Imagine the farmer who sits down at his kitchen table and uses a computer to plant his crop.
Out in the field are a series of small robotic machines that have been working day and night to meticulously and precisely plant and pack.
Now imagine an autonomic robot moving up and down the rows of the growing crop, seeking and destroying weeds.
Sounds like science fiction? Think again.
Reza Fotouhi, an assistant professor of mechanical engineering at the University of Saskatchewan, thinks farmers could be in the next great agricultural revolution within 10 years.
Read Also
Agriculture ministers agree to AgriStability changes
federal government proposed several months ago to increase the compensation rate from 80 to 90 per cent and double the maximum payment from $3 million to $6 million
Fotouhi’s robot, which is about the size of a small coffee table and weighs 45 kilograms, recently navigated the halls of the engineering building at the university. It was able to map out and memorize its surroundings and then navigate the area on its own, finding the quickest route.
A flexible robotic arm located on top of the robot can hold up to two kilograms.
Fotouhi was able to view the robot’s location on his laptop in a laboratory and control it if needed. He was also able to use an attached video camera to see what was ahead of the machine.
The robot is called PowerBot, which is the brand name of the base on which the technology sits.
Fotouhi, who has been working in robotics at the U of S for eight years, specializes in developing algorithms for robotic applications.
“We tell the robot which paths to follow. If there are obstacles, like someone walking in front, it will either stop or go around,” he said.
Fotouhi originally created the robot and its accompanying algorithms for a medical application.
“In a remote setting without nurses, this robot can play the role of a nurse carrying equipment,” he said.
“Further down the road we have a manipulator attached to the robot that this robot can go help patients. For example, (with) the SARS scare in Toronto and contagious diseases, they wanted to keep contact between the patient and caregiver to a minimal. That kind of robot can play a role. Send it, give a pill, glass of water, food, these kinds of simple jobs. Necessary but simple.”
He said expanding the technology into the agricultural arena is more ambitious. Instead of navigating a flat, smooth floor, the robot must now decipher a graded, unmapped field littered with rocks, weeds and crop.
“The idea is to develop a robot that can navigate in the auto setting.
“It’s not flat. It could be graded and we don’t have a map of the area. The idea is to develop a robot for weed control. That’s one application we are going for,” he said.
“This robot would navigate through a cornfield, for example, and identify the weed from the crop and destroy the weed.
Destroy could be done chemically or mechanically. We’re going for chemical because it’s easier for now.
“Navigation is harder in the auto setting because we have to use the GPS system. We have to identify where the crop rows are. There’s so many challenges.”
Fotouhi is concentrating on seeding and packing.
“Taking corn, for example, the seeds are placed in not a precise manner. Some seeds are placed six to eight inches apart, sometimes double,” he said.
“With a small robot, we can increase the precision of seeding. So another student’s PhD project is developing that ground engagement tool.”
The next phase will be to develop a robot that can identify and destroy weeds. It’s a challenge because it requires specialized vision.
The work’s novel aspect is key to attracting and receiving corporate and government funds.
“This research is new in Canada, but around the world a few people attempt to work in this area, but still in baby steps, a few in the U.S. and mostly in Europe,” he said.
Fotouhi expects that navigation and ground engagement will be completed in three years if we can raise the necessary funding.
“Putting them together and testing the algorithm that we developed will take about two years,” he said.
“I’m thinking five years from now, we’ll have the first prototype that can actually do something. Debugging and making the robot perfect will take maybe a few more years.
“The second phase will be going to perfection and commercialization and so on.
“If everything goes well, perhaps we’ll have robots roaming in the agriculture fields helping farmers. Electrical batteries could power these robots. They’re green, very precise and they can work day and night.”
Fotouhi said humans will remain in charge. Someone still has to start the robot, feed it and make sure everything is working.
“Instead of the farmer sitting on the seat of a tractor, he’s in a control room with 10 of them working at the same time and all he has to do is make sure they are working properly,” he said. “It will be a different era for agriculture.”
