Solar-powered robot fights weeds with AI and hoe

The robotics start-up Aigen has declared war on weeds on agricultural land. The company uses a solar-powered robot that recognizes weeds with the help of artificial intelligence (AI) and kills them with electrically powered hoes.

In agriculture, weeds in the field are usually controlled with herbicides. This is neither healthy for the soil nor for the consumer, who consumes the contaminated crops through food.

"Everyone eats food that has been sprayed with chemicals," says Richard Wurden, co-founder and CTO of Aigen, summarizing the problem succinctly. Wurden worked as a mechanical engineer at electric car manufacturer Tesla for five years. He then began developing a robot that could combat weeds in agriculture. He got the idea from relatives who have a farm in the US state of Minnesota, grow crops and complained about the costly weed control. After a while, weeds can develop resistance to weed-killing herbicides. Other, less dangerous methods, such as removing weeds with a hoe, are too expensive. There is simply not enough manpower to do this work. This leaves the use of chemicals, although their effectiveness decreases over time and the majority of farmers reject them.

Fighting weeds with AI and a hoe

The robot from Aigen is designed to solve the weed problem in an environmentally friendly way. Solar cells are installed on an autonomous, electrically mobile frame. They charge a battery that powers the electric motors and the electronics required to control the drive, cameras, signal transmission and weed removal. In principle, the wheeled robot imitates a person weeding.

Empfohlener redaktioneller Inhalt

Mit Ihrer Zustimmung wird hier ein externes YouTube-Video (Google Ireland Limited) geladen.

YouTube-Video immer laden YouTube-Video jetzt laden

Ich bin damit einverstanden, dass mir externe Inhalte angezeigt werden. Damit können personenbezogene Daten an Drittplattformen (Google Ireland Limited) übermittelt werden. Mehr dazu in unserer Datenschutzerklärung.

To do this, the robot moves across a field, also coordinated with other robots, and searches for weeds among the crops. This works via an AI that evaluates camera images using image recognition and identifies weeds. However, this is not done on the robot itself, but in a control center to which the robot transmits the image data wirelessly. This is supported by Amazon Web Service (AWS), which supported Aigen as part of the "Compute for Climate" program. Amazon's cloud offshoot uses this program to help start-ups dealing with environmental problems and provides them with AI tools, data centre capacity and technical assistance.

With the help of weed detection, the robot can then begin to combat the weeds. To do this, it uses cutting blades, similar to a conventional hoe, which chop off the roots of the weeds so that they die. It then remains on the field. The three hoes on the robot are aligned and operated by an electric motor according to the position of the weeds. Almost in real time.

If something goes wrong, the robot sends feedback to the control center, which can take further action. In principle, the robot should carry out its work continuously – at least as long as the sun is shining and the solar cells are providing an energy supply. After dark, the robot stops and does not resume its work until the next day.

Aigen makes it clear that the weed robots are not intended to replace human labor. Rather, employees who were involved in the application of herbicides, for example, could be deployed to monitor the robots in the control center.

Five robots are needed for a 65-hectare field. However, this adds up to a considerable sum for the farmer: A robot is said to cost around 50,000 US dollars to purchase. However, expensive agricultural machinery, such as combine harvesters, which are not used continuously, are usually loaned out by companies for a certain period of time for a fee. It can be assumed that this will also become Aigen's business model.

(olb)