Our dealers have been out in the field all summer demonstrating the sophistication and speed of the AgEagle flight platform. There's never a shortage of ooh's and aah's when we launch the AgEagle wing and people see how quickly it can gather high-resolution imagery.
But all UAV systems for agriculture only become truly amazing when they let you see things from the air that you could not see from the ground.
When we fly imagery for a grower, they always see something that captures their attention. They often immediately point out things that they know on that field and see how obvious it is in an aerial shot. But even more interesting are those instances where they see something from the air that they hadn't seen from the ground.
This week, we were invited to a demonstration of canola genetics at the Viking Colony in Alberta, with the entire width of a quarter section planted to different genetics. The yield and other details will be captured this fall as information to be combined with other sites for use by producers across the region. But what really captivated our attention were the excellent and interesting fertility demos. The plots were designed to show the effect of different nitrogen rates (from zero to over 200 lbs of nitrogen per acre) and the effect of seed-placed phosphorus fertilizer (up to 50 lbs/ac).
The impact of both was visible from the ground. The zero-nitrogen check is on the right in the foreground, with higher N applications in the background. On the left are the seed-placed phosphorus plots -- rates increasing towards the back (west). The effects are clearly visible while on the ground and the demo plots certainly provided the desired opportunity to learn about fertility.
But we were on-site to demo our UAV's - so we took another look from the sky. First, we used the DJI Inspire 1 with a standard visible-spectrum camera. The effects seen on the ground are even more obvious from the air. Look at the impact the seed-placed P had on plant counts (1.25 plants per square metre instead of the standard 7 plants for the rest of the field). That plot at the southwest corner is still well into flowering in early August!
We also flew the entire quarter section using the AgEagle RAPID Carbon Pro (in 29 minutes). The different genetics on other parts of the field were visible not only through differences in NDVI (normalized difference vegetation index) values, but also through height differential visible in the digital surface model generated by the DroneDeploy software platform.
The entire site was well managed, so variability was not high given this year's crop circumstances (little spring moisture). But by teasing out some of the variability (NDVI values 0.55 to 0.75), it is possible to see even more using NDVI imagery:
It became clear that there are some site impacts on the west side of the plots, likely from microtopography and/or soil (further investigation would be required). We also noted the counterintuitive NDVI value for the plot in the far southwest. It has low plant counts and will clearly produce inferior yield, but shows up as the darkest green in the NDVI imagery. The delayed maturity means that it is still flowering and more actively photosynthesizing than the more advanced, stronger crop. While earlier in the season, high photosynthetic rates are usually positive and indicate good crop, that is not necessarily the case at this point in the season.
What this has reinforced for us:
aerial NDVI imagery is incredibly useful at determining where crop is impacted
unless the restricting factors are well known and controlled,
two things are still required to fully understand what the impact is:
Many thanks to the organizers of these trials (Viking Colony, Canola Council of Canada, and DEKALB) for another great learning opportunity! Special thanks especially to Paul & Michael Wipf for inviting me to the event.
Interested in seeing more of the imagery captured at demos this summer -- click here.