Faster than people
Hitherto, crowds of human helpers are needed to harvest mature cauliflower, while young heads are still standing still.
This can soon be a machine. With the aid of hyperspectral cameras, it detects whether the vegetable is ripe, even though the head hides under cover sheets.
The “asparagus panther” of Wolfsburger agricultural machine builder ai-solution learns.
Researchers from the Fraunhofer Institute for Factory Operation and Automation IFF in Magdeburg, together with their colleagues in Wolfsburg, modify it so that it also reaped cauliflower.
That does not sound like a sensation. The machine needs only to drive across the field and cut off one cab after another.
But that would be a bad business for the farmer, because the vegetables are treacherous. The heads are mature at different times, even if they are immediate neighbors.
In order to prevent heads being cut off, which still need a few days on the field, harvesters must come. They lift the protective leaves, test the degree of maturity with an experienced look, and only cut off the heads that are ripe enough.
This is a personal intensive and therefore expensive affair. In addition, harvesters are harder to find.
Fine differences of the cover sheets VitaPanther is the name of the untiring harvester of the future.
It is equipped with hyperspectral cameras which record images of very many wavelengths, as well as the infrared and ultraviolet areas, which are hidden from the human eye.
The cameras record the reflected light.
We are using an effect that we have discovered in preliminary studies, “says Professor Udo Seiffert, Head of the Department of Biosystems Engineering at the IFF.
“The leaves of a mature cauliflower are biochemically different from those that envelop the immature heads.” This difference is recognized by the image evaluation unit on board.
If everything points to a mature cabbage, the machine pulls the knife and cuts it off.
Harvest takes many days Human harvest assistants need a week or longer to complete a field.
They have to do it again and again. The VitaPanther does the same.
Only with two differences: he does not get back pain and is also operational at night.
The mathematical model, which evaluates the images of the cameras, is based on algorithms derived from machine learning.
The researchers use the camera to record differently ripe cauliflower heads, which is simultaneously examined by a human expert. Over time, the software learns to distinguish mature from immature minds.
Accessories for all vegetable varieties The VitaPanther should not be the only offshoot of the asparagus panther.
“Later, other crop units could be added for further vegetable varieties,” says Christian Bornstein, Managing Director of ai-solution.
“Our goal is to build a module that can be adapted to the existing device.” The farmer would then come out with a single vegetable harvesting machine. In 2017 the VitaPanther will be launched.
Fertilize with drones and destroy them with robotic weeds Smart technologies are increasingly gaining ground in the management of fields.
Thus, a Chinese company has developed the world’s first drone for use in the field. It is supposed to cost 15,000 dollars and still become a hit.
And in the machine called Bonirob has a highly specialized image recognition system that reliably distinguishes between crops and weeds.
This has to be brought to the robot first, which is an elaborate matter.