By Ann Filmer
Weed control is a continual struggle for vegetable growers, particularly hand weeding. Hand weeding is a major expense for crops like lettuce, and accounts for about 50 percent of the weed control costs for growers, according to Steve Fennimore, Cooperative Extension specialist and faculty member with the Department of Plant Sciences,
University of California, Davis.
There are three factors driving automation in vegetable crops: Need due to employee shortages, availability of new technology and no new vegetable herbicides.
“We have a tremendous labor shortage,” Fennimore said.
In the 1980s, employees came from Mexico, in large numbers, but those days are gone, he said.
“It doesn’t mean the end of people coming up here from Mexico, but it’s not going to be like it was,” Fennimore said.
The fact is, there just isn’t enough labor, which impacts farming operations from weed control to harvest. This makes mechanization very important, he said.
The technology is here with robotic weeders, but as far as herbicides, most used in vegetable crops were registered 40 to 50 years ago, and there are no new ones coming down the pike, Fennimore said.
Small specialty crops like vegetables with small acreages discourage the agrochemical sector from investing and developing new herbicides, he said.
What this means to vegetable growers is, the priority for vegetable herbicides isn’t even on the list for most agrochemical companies, Fennimore said.
With a reduced labor force and no new chemicals on the horizon, there is a lot of potential for robotic weeders for conventional and organic vegetable crop growers, he said.
“The automated weeders are real, they’re happening, and they’re in the field right now,” Fennimore said.
Automated weeders don’t need permission for use — other than from the grower, Fennimore reminded growers. There’s no government agency overseeing it, so that makes it much more straightforward, he added.
Automation for weed control has two components — detection and actuation (how the weed is killed). Automatic detection most commonly uses two-dimensional image processing to differentiate the plant from soil by color or light reflectance to detect the crop from the weeds by size differences and crop row pattern.
Actuation has two methods. One is to spray the weed and kill it with an herbicide. The other is mechanical that uses a cultivator knife to remove the weed.
There are three methods of physical weed control that are compatible with automation, and they will work in conventional and organic operations: Mechanical intra-row cultivation, thermal weed control and abrasion — sand blasting.
In the past, intra-row weeding has been done with hand weeding and selective herbicides. Currently, there are two designs for intra-row cultivators on the market.
One has reciprocating knives that reach in and out of the crop row using machine-vision guidance. The other has a rotating disk that is controlled by a vision system to detect the crop plant and align the disk cut-away section with the crop plant.
For best performance, machine-vision guided systems need a uniform, well established crop, low weed pressure and crop plants that are larger than the weeds, Fennimore said.
Tractor-mounted robotic weeders have been around for some time, and they are controlled by a driver, he said.
There are several tractor-mounted robotic weeders commercially available: Ferrari Remoweed weeder, Robovator weeder, Garford In-row weeder and Steketee IC weeder.
Fennimore has used automated weeders on 20 different crops, from lettuce and broccoli to Brassica vegetables like bok choy, gaylon and radicchio.
“There’s a possibility it could even work in cotton,” Fennimore said, “but high density crops like onions, spinach and carrots are more challenging, and a new weeder design will be needed to deal with these crops.”
Autonomous weeders are still in the development stage. Bosch Robotics has developed an autonomous weed control robot, called BoniRob. The robot is about the size of a compact car, and it is able to remove two weeds per second using laser-guided machine vision and GPS-based navigation.
Another robotic weeder under development is by ecoRobotix. The robot is also autonomous, solar powered and uses a weed crop detection system. It has two mechanical arms that selectively spray the weeds.
FarmWise has an autonomous weeder, too. A large, orange machine drives down the rows of vegetables, identifying weeds and pulling them out with a hoe-like arm.
The FarmWise robot uses artificial vision to move through the fields to determine weeds from crop plants. The robot can recognize the plants in different stages of growth and different types of crops.
A field robot that is used for weeding could potentially have other uses, too. For example, it could scout for insects or foliar diseases, which could lower the per unit cost of the robot.
A cost breakdown is being developed, but as yet it isn’t available, Fennimore said.
Robotics are a more expensive process, he continued.
“As the technology is developed, it will get cheaper because there will be more competition, and there will be better machines,” Fennimore said. “When I started doing this five, six, seven years ago, people just barely knew how to do it, and now they know how to do it.”
There is a common misconception that these machines will replace herbicide treatments, he said.
“They might be able to do that, but that’s not what growers are doing. In conventional fields, they’re using the herbicides,” he said. “What they’re doing is, they’re reducing the need for hand weeding.”
Fennimore said there are several things that need to happen to overcome challenges in weed automation:
- Better crop/weed recognition to improve the technology.
- Public funding to support research in specialty crops.
- Development of crop-marking systems to improve machine-vision recognition of the crop, using a unique marker with breeding or a physical marker placed on the crop during transplanting.
- Improved physical weed control actuators like abrasives, cultivators, high-pressure water, lasers and propane flaming.
- Changing weed science curriculum to train undergraduate and graduate students on the basics of robotic weeding.