Hey, it's Greg from Scholar Farms, and I'm coming to you from Hawaii, where I'm here doing some mapping of Kona coffee. So we're in the Kona region. I'm wearing my Alloha shirt. I traded in for my standard flannel. That way, I can blend in with the retirees here on the island. But I want to walk through and use this time as a tutorial for mapping a high value crop like Kona coffee. And if you don't know about Kona coffee, it's one of the gourmet coffees in the world. It retails from anywhere from 30 dollars up to 75, 80 dollars per pound. And so it's a perfect example of where drone data for that high resolution on demand imagery can really pay off. So I want to walk through the sensor, the drone, the mission planning and then the processing really quickly for you. So here we're using the Micasense Rededge camera. This is the Rededge MX and I've mounted it to the DJI M200 series. And so there's a mounting kit that you can get from Micasense and it provides power to my Rededge. The Rededge camera is a five-band multi spectral camera. So those bands are red, green, blue. So you can get your standard RGB imagery or color imagery. And then it also has red edge or near infrared. And with those bands, we can combine those and we can get our vegetation indices like NDVI or NDRE or any other number of the dozens of indices that are out there. And we can start correlating those indices with the productivity here of the coffee out in the field. So I mounted it to the M200 so that I get a slightly bigger platform. It's flying for around twenty five minutes or so with the TB55. Those bigger batteries that come with it. It also allows me to do vertical takeoff and landing, which in kind of the variable terrain of coffee can be important. So once I've mapped the entire farm with the Rededge, I might select core target areas then for the Micasense Altum sensor. And this provides me those same five bands that I just mentioned, as well as a lower resolution thermal. And I'll talk about the thermal in just a bit. But the five bands are much higher resolution. And so if I want to do more individual plant measurements, then that higher resolution actually allows me to do that. The thermal camera is actually pretty interesting here in Hawaii as well as other tropical areas. And so there's a lot of variability in temperature as well as moisture on the island. So as we go down in elevation or up in elevation, we get changes in temperature. Its pretty drastic. And so those changes can really affect the coffee, both the phenology, the productivity, a lot of different aspects of it. It's also much wetter higher up. So you get more of the rain shadow here from the volcano, but you get more of the rain coming over from the wetter side of the island at higher altitudes. And so that wetness also can affect the temperature and thus the coffee. So having the Altum sensor is actually very useful. What is also useful is to account for a lot of the variability in lighting conditions here on the island. In fact, the reason I'm shooting this video is that I'm waiting for the shadows, that equatorial sun right now early in the morning in it's low down. The shadows are actually pretty long and they're overcast on my plants, the lighting conditions also change throughout the day. Earlier in the day, it's pretty sunny and there's no clouds in the sky. But come around lunchtime, the clouds are going to start rolling in and they're going to cast shadows on my map. Also, I might have a sunny day versus a cloudy day. Back to back. That's pretty common. So we use the DLS2. This is a sensor that comes with both the Rededge and the Altum, and it has light diodes on it that calibrate the imagery by capturing how much light is coming in from the sun while I'm flying my mission. That way I can capture from a sunny day to a cloudy day. I can calibrate my map so they're radiometrically accurate. Some jargon that means correct amount of light is bouncing off of my plants and I can compare apples to apples or coffee to. coffee, in this case. And so I can use that light sensor when I'm flying. And then I also use a calibration panel. So the calibration panel is kind of backup. You can use this or you can use the light sensor. And what it is, is a white painted panel. Hold the drone over my panel and I can shoot a picture before I fly. And then I can use the amount of light that's coming off of this panel. And I know in each of those five bands the exact amount of light that's coming off the panel and I can calibrate my maps using the information for my panel as a baseline. So we've talked about the sensor. We've talked about the drone. Let's talk about mission planning. So mission planning in coffee environments can be a challenge. So coffee is often grown in steep terrain and Hawaii is no exception to that. But if you go to Mexico or Colombia or Rwanda or Indonesia, any of these coffee growing regions, it's pretty steep country. And so for drone flying, you want to make sure that you're maintaining your elevation above your plants at a standard elevation. So for that, we need to account for those differences in terrain. And so in this case I'm using the Map Pilot app by Maps Made Easy. And I'm accounting for the terrain using SRTM data. So every 30 meters on the earth we have a point value that gives us accurate terrain information. We can upload that terrain information to our automated flight or lawnmower pattern and the copter will then adjust for those terrain differences and keep our pixel sizes on our maps the same. I'm using timing triggering on the camera. Actually, I set my target altitude, my flight speed, and it will tell me how many seconds I need to trigger. In this case, I'm triggering it about every one and a half seconds or so for this particular mapping. Then comes the processing side, working in coffee areas. Often there is pretty poor internet. That's where having local desktop processing can be important. So here I'm processing with Pix4Dfields, Pix4DFields is actually a nice rapid processing tool that takes both RedEdge and Altum data and I can process locally very quickly on my computer. I can generate both color and index maps as well as zonation maps. And those zonation maps are really nice to show the local growers or the farm managers to look at core target areas that they're managing. So one of the major aspects that we can use drone data for is looking for disease. And one of the disease problems that happens here in Hawaii are nematodes. These are a little parasitic worms that get into the soil and into the roots of the coffee plants. And it causes this kind of spindly shape or so that you can see right here in this particular plant. It's stretched out. We don't have a lot of leaves in there and it greatly reduces the production of this particular plant. And so there are major issue. And so managing nematodes is a core focus here of using the multi spectral data, identifying areas where you have this spindlyness. You can really see the lack of foliage that shows up in your multi spectral camera and then really accurately estimating both the amount of damage or attack that might be on the farm and then whether it's spreading or not, as well as how well the management is working. So that's disease. The other aspect then is going to be nutrients. So in particular, when we think about nutrients, there's not a lot of nutrient retention in these tropical volcanic soils. And so it's basically a rock or lava and the depth of the soils varies considerably across the farm. And you can really pick that up in terms of how they retain their nutrients and how much amendments needed to be added to the particular management sounds. That's where those management zones come into play is really accurately depicting where areas of fertility are high or low and how to accurately manage your amendments. The plants are actually pretty stressed right now. It's the end of the picking season. So you can see that the pickers have come through and they stripped most of the cherries off and you can tell by the chlorosis that's happening in the leaves and that's mostly a nutrient issue. They've been shunting all of their energy into producing coffee cherries or beans at the end of the season or into the picking season. The plants look like this. They're pretty stressed out. And we can see that stress in the digital data. You can see changes in photosynthetic activity or in INDVI for example, you have lower values of NDVI. When the plants are stressed out like this. And so when we map, we're able to come through and accurately estimate different areas of the farm and the level of stress on the farm for nutrients. And think about nutrient planning for the next growing season or in the next crop. It comes about in another six months or so. So a final aspect that I want to talk about for coffee is thinking about plant size. So every six to eight years or so, we're coming through and we're cropping or pruning the plants and then the plants regrow with new sprouts. We can accurately measure the density of the plants, the size of the plants, and then also that regrowth by flying repeatedly throughout the year. So that's a very practical application of using the digital drone data. That's just a quick walkthrough on using multi spectral cameras such as the Micasense Rededge or the Altum sensor for mapping high value crops and some of the returns on the investment for precision drone data. So I'm Greg with Scholar Farms coming to you from Hawaii. Beautiful location. You can see the ocean for miles behind me and I'm going to get up in the air, start mapping and then head to the beach and have a Mai Tai. So we'll talk to you again soon.
