Drones provide bird’s eye view of fields


Issue Date: June 7, 2017
By Kevin Hecteman
Khai Chang, left, and Freddy Lopez look over their drone during a University of California Cooperative Extension field day at a Bowles Farming Co. cotton field in Los Banos. The group from California State University, Fresno, is working with UCCE farm advisor David Doll and students from UC Merced to investigate ways farmers can use small drones to assess their fields and orchards, particularly to monitor for water stress.
Photo/Kevin Hecteman

You're responsible for hundreds of acres of orchards. How will you keep track of the farm's water needs? You could drive or walk the field, or you could send up your trusty drone, and analyze the photos and data it brings back from a 10-minute flight.

That's the idea behind a multiyear project in which David Doll, a University of California Cooperative Extension farm advisor in Merced County, has teamed with professors and students at UC Merced and California State University, Fresno, to investigate ways farmers can use small drones, also called unmanned aerial vehicles or UAVs, to watch over their fields.

The project has five objectives: reducing equipment costs; determining the proper flight angle and height for different crops; using the drone for data collection on a farm; developing user-friendly hardware and software; and sharing the knowledge with farmers and seeking their feedback.

UCCE hosted a field day at Bowles Farming Co. in Los Banos last week, with Doll and teams from the two campuses in attendance.

"We want to provide farmers awareness that, although this technology is out there, it's still new and we're still doing some research alongside with it," Doll said, " as well as showcase some of the technology that we have in the field and how we're able to use it to collect data on the fly from sensors, as well as take images."

He said researchers are also developing new vegetation indexes to provide a better sense of crop stress in almonds, walnuts and pistachios. Doll said many systems use the Normalized Difference Vegetation Index, which works well for annual crops such as corn but has come up short on tree crops.

"Within perennial crops, NDVI is not able to detect mild and moderate stress levels due to the canopy density and tree architecture," Doll said. "These are the levels of stress in which the plant's water status needs to be maintained and managed. It can detect severe stress, but by this point crop load and next year's crop load is negatively impacted."

The Fresno State crew used a drone to collect information from sensors placed throughout a 9-acre cotton field.

"Out in this field, we have 20 sensors spread out," said Freddy Lopez, a Fresno State graduate student. "They're taking temperature and humidity readings and moisture samples."

The sensors can be programmed to take readings at set intervals, he added. The drone hovers overhead, wakes up the sensor, collects its data via Wi-Fi and saves it to a memory card.

"We have a few programs that will plot the data over time and show you what the moisture was like out in the field," Lopez said. "We can detect overwatering. Hopefully, we can detect underwatering before it becomes an issue."

Lopez described the project as both practical and research-based.

"We're trying to tune in our project, see how much battery life we can get out of each module, see what kind of wireless range we can get out of each node, see how fast we can collect the data, and honestly as engineers also learn a little bit about farming as we develop tools for farming," he said.

The UC Merced drone looked for signs of water stress.

"I'm working on trying to extract the water status or early water stress detection for crops using drone imaging," said Tiebiao Zhao, a doctoral candidate who piloted his team's craft.

"I flew two cameras," Zhao said. "One is a red-green-blue camera. The other is a near-infrared camera. We flew at 60 meters (almost 197 feet) with the resolution of 2 centimeters per pixel, and then we covered the whole cotton field within 10 minutes."

The resulting photos can help a farmer identify patterns, Zhao said, and figure out where problems might be occurring with water, soil or disease.

Per federal regulations, the drone pilots kept their crafts within their line of sight while they were airborne. Commercial drone pilots need a remote-pilot certificate, which requires passing an aeronautical knowledge exam.

The prevalence of drones has also placed emphasis on the need for coordination with manned agricultural aircraft. Though there have been no reports of midair collisions between manned agricultural planes and drones, pilots have reported a few close calls. Reid Potter, manager of Lakeland Aviation in Corcoran and president of the California Agricultural Aviation Association, said low-flying agricultural pilots won't be able to see a drone unless it's properly equipped.

"If someone's going to be operating a drone, that vehicle has to have some sort of beacon, flashing beacon, so that any pilot can see it," Potter said, adding that radio communication would also help (see text box).

Doll and his colleagues are looking for ways to make UAV systems easier to use and easier to afford.

"We've been trying to identify a low-cost platform, such as a cheaper drone or low-cost drone, low-cost imager using just a point-and-shoot camera, and then trying to utilize that and develop the technology to process those images into something a farmer can actually use and interpret and apply to their field," he said.

As for the next step, "what we're looking to do is to work on that post-image processing," Doll said. "We know that's a major challenge for many farmers. The technology is there to georeference, to geotag, to take images, to stitch the images, but where we're struggling is to put that all into one package for the farmer to utilize. We want to simplify this process."

One company trying to do that is Air Tractor, a Texas-based builder of agricultural planes. Last year, Air Tractor bought drone startup Hangar 78, maker of the Yield Defender. The system includes a small quadcopter drone, a near-infrared camera and processing software.

"This is the first growing season we've actually had the analytical tools to provide the turnkey system," said Wes Hall, vice president of UAS technologies for Air Tractor. Hall said the processing software geolocates each image, pinpoints its location on a satellite overlay and processes into NDVI.

The great unknown, he said, is who will fly the Yield Defender.

"The market hasn't really decided who those people are," Hall said. "Is it agronomists? Is it seed companies? Is it chemical companies? Is it the grower? We don't really know right now."

(Kevin Hecteman is an assistant editor of Ag Alert. He may be contacted at khecteman@cfbf.com.)

Safety tips for drone pilots 

The National Agricultural Aviation Association recommends the following for drone pilots:

  • Become certified and well-trained in operating a UAV (mandatory for commercial operators).
  • Contact local agricultural aviation operators before flying by consulting www.agaviation.org/af_AerialDirectory.asp.
  • Equip a UAV with a tracking device, such as ADS-B Out, to send a signal to agricultural aircraft with similar tracking-reading technology.
  • Equip UAVs with strobe lights.
  • Give the right-of-way to a manned aircraft. It's the law.
  • Land the UAV immediately when a low-flying aircraft is nearby.
  • Carry UAV liability insurance.
  • Review the NAAA UAV safety video and other information at www.agaviation.org/uavsafety.

Federal Aviation Administration regulations concerning commercial drone use are summarized at www.faa.gov/uas/media/Part_107_Summary.pdf.

Permission for use is granted, however, credit must be made to the California Farm Bureau Federation when reprinting this item.