Precision farming: How farmers benefit from drones in plant breeding

The future of breeding is up in the air, among other things: KWS uses drones over its fields in order to produce new varieties even faster. The pictures taken by their cameras are analyzed on a computer to permit precise findings about plants’ growth. That helps the company offer new seed to farmers sooner – a great advantage for agriculture.

The air’s vibrating. It sounds as if a swarm of bees is coming. For a few seconds the four rotors on Marius Burkhardt’s bright-orange flying machine cause so much turbulence on KWS’ trial plots in Einbeck, Lower Saxony, that the plants on the ground bend and sway in the airflow.

Then the drone climbs to lofty heights and, much more quietly, begins flying mathematically precise paths over the field where tender corn plants a few centimeters high are growing. They are the reason why the 40-year-old Burkhardt is flying the quadrocopter over Einbeck this morning.

Several sensors are suspended from it, scanning the plants on the field so as to ascertain their growth. The drone systematically records an area of between three and eight hectares on one full battery charge. That’s enough for almost half an hour in the air, depending on the application and the altitude, sensors and flight plan. A multispectral camera and a thermal imaging camera measure the amount of light reflected and absorbed by the plants in various wavelength ranges. The photos of the field are also used for evaluation.

  • 1 / 3

    Ready for takeoff: Drones will be used more frequently in breeding in the future, since they enable precise findings on plant growth.

  • 2 / 3

    The drone flies at a constant height of 25 meters. As a result, the software can determine the exact GPS position of each plant.

  • 3 / 3

    A watchful eye: The drone’s cameras record every plant and permit conclusions on its growth and any diseases.

Drone and image analysis: aids for better breeding

The expert Marius Burkhardt monitors whether the photos are taken from a constant altitude of 25 meters. Only then can the software later determine the exact GPS coordinates and create a 3D model of every single plant.

This combination of technical innovation and breeding experience enables new disease-resistant plant varieties to be bred in a shorter time. After all, the goal is for the seed to produce plants with high yields, even under the toughest farming conditions.

“Farmers want to use seed that precisely meets their individual requirements,” says Dr. Christoph Bauer, who is in charge of developing digital phenotyping technology at KWS and coordinates the drone flights and evaluation of the data obtained from them. The new digital tools developed in-house at KWS support breeders in picking the most suitable plants for their work from among the hundreds of thousands of possibilities. “All this helps our breeders identify the best plants,” says Bauer. “Drones and image analysis are increasingly valuable aids.”

Precision farming: greater accuracy in agriculture

The method behind it is as old as plant breeding itself: Phenotyping delivers answers to questions such as whether new varieties cope well with the environmental conditions at a location. Or whether and where they are infested by pests.

“Throughout KWS’ more than 160-year history, its experts have assessed plants in the field,” says Bauer. “That’s vital in developing new varieties.” Breeders sow new plants every year and then evaluate how they develop in the field. They record the plants’ size, color, growth rate, the number and shape of their leaves, and other features. In many small, often very laborious steps, breeders can thus ultimately supply varieties tailored perfectly to the farmer’s needs.

The aerial photos and software analysis speed up that process significantly. And they have more advantages for farmers as well, explains the technology expert Bauer. Fungal infestation, dry areas, low chlorophyll content: Nothing goes undetected by the digital sensors. And because the measurement devices and software never get tired, the results of analysis of the plants’ rate of growth and the number and size of their leaves are often more precise and available faster than without technical assistance.

As Bauer puts it bluntly: The human eye is not accurate enough. “If several people walk through a field to assess development of the foliage, you get results that are usually very similar, but that nevertheless have slight differences.”

Technology is always objective and helps rule out those differences – and also has the advantage of speed: “No human can assess such huge areas with such unerring objectivity.”

  • No human can assess such huge areas with such unerring objectivity.

    Dr. Christoph Bauer, Chief Engineer Phenotyping Technologies KWS
    KWS_drohne_pflanzenzuechtung_dr_christoph_bauer_leiter_technologieentwicklung_buero_portrait.jpg

Hundreds of drone photos, one big picture

Despite drones and software, plant breeders are still indispensable. First of all, they have to calibrate the measurements: “Of course, it’s easy for us humans to tell two beet or corn plants apart,” says Bauer. “However, transferring this knowledge to machines is a laborious task.”

An example: The sun is especially low over the fields in the morning and evening and its light is then more reddish than at midday. At 12 noon, however, the sun is very high in the sky and there is a greater percentage of blue light. If there’s dust on the leaves in the morning on dry days, but it’s washed away in the afternoon by a downpour, the plants’ appearance changes again in the space of a day. Or shadows are cast differently in the morning and at noon. “We have to take all of that into account and make adjustments,” says Bauer. “Only then can we accurately identify the state of the plants and leaf diseases.”

And people are also vital when it comes to assessing and using the results. After returning from the field, drone pilot Burkhardt feeds the images into the analytics software. It combines hundreds of photos to create one big picture. Fine green, red and blue lines extend across the virtual field on screen. The team then discusses the data: What is the mean leaf coloring? What area do the plants cover? And what does that mean for further breeding work? The aim here is to identify concrete successes and challenges.

  • 1 / 3

    Despite all the technology: The human eye is vital in analyzing the recordings. Experts like Christoph Bauer are responsible for drawing the right conclusions for breeding.

  • 2 / 3

    Fungal infestation, dry areas, low chlorophyll content: Nothing escapes the digital sensors as they scan the sugarbeet.

  • 3 / 3

    The photos from the drone are loaded into special software and analyzed on the computer.

Interdisciplinary research into phenotyping at KWS

KWS’ interdisciplinary working group that deals with the drone photos includes several experts in the field of image analysis. Christoph Bauer, a physicist, leads the group. His employees and their technologies are deployed at breeding stations worldwide and have already helped gather a lot of positive experience: “We combined our strengths coming from the input of breeders, biologists and engineers,” says Bauer. “That was the reason for our joint success in analyzing the data.”

  • At KWS colleagues from a diverese field of disciplines work together.

    Marius Burkhardt, KWS Drone-specialist
    KWS_drohne_pflanzenzuechtung_marius_burkhardt_drohne_zusammenbau.jpg

Yet plant breeders will still need to work in the field: They have to check the conditions for the measurements on site and derive conclusions for breeding from their personal observations. “We’ll never make humans superfluous,” adds Christoph Bauer. “But the technology we’ve developed will assist people in making breeding decisions.”

KWS has a large research budget for precision farming

Drones now fly over the fields of more and more farms. Precision farming helps save fawns and moles from harvesters, detect impending aridity at an early stage and chart molehills as rough terrain so as to draw conclusions about potential harvest losses. KWS is constantly optimizing the complex system of drones, sensors, software, image recognition and servers it uses in plant breeding. The research budget in fiscal 2016/2017 was around €190 million or some 17 percent of net sales.

Back to the "World of Farming" main page
World of Farming

Common icons/confirm-email