Since the Industrial Revolution, people have sought to mechanize labor, from the power loom and cotton gin of the late 1700s to the robotic grocery store assistant you might catch strolling aisles today. While human work can be cheap, machines are cheaper—especially because they don’t demand things like health care and paid time off. One of the industries that’s been most transformed by mechanization is agriculture, with innovations like tractors and milking machines speeding up production and increasing yields. Unsurprisingly, the use of robots on farms is an increasingly sought-after technology, as farmers face complications like the expensive unpredictability of climate change and a decreasing interest, on the part of the new generation, in continuing to farm. The latest robot to come on the scene is a lettuce-peeling machine, and while it’s not yet as skilled as its human counterparts, it heralds a shift in agriculture that could one day mean more robots and fewer humans.
Announced last week by engineers at the University of Cambridge, the lettuce-peeling robot is an attempt to speed up the time-consuming work of harvesting iceberg lettuces, which grow close to the ground, have to be cut from their roots by hand, and then must be stripped of their dirty, unattractive outer leaves before going to market. When you’re paying a human worker by the hour, all that labor adds up, an issue the robot’s engineers have attempted to address with their invention.
“Post-harvest manipulation of produce is something done by hand in many crops,” Luca Scimeca, one of the graduate students who worked on the project, told MUNCHIES. “The manual operation requires specialized personnel, is expensive and is hard to scale. In our laboratory, we look at solving real-world problems and harvesting, post-harvesting techniques and agri-robotics pose challenges whose solutions could very much benefit the industry of the future.”
Historically, it has been hard to mechanize the harvesting and processing of produce: Fruits and vegetables are delicate and easily bruised, leaving them vulnerable to damage that could send them to the compost bin instead of the supermarket. But the university’s robot, developed by a team led by mechatronics professor Fumiya Iida, utilizes new technology such as a 2D web camera that acts as the robot’s “eye,” a 3D-printed circular nozzle, and a suction system to help it grasp, and peel, the lettuce efficiently.
“In a delicate manipulation task, such as peeling an outer leaf of a lettuce, you are encouraged to find solutions which are reliable against such diversity,” Scimeca said. “This makes this a challenge worth exploring.” The technology, the researcher noted, could someday be applied to similar crops, such as cauliflower, that also need to be peeled before being packaged.
But while Cambridge’s robot represents an advancement in technology, it still doesn’t compare with the speed and accuracy of a human worker. It takes the robot an average of 27 seconds to complete its task, which it performs accurately only about 50 percent of the time. That’s pretty pathetic in comparison to a person, who can peel a lettuce in a few seconds flat and with total accuracy.
For that reason, robots like these will need some more work before they come close to replacing human workers. More likely, they’ll first take their place alongside people, making their tasks faster and easier, until technology advances to the point where the machine is comparable to the man.
“In my opinion, the agri-robotics of the future will not entirely replace human labor, or not for a long time,” Scimeca said. “I see these technologies aiding in farms, or in packaging factories, alongside human operators. I see the potential for labor in farms to become less strenuous to the body, improving the health conditions of workers which may be supervising, rather than doing, the labor intensive-tasks.”
Dr. Philip Martin is a professor emeritus at UC Davis’ School of Agricultural and Resource Economics whose work has examined the impact of machines on US agriculture. He said that while robotics are going to continue to have a transformative effect on farmwork, it’s going to be a marathon, not a sprint.
“Mechanization is going to be evolutionary rather than revolutionary,” he told MUNCHIES. “It won’t be a case of this year there are 10,000 farm workers and tomorrow zero. But it might be a case of next year there are 9,500 workers, and the year after that, 9,300. In ten years, it might be a threat.”
Martin noted that due to their high operating costs and dependence on many hired workers, farms have always sought to integrate machinery, and cut costs, wherever possible.
“Agriculture is a history of labor-saving innovation,” he said.
For farms, Martin said, the most attractive aspect of technology is its fixed cost. While wages, and workers’ demands for them, have increased over the decades, a machine does work for free, after an initial investment and any necessary upkeep.
“In most countries, the big period of labor-saving mechanization happened in the 1960s, when wages went way up, in the US they went up by 40 percent,” he said. “Then, when immigration from Mexico surged in the 1970s, mechanization slowed.”
Today, with the minimum wage, in many states, at an all-time high—in California, the nation’s breadbasket, it’s up to $11 per hour, to increase to $15 per hour by 2022, and in Washington, it’s at $11.50 and set to increase to $13.50 by 2020—“there’s a new push for mechanization,” Martin said.
It’s true that in western states especially, with their heavy reliance on the inexpensive labor of undocumented Latin Americans, some farm owners might hold out longer and continue to rely on traditional workers.
“There might be some labor contractors who are going to be somehow shading things a little bit, that might say, ‘Don’t mechanize,’” Martin said. But, he noted, they, too, will eventually have to succumb to the reality of increasing costs. “It’s not gonna get cheaper. It’s gonna get more expensive.”
In the meantime, scientists and engineers will surely work to improve the speed and accuracy of the clumsy, fumbling robots that currently represent the highest echelons of technology.