Despite FAA dithering, a drone economy sprouts on the farm

September 16, 2014, 10:31 PM UTC
Farming alfalfa harvest
Farming alfalfa harvest
Photograph by Thorney Lieberman—Getty Images

In June, the oil and gas giant BP (BP) began flying drones over an oil field in Prudhoe Bay, along the northern coastline of Alaska. It was the first commercial drone operation over land in the United States that was approved by the Federal Aviation Administration.

It’s a small step. There remains a long road toward integrating commercial unmanned aerial systems, known as UAS but more widely referred to as drones, with the U.S. national airspace. Progress has stalled for developing a new set of FAA regulations that would open up American skies to the devices. On June 26, an audit by the U.S. Department of Transportation found that the FAA will miss its September 2015 deadline to accomplish the task.

For most proponents of commercial drones, the waiting has become interminable. The BP exercise was by a single company over a small sliver of sky in a far northern hinterland—a painful reminder that progress is sluggish at best. UAS technologies will unleash $82 billion in economic impact and create more than 100,000 new jobs in the years immediately following the legalization of drones for commercial use, according to the Association for Unmanned Vehicle Systems International, or AUVSI. Bureaucracy, drone supporters insist, is costing people money.

Down in the lower 48, the drone economy is developing more rapidly—albeit quietly, and in an altogether different industry than energy. Farmers and agronomists eager to make precision agriculture even more precise are developing and deploying a range of new UAS technologies—some off-the-shelf, some homegrown—to boost yields, battle common crop ailments, and drive overall farm efficiency. Growers and researchers are able to fly their drones without violating the FAA’s near-outright ban on commercial drone use. In doing so, they are fast becoming the first in the U.S. to realize the economic potential of the drone economy yet to come.

Drones on the farm

Robert Blair, the proprietor of Three Canyon Farms in north-central Idaho, has owned and operated various small unmanned aircraft since 2006, when he first saw an ad for a farm-focused UAS in an agriculture magazine. To Blair, the benefits of capturing on-demand overhead imagery of his fields were immediately apparent. Even gathering only the most rudimentary data—such as photos—Blair can glean a good deal of information from a single pass over his fields and reduce the crop damage he might inflict using a more conventional method.

“I’d have to go out and scout my fields anyhow,” Blair says, “but this allows us to obtain information during periods when you would not be able to physically go into the fields—such as after a rainstorm when it’s too muddy or when the crop is at a certain maturity when you would damage it by driving a four-wheeler out there. A UAV is easy to deploy and that info is back in your hands as soon as it lands.”

The on-demand aspect of a drone is particularly useful. Satellite imagery that is days or weeks old isn’t useful after a big storm or during a particularly dry spell. Hiring a conventional aircraft is expensive and often requires several days of lead time. “We just had a storm yesterday,” Blair says. “And I want information today.”

Several research institutions are hacking together new kinds of UAS from a range of off-the-shelf aircraft components and sensors in an attempt to beam new kinds of on-demand information from field to farmer. Researchers are custom-building drones that can deliver new kinds of data to growers, which they can then use to identify a range of agricultural problems.

Dennis Bowman, an agronomist with the University of Illinois Extension, has experimented with both the camera-equipped Parrot AR Drone (a recreational model available online for less than $200) and a Phantom quad-rotor (a slightly more sophisticated recreational drone manufactured by Hong Kong-based DJI). He has since constructed his own hexacopter drone from parts available online, equipping his creation with a custom-modified Canon Powershot camera that shoots imagery using infrared (IR) light. “For some of those things that would seem very daunting, I’ve found YouTube videos for every step of the process,” he says. “It’s kind of amazing.”

Capturing conventional aerial imagery can help farmers identify problem areas for closer inspection, but capturing infrared imagery begins to expose drone-driven agriculture’s real potential: the ability to acutely diagnose problems in a precise location. “When you apply herbicide at the wrong rate, pest problems, weed problems—these are the things you’ll see,” Bowman says. Which in turn help a farmer make better crop management decisions.

