Using Drones To Increase Farming Yields

Virginia Tech Researchers


Researchers at Eastern Virginia Agricultural Research and Extension Center in Warsaw are using drones to collect aerial imagery with the goal of helping farmers to grow small grain crops more efficiently.

Eastern Virginia AREC is one of Virginia Tech’s eleven research centers located across the state which were formed to facilitate collaboration among the Virginia Agricultural Experiment Station at Virginia Tech, Virginia Cooperative Extension, and agricultural industry partners. The AREC in Warsaw focuses on researching the growing of small grain crops and soybeans and breeding improved varieties. The funding for the drone program in Warsaw is from the Virginia Agricultural Council.

Joseph Oakes, Superintendent of the Eastern Virginia AREC, oversees the small grain and soybean crop research done on over 200 acres of cropland. He is experimenting with drones to improve data collection in the field and to make decisions regarding when to apply fertilizer and how much to apply. Assessing fertilizer needs in a more precise way increases yield, reduces manual labor, and reduces the amount of applied fertilzer, which saves money and the environment.

For the past two years, Oakes has been mainly focusing on developing ways to use the drones to optimize the production of wheat.

Oakes explained how wheat must be managed carefully throughout the growing season to optimize yields. The plant has a main stem with several tillers. Each tiller has the potential to produce a grain bearing head. Research has shown that the ideal number of tillers is around three. If there are too many tillers, the heads are too small and just end up competing with each other. A plant which produces two to four tillers has the best yield.

Wheat is planted late October through early November. “Early on, you want to get those tillers established,” said Oakes. “Ideally, you want to produce three or four tillers by January 1st. January and February can be a key time to apply nitrogen. Research has shown that the best way to decide when to apply fertilizer is to count tillers. If there are less than 50 tillers per square foot, you need to apply nitrogen.”

Obviously, measuring square feet and counting tillers in the field is very labor intensive, especially since tiller counts need to be taken five or more times throughout the growing season. It became a little easier when scientists found that they could estimate tiller density by calculating something called normalized difference vegetative index (NDVI). Researchers use a handheld device which assesses plant density using red and infrared light. The device measures the amount of each type of light reflected back from the plant. Healthy green plants, which are more dense, absorb most of the red light and reflect back more infrared light. Since there is a strong correlation between plant density and tiller density, measuring NDVI can be used to determine nitrogen needs.

Having these sensors reduces labor somewhat, but Oakes realized they could use a multi-spectrum drone camera to collect the same information more efficiently. The researches first used drones to collect data from their small 50 acre test plots last year. This year they are testing it on larger fields at the Eastern Virginia AREC and at four volunteer farms in Essex, Dinwiddie, Prince George and Suffolk counties. They are also still collecting the manual data so they can compare the feasibility of the two methods.

Oakes flies the drone over an entire field to collect the NDVI data information. The first step is to program a flight plan for the drone to follow. The camera takes 400-500 overlapping photos of the field. They then use a software program to merge the pictures into a composite photo. This technique enables them to fly the drone at a much lower height.

After the composite is produced, it is analyzed using software which helps them determine when, where, and how much fertilizer to apply. “One thing to keep in mind is that the majority of the work is done after the flight,” Oakes said.

There are other applications of the drone technology which are useful to the plant breeding programs at the center. Oakes said they are very close to being able to use the data to determine plant height. It’s important to know the height of a plant because if it is too short, the yield is low and if it’s too tall, it can lodge. Already they are able to determine from the photos if any plants have lodged, which reduces a lot of time spent walking in the fields.

The center is also in the early stages of looking at things like maturity dates. For example, soybeans mature over a long window of time from late September through early November. Right now they have to walk through thousands of different breeding plots several times a week to determine the maturity date of each different line. “We’re going to be able to collect some of that data with the drone,” said Oakes.

Eastern Virginia AREC is working with computer scientists at Virginia Tech to write software for practical applications of the drone data which will be helpful to farmers. They are developing a shape file which can be read by other software programs so that a farmer can do things like just plug an SD card into his equipment that will tell the sprayer what to spray.

Oakes said they should be ready to release some of the applications to the public in the next year or so. He noted that there is a cost involved and it may only be worthwhile for farmers with several thousand acres of crops.

Oakes said that there are several startups who plan to fly drones to collect data for farmers, but stressed the importance of having someone who knows how to properly analyze the data.

“After you fly and collect the images, there’s quite a lot of work. It takes someone who is trained to do that,” Oakes said.

Leave a Reply

Your email address will not be published. Required fields are marked *