homepage logo

Automated weeding machine developed

By Staff | Apr 16, 2010

AMES – Among all the speakers offering their expertise for niche crop and livestock marketing, it was a robot that had most peoples’ attention.

Vegetable growers throughout Iowa had a chance to look over an automated weeder, designed for small and mid-sized vegetable farmers, at the April 1 Marketing and Food Systems Initiative Workshop, in Ames. The event was sponsored by the Leopold Center for Sustainable Agriculture.

Dr. Lie Tang, an at Iowa State University ag and biosystems engineer, designed the robot to dig out weeds between vegetable plants. Dr. Lie said he is still testing an array of different scanners to see which works best to help the machine determine what is a vegetable plant and what is a weed.

The machine can be used inside high tunnels, or in the open field.

Lie, a specialist in automation robotics, said Europe is well ahead of the U.S. in bringing robotics into the farming industry. “We were looking for a solution for auto control of weeding in vegetable crops, and keep it on a scale affordable for small and mid-sized farmers,” Lie said.

He added that other goals included the machine caused minimum soil disturbance, a minimum of crop damage, while using the least amount of energy input.

His machine is geared to dig out weeds “while they are very small,” he said, making it easier to root weeds out.

Powered by a bank of 12-volt batteries, the machine is relatively quiet, lightweight, and, with an infrared sensing system, it can run at night, as well as during the day, extending the work window, Lie explained.

It is also a hybrid machine, able to run off batteries, solar power and petroleum.

Being GPS equipped, it is operated by remote control. “Farmers could sit on their porch and run the machine,” Lie said. It can also be towed behind a garden tractor.

By catching weeds when small, the topsoil would only be disturbed within the top inch, he told his audience of about 60 growers.

Lie told Farm News that the machine will be field tested this summer and the results will be made available. “We think it will do a pretty extensive job,” Lie told Farm News. “If producers have to do some weed pulling, it’ll be very little.”

One potential problem to be determined is if the sensor system can determine a weed, if the vegetation canopy is fully developed. But then, the canopy may be able to control the weeds by itself.

Another factor that Lie and his research team of ISU students are eager to determine is if the sensing system will be affected by going from sunny to shady conditions.

Lie’s team has experimented with four different sensor systems including stereo vision, graphical user interface, time of flight of light 3D imaging and active stereo using fringe patterns.

The latter, Lie said, seems to be the only sensor that can work at high speeds. The others work but at speeds that growers may deem are too slow, he said.

Contact Larry Kershner at (515) 573-2141 Ext. 453, or by e-mail at kersh@farm-news.com.

Please Enter Your Facebook App ID. Required for FB Comments. Click here for FB Comments Settings page