pretty much the rest of the time
[00:09:53.100]that we have talking about the swath width
[00:09:56.760]and how we set this up for uniformity coverage
[00:09:59.710]and more or less how difficult
[00:10:04.140]that is to get a uniform pattern.
[00:10:05.880]So, I won't go into lots of detail
[00:10:08.440]but just to give you a feel for what it looks like
[00:10:11.805]this top part of the figure up here is what we get
[00:10:17.460]in terms of distribution across that boom for a single pass,
[00:10:20.710]the black box that you see as the,
[00:10:23.995]what we would consider the optimum distance to optimize
[00:10:29.220]that swath width, we're looking at,
[00:10:30.970]in this case at a 5.5 meter, roughly 16 to 17 foot swath.
[00:10:37.536]And if we look at that CV value,
[00:10:39.810]and I talked about them being between 25 and 75%,
[00:10:43.700]here we're right at 55% with a mean coverage of 10%.
[00:10:49.150]You can see also inside this,
[00:10:51.517]the distribution of the droplet sizes
[00:10:54.220]across that spray swath
[00:10:56.480]and the green representing the largest droplets
[00:10:59.360]that we see in this particular application
[00:11:01.160]or the extra course.
[00:11:02.890]The blue being the very course, the yellow being the course.
[00:11:06.520]And so you can see that those extra course drops
[00:11:09.700]are deposited right underneath the equipment.
[00:11:12.570]Maybe with just a little bit of shift to the left side
[00:11:15.320]of that sprays wall, the larger droplets,
[00:11:19.372]the smaller droplets getting pushed out away from the center
[00:11:23.310]of the unmanned aircraft.
[00:11:25.380]And then smallest droplets are those core droplets being
[00:11:28.310]on the very outside edges of that.
[00:11:31.390]Now the next figure down shows the,
[00:11:34.490]what nine overlapping patterns
[00:11:36.305]would be with that 5.5 meter swath width
[00:11:39.620]and the red bar now represents
[00:11:42.009]what the optimum coverage should look like.
[00:11:46.400]This is for 10% coverage
[00:11:48.005]and where you see gaps below that red bar,
[00:11:52.280]that's a lower dose than what we were targeting,
[00:11:56.620]where you see the spikes,
[00:11:58.225]those green spikes sticking up above the red bar.
[00:12:01.940]This is where we're getting more output than what we desire
[00:12:05.720]and what this starts to lead to is potential for where we
[00:12:08.600]get higher doses or improper injury
[00:12:10.842]or damaging desirable vegetation
[00:12:15.610]or where we see gaps down below,
[00:12:17.710]this is potentially an ineffective dose,
[00:12:21.070]which could lead to selection, pressure for resistance.
[00:12:24.140]So, as you can see here,
[00:12:25.210]should we start to set this up,
[00:12:26.770]this scenario where we really aren't getting
[00:12:29.800]a nice clean application.
[00:12:31.520]Now, real briefly with the last minute or two,
[00:12:33.650]I just wanna show you what that looks like in terms
[00:12:36.350]of modelling this out.
[00:12:37.500]So this is what the EPA would look at for risk assessments.
[00:12:41.020]And I don't, won't go into details here,
[00:12:43.040]but what you see are a number
[00:12:45.250]of different drone applications
[00:12:48.050]and those blue lines that you see swirling
[00:12:50.520]across each of those four figures,
[00:12:52.118]are the airflow coming
[00:12:55.150]off of in this case at multi-rotor drone.
[00:12:59.080]And you can see as we change the application speed,
[00:13:02.646]we change different practices
[00:13:06.080]in that application in this example,
[00:13:08.759]two meter, three meter, five meter, a boundless height.
[00:13:15.620]You can see how much that air flow dynamics changes
[00:13:18.534]as we're spraying or releasing spray into that.
[00:13:21.580]You can imagine how difficult that then becomes
[00:13:24.030]to get those droplets to the ground in a uniform way.
[00:13:30.770]So we'll just skip forward here.
[00:13:33.570]This is us modeling done by Milton Teske and his group.
[00:13:37.450]This is another multi-rotor air flow dynamics team.
[00:13:42.140]And you can see that air flow coming off that drone.
[00:13:46.100]Now, when it's in flight and the prediction
[00:13:49.530]of what that release would look like,
[00:13:51.730]and you can see,
[00:13:52.860]just imagine if we're releasing droplets into this,
[00:13:56.410]where those droplets might go, and it's all over the board,
[00:14:00.160]which leads to those high CV values
[00:14:02.210]and very imprecise applications that we're talking about.
[00:14:06.160]So once we start to think about the future of our program,
[00:14:09.700]now we're gonna be doing a lot of work looking
[00:14:11.430]at field efficacy, pairing different nozzles
[00:14:15.030]and product combinations,
[00:14:17.070]continuing to do work on deposition
[00:14:18.860]and swath width,
[00:14:19.910]as you can see,
[00:14:21.480]we're nowhere close to what a ground
[00:14:22.880]of aerial application would be.
[00:14:24.110]So the question becomes,
[00:14:25.690]what we do to start to make it look more like that?
[00:14:29.670]Definitely need to work
[00:14:30.720]on getting standard application parameters,
[00:14:33.215]including the heights and application speeds.
[00:14:36.450]And then the other thing that we didn't even touch on today
[00:14:39.270]is understanding pesticide drift
[00:14:40.800]and where these products are gonna drift to.
[00:14:45.170]So with that, I'm gonna wrap up,
[00:14:47.850]we'll open things up for questions.
[00:14:49.840]Thank you.

The screen size you are trying to search captions on is too small!

You can always jump over to MediaHub and check it out there.

Comments

0 Comments

The Imprecision of Using Drones for Precision Pesticide Applications

Description

Searchable Transcript

Comments icon comment

Related Channels

Comments