UAVs team with artificial intelligence to boost crop scouting efficiency
Wading through crop fields searching for insects, diseases, weeds, nutrient deficiencies, uneven emergence, and other maladies consumes time and effort.
“Even when the corn is just knee high, you can only see a couple hundred yards in each direction,” points out J.D. Bethel, an agronomist with Integrated Ag Services (IAS), Milford Center, Ohio.
This makes it almost impossible to see emerging weeds like giant ragweed that quickly become, well, giant. IAS aims to nix this scenario and others by pairing artificial intelligence developed by Taranis with flights of unmanned aerial vehicles (UAVs) during the growing season.
Taranis officials say its AI2 SmartScout captures 0.3 millimeter per pixel resolution from UAVs at a speed of 100 acres in six minutes. In comparison, the best satellite resolution is about 1.2 meters per pixel, says Mike DiPaola, Taranis general manager of North America and vice president of global sales.
“It can easily identify a bean leaf beetle or a Japanese beetle on a soybean leaf,” says Bethel. “We have even been able to count the hairs on a soybean leaf or the colors of the flowers on soybeans. That’s the sort of resolution it can achieve.”
Confirmation still is required, of course. “You still want to go out and check if it is indeed a waterhemp plant that the program has identified,” says Bethel.
Even when making a field visit, this technology boosts scouting efficiency, according to IAS and Taranis officials. Taranis software also contains a feature that farmers and agronomists can use to prioritize field visits.
“There may be only 15% of fields that they [agronomists and consultants] need to immediately visit,” says Evan Delk, IAS vice president of sales and marketing. “If they’re instead trying to get across every single acre, their time is not being utilized as it should be. Our consultants need to be in front of the grower, helping them make better decisions.”
Josh Guy with Integrated Ag Services readies an unmanned aerial vehicle for flight.
Keying all this is artificial intelligence (AI) developed by Taranis. Its image bank contains more than 50 million submillimeter high-resolution images of crop disease, insects, weeds, nutrient deficiencies, and other issues compiled by more than 100 agronomists. Through its AI engine, Taranis leverages machine learning and computer vision to help farmers and consultants identify field maladies.
“For example, we will take pictures of a Japanese beetle and run them however long it takes for the AI to see a pattern,” says Ofir Schlam, CEO and cofounder of Taranis.
Once the computer records the pest or malady, though, it remembers it.
“What’s great about artificial intelligence is it doesn’t think like us,” says Josh Guy, IAS operations manager. “It may detect soybean diseases in a field that human eyes may not see.”
Imagery captured by unmanned aerial vehicles (UAVs) enables maps to be made that monitor crop emergence or emerging weeds deep in the canopy.
How It’s Used
IAS offers early-season and late-season scouting packages for $9.75 per acre, while a full-season package costs $13.50 per acre. A full-season package is the best way for farmers to monitor their fields, says Delk. In the full-season package, IAS flies UAVs across fields about every 14 days, depending on weather and crop growth progression. This service also provides an aerial overview video of each field.
“Whether you’re scouting on foot or with a drone, you have no idea what is happening after you leave the field,” says Bethel. “There can be weeds coming up, plants dying from disease, or plants still emerging. Using the IPM [Integrated Pest Management] approach, we scout every two weeks with a drone looking for weeds that may influence changes to the existing herbicide program.”
The UAV and AI combination can help a farmer decide whether or not to apply a fungicide, while nutrient scouting can influence whether to apply late-season nitrogen, says Delk.
“The idea is to constantly have eyes on the field,” he adds. Cost savings from making or forgoing a chemical application or late-season nitrogen pass or seed savings from a selective replant (see “Easier Replant Decisions”) can quickly surpass the $9.75 to $13.50 per acre cost, he adds.
Umanned aerial vehicles can provide images that enable farmers to quickly make decisions.
Some 12 to 24 hours normally pass between a drone flight and the time maps are digitally delivered to a farmer’s desktop computer or mobile device, says Guy.
“We plan flights in advance so once we get out to a field and set up, it’s as simple as hitting play on the flight plan,” he points out. “We still need to keep eyes and hands on the controller, but it is basically a preplanned flight with the UAV.”
Challenges exist. “One of the major impediments in getting good imagery is wind speeds,” says Guy. “Technically, this equipment can fly in winds up to 25 mph. Once you get past 10 to 15 mph though, the crop moves just enough for the camera to pick up that motion and blur the image.”
So much information is collected that it can be overwhelming for the farmer.
“The important part of getting the value out of the data is to make sure you have a trusted adviser to make sense out of it,” says Delk.
“Farmers are busy,” adds Bethel. “We can text a farmer with a report that says, ‘These four fields look good, but you really need to look at field five.’ This is a huge time savings for them. They can better allocate the amount of time they do have for more important tasks.”
Easier Replant Decisions
Replanting is where unmanned aerial vehicles (UAVs) teamed with artificial intelligence particularly shine, says Evan Delk, vice president of sales and marketing for Integrated Ag Services (IAS).
“Before, we went out in the field to do five plant stand counts in a 100-acre field,” he says. “Now, we take high-resolution images [with the UAV] every one-half acre, which creates many data points that the farmer can use to decide whether to replant.”
“It takes a lot of the emotion out of the replant decision,” adds J.D. Bethel, IAS agronomist. “Instead of driving back and forth through the whole field wondering where they need to plant, the map shows them the worst parts of the field where they need to replant.”