For decades, the term “precision agriculture” was used for agricultural practices that used GPS and other geospatial technologies to observe and manage variability in crops. Drones offer a more acute way to do that—and combined with data analytics technologies, could actually diagnose problems and recommend crop management strategies that farmers can consult.

For example, the Utah Water Research Lab at Utah State University recently struck a deal with a major California grape-growing operation to use drones to collect data on how much water the company is losing through evaporation versus plant transpiration. The information is consequential as water scarcity becomes an increasingly pressing issue along parts of the West Coast. Thanks to tools developed by the Utah Water Research Lab team, it’s not all the collaboration will produce.

“What we want to be able to do is to fly one day and within 24 hours deliver to the manager of the farm a daily schedule for the next seven days recommending how much water he needs to put down each individual drip lateral,” says Mac McKee, director of the Utah Water Research Lab. “We could tell them how much water to give to each individual grape vine, but they can’t respond at that level of detail. Right now they can’t even respond at the drip lateral because they lack the information to differentiate from one to another. We’ll be able to give them that detailed information.

“At the same time we’ll be able to tell them how much nitrogen fertilizer should be contained in that irrigation water. We’ll be able to give them a schedule for when we think they need to trim the grape leaves out of the vine canopy to provide optimum sunshine. And a whole series of other things.”

In other words, drones will soon be feeding growers more high-resolution data than they’ll know what to do with. It’s not some distant future: Pending FAA approval, the Utah State team will commence the California flights before the end of the year.

Harvesting data

In countries like Japan and Canada where governments have been more proactive in embracing drone technologies, UAS have been flying over farms for years. But those technologies aren’t quite geared for the kind of large-format farming conducted in the American Midwest: The average Japanese rice paddy, it’s safe to say, is quite a bit smaller than the average Midwestern cornfield.

“What we don’t have right now is the right products in the United States,” says Darryl Jenkins, a longtime aerospace industry consultant and analyst and co-author of the AUVSI report on the potential economic impacts of commercial drone integration in the U.S.

Analytics in particular are lacking. Today, imagery collected in non-visible spectrums like infrared can detect variations in plant stress across a given field, but tools to diagnose exactly what the data mean (let alone produce meaningful recommendations for growers) are not widely available. Farmers can see that crops in one corner of a field are under a different amount of, or different kind of, stress than plants in another area. Soon, thanks to research efforts like those conducted by McKee and Bowman and their academic peers, databases will exist that can tell farmers what is causing a certain kind of plant stress and recommend a fix—all without having a professional agronomist on hand to translate the data.

“It’s really hard, technically, to translate multispectral imagery into products that can be useful for supporting decisions that farmers and growers need to make,” McKee says. “Only in the past few years have we been able to bring the analytics far enough along to make a dent.”

The challenge has created something of an expectations gap for some farmers who want to use drones made or sold by 3D Robotics, Honeycomb, Precision Drone, and others to improve their resource efficiency and crop yields. Online message boards on forums like DIY Drones are rife with questions, answers, advice, and shop talk surrounding the use of drones in agriculture.

“When you start needing to make agronomic decision on the crops themselves, you can’t get that just visually, it takes a bit more sophistication,” says Mark Blanks, the UAS Program Manager at the Applied Aviation Research Center at Kansas State University. “People are buying stuff that they think is going to basically cook their dinner, and then they’re finding that the data isn’t as good as they thought it would be.”

The misperception suggests future opportunity. The Utah State Water Research Lab team led by McKee plans to spin out a business within the next 12 months offering drone hardware and software solutions to precision agriculturists. Fixtures of the drone hardware world such as DJI and 3D Robotics are poised to sell a many more aircraft systems if they can provide the solutions that farmers need. And software analytics, whether sold to farmers individually or provided as a service, are expected to become major agricultural service—just as soon as the FAA develops a set of rules that allows companies to sell drone-related services commercially. The transformation of both domestic agriculture and the U.S. small drone industry is already underway.

“We see it in the community,” he says. “There are a whole lot of startup companies that are going out and selling services or hardware to farmers. And some early adopters are moving into it. But I don’t think the return on investment has proven itself yet. We need to prove the return on investment on the data, and that’s coming soon.”