HENRIETTA, N.Y. (WROC) — Professor Jan van Aardt has received a grant — for $357,000 — from the United States Department of Agriculture for a drone that uses imaging sensors to determine the health of grape vineyards.
Van Aardt is part of a larger consortium led by Washington State University. This consortium worked van Aardt and his team in determining what this drone would need to sense to be an effective tool for the farmers. This partnership also extends to Cornell University in surveying and collecting samples.
The technology — hyperspectral imagining and LiDAR — has been been tested by RIT student Rob Chancia.
But as we do when we tackle these science stories, let’s first meet our expert, go over these terms, the technology, and how this will help farmers grow the best wine grapes possible.
Van Aardt is from South Africa, and after coming to the US to study forestry and imagery at Virginia Tech, he stared at RIT in 2008.
“I’m the one who uses remote sensing data to analyze forests,” he said.
Through this field, he is able to examine the health of a forest or farm, determine if harmful chemicals are in a nearby water source, or if plants are getting the right kinds of nutrients (for grape farmers, that would be nitrogen, magnesium, potassium, to name a few).
That blossomed into a love of growing a research solution into a practical solution. This often means taking incredibly complex technology — like this drone — into something that an everyday person can use and can benefit from.
“I work with two technologies,” he said. “Principally one is hyperspectral meaning many color channels. You and I see blue, green, and red; a hyperspectral sensor sees outside what you and I can as humans. See, but the catch is, it sees it does that in an many narrow channels, contiguous channels. So it really enables us to sense the entire electromagnetic spectrum.”
“(The other is) LiDAR: light detection and ranging,” he said. “So a laser that pointer that blinks really fastm, and generates a 3D point cloud of an object or a surface or a crop.”
How this science fiction works
Another key concept in imaging science is understanding how different material reflects and absorbs different colors of light.
“The green leaf is green because it reflects a little bit of green energy, but it’s a heavily absorptive in, in blue and red,” van Aardt said.
Van Aardt says that the hyperspectral imaging can detect visible and non-visible light in such a discrete measurable way, they can use this incredibly sensitive data to conclude information with just a picture.
“I’m just using a silly example, but if we find that 950 nanometers — which is beyond what you and I can see — is tied to nitrogen absorption or nitrogen content… We can now design a sensor that senses that specific wavelength or range of wavelengths to detect that absorption feature on a routine basis for nitrogen mapping,” he said.
How it helps the farmers
As we learned from van Aardt, his passion comes from making this technology accessbile, and easy to use.
“So many of the drones we fly are $250,000 and upwards per flight,” he said “The challenge to us is to distill it from that over-sampled spectral dataset, that expensive package, to something useful for a practitioner or a service provider to the industry.”
So while he and the team are working on making this technology more accessible, he says that once it is, it can be a massive benefit to farmers.
Currently, the best way that farmers (especially for grape farmers, more on that in a moment) can best determine the health of their fields is to employ what van Aardt calls a destructive process. Farmers have to take leaf clippings and soil samples to send to a lab, which harms the plants and soil.
It also takes a long time; they have to plot out a field randomly to make sure the sample is statically viable and a representation of the whole field, they have track down the plants, make the cuttings, package it, send it to a lab, then await the results.
“Then the chemical analysis starts and that takes sometimes a week,” he said.
Van Aardt that says that this time frame still can work for most farmers, as growing seasons are often long enough. But for grape farmers, they’re more on the clock. So when you have technology like this drone that can give them the same information in one day — a difference van Aardt calls “orders of magnitude” faster when you take into account the whole process — it can ensure the best grapes.
“But if you want that flavorful grape, you actually need to sort of exert a little bit of stress on the vine so that it works that extra little bit harder to extract moisture and flavor from the, from the soil,” he said. “So it’s a critical to a farmer, for instance, in that case to know exactly when, what the moisture status of a vine is (along) with nutrient count.”
When it will become commercially available
Van Aardt says that John Deere actually has a very simple version of this technology already, it’s a small sensor on the front of a tractor that measures “vegetation vigor, and health.”
“As you’re driving your tractor through the field, (the tractor) relays that information to the fertilizer or the spreader on the back, you can differentially apply fertilizer based on what the camera sends in front of the tractor,” he said.
While van Aardt didn’t give a specific timeframe for commercial release, he does say that once they bridge the gap of the drone’s “functionality and adaptability” by possibly creating adjustable filters, then it will be ready to go.
Remember that example with the nitrogen? Instead of a giant machine that detects everything at once, a farmer would be able to pick and choose what the drone needs to see, which will make it cheaper and more effective for the farmer.