ROBO-FARMERS capable of identifying each individual fruit on a tree and its degree of ripeness could reduce the manual tasks of farming and increase efficiency, developers say.
University of Sydney mechatronics experts are developing robotic devices with the ability to autonomously sense, analyse and respond to their surroundings.
A push for Australia to become the “food bowl” of Asia and meet the increase in demand for fresh produce could be restricted by labour costs and technology.
“I think in agriculture, because there are low margins and the guys are having a tough time anyway, that they want these solutions,” said Salah Sukkarieh, a professor of robotics and intelligent systems at the faculty of engineering and information technologies.
Professor Sukkarieh is leading a team that has developed robotic systems, sensors and intelligent devices. The devices are being trialled on an almond farm in Mildura, Victoria, and an apple orchard near Melbourne.
The team is one of a handful of groups around the world working on automated agriculture for orchards.
At present, driver-assist tractors are used in broadacre cropping such as wheat and rice.
The group’s three-phase program initially was focused on delivering autonomous perception to enable robotic devices to read and understand their surroundings.
“The robot would go up and down these tree orchards collecting information and data, and we would develop machine learning algorithms that would help classify the health of the tree, the ripeness of the fruit as well as fruit counting,” Professor Sukkarieh said.
He said within a year there would be working perception algorithms.
“The plan is within the next two to three years we would have a commercial autonomous ground vehicle operating on a farm that would do the perception task and within five to 10 years starting to look at machinery that can harvest the fruit as well.”
The work has received funding from Horticulture Australia as well as the Australian Research Council and the university.
“From an R&D perspective, the initial parts are very much more developed than they are research,” Professor Sukkarieh said. “So the technology is available, the understanding is available, if the funding was there you could progress along that path and build these autonomous tractors that do yield perception. ”
Professor Sukkarieh said farmers faced a larger capital cost upfront to adopt automation.
“It is operationally then when the reductions start to happen, so things like operating 24-seven, better fuel efficiency on the vehicles that are driven.
“When you model these tractors and so forth better you can design control algorithms that make them run more efficiently, so that improves maintenance as well as fuel costs.”
He said farmers’ acceptance of new processes and practices related to the use of robotic vehicles was also crucial.
“Whether somebody is willing to step up and take that risk and push it forward is what is going to govern whether or not it becomes successful,” he said.
Source: Jennifer Foreshew, The Australian http://www.theaustralian.com.au/australian-it/government/robots-find-themselves-at-home-on-the-farm/story-fn4htb9o-1226580619487