On high-tech small farms, robotic milkers, automated greenhouses are just the start
BULGER, Pa. – About 150 Jersey cows in the rolling terrain at Rivendale Farms in Bulger, some 25 miles west of Pittsburgh, wear Fitbit-like collars that monitor their movement, eating and rumination patterns. They are milked not by humans but by robotic machines.
A nearby greenhouse, about a quarter-acre in size and filled with saladbowl crops like kale, arugula and baby carrots, is automated. The temperature, humidity and sunlight are controlled by sensors and retractable metallic screens. And soon, small robots may roam the farm’s 8 acres of vegetable crops outdoors to spot disease and pluck weeds.
Farming in America is increasingly a high-tech endeavor. Combines guided by GPS, drones, satellite imagery, soil sensors and supercomputers all help the nation’s food production. Yet that technology is mainly tailored for big industrial farms, where fields stretch as far as the eye can see.
Rivendale Farms, which has just completed its first year of full operations, offers a glimpse of technology coming available for smaller farms.
Technology for giant farms is all about increasing yields and cutting costs. For smaller farms, too, efficiency is paramount. But technology can also eliminate a lot of tedious, routine labor – a lifestyle payoff that can help persuade a younger generation to stay put on family farms rather than sell out.
Smaller farms typically raise specialty crops on limited acreage. Specialty farming requires a scaled-down approach, like the small robots being developed for Rivendale by scientists at nearby Carnegie Mellon University and the expanding array of equipment from the “slow tools” movement, a group of farmers and engineers designing affordable tools for small farms. The goal at Rivendale, said Thomas Tull, the farm’s owner, is to create a “boutique, cutting-edge farm, enabled by technology, that produces great food.”
Rivendale can afford its combination of cutting-edge commercial technology and science experiments because Tull is a billionaire serial entrepreneur, investor in tech-related ventures and former film producer. He is also on the board of Carnegie Mellon. He has spent several million dollars on Rivendale so far. But the plan, Tull said, is for the farm to become self-sustaining by 2020.
So Rivendale can try more things at once than others can. But its efforts, experts say, are part of a broader trend among small farmers seeking to raise healthy food and livestock using less fossil fuel, fertilizer and processed feed. “We’re seeing greater use of modern technology and tools in smaller, soil-based farming, and that vision is being wholeheartedly embraced at Rivendale,” said Jack Algiere, farm director at Stone Barns Center for Food and Agriculture, a nonprofit farm in Pocantico Hills, N.Y., that has been a leading advocate for sustainable agriculture on small farms.
Walk into the dairy barn at Rivendale and there are no people – only cows, an automated feed system and three robotic milking machines.
The Rivendale cows are milked four times a day on average, when they feel ready, compared with the traditional twice-a-day regimen when humans manage the milking. And its Jersey cows produce 15 percent more milk than the average for the breed, with a higher protein and butterfat content, said Christine Grady, general manager of Rivendale.
“They eat when they want, lie down when they want and feed when they want,” Grady said. “And a happier cow produces more milk and better milk.” Robotic milkers have been available for years. But the technology has steadily improved, requiring far less human assistance than a few years ago.
The machines cost about $200,000 each. Without them, and an automated feeding system, the milking barn at Rivendale would require five workers instead of being mainly overseen by one, Grady said.
George Kantor, a senior systems scientist at Carnegie Mellon’s robotics institute, is leading the effort at Rivendale to develop “scouting robots” to identify disease and weeds in the vegetable field, and then send smartphone alerts if there is a problem.
His team did field work and collected data in the fall, then shifted to a university lab for the winter. Computer vision and machine learning, Kantor said, will be deployed to distinguish healthy plants from diseased ones and weeds. A next step would be to get rid of weeds. Organic farming avoids pesticides. And digging up or pulling weeds, Kantor said, would require hand-like grasping and holding, a more daunting robotic task, where progress has been slower.
To most longtime farmers, AI stands for artificial insemination, not digital wizardry. But a new generation of family farmers is welcoming robotic assistants and smartphone apps.