Rows of hedged apples trees at Rod Farrow’s Lamont Fruit Farm in Waterport, New York, in 2016. (TJ Mullinax/Good Fruit Grower file photo)
Start getting your orchard robot-ready now: Mechanical harvesters will be used commercially in your neighborhood within five years.
“The orchard you plant today is going to be the one that needs to be robot-ready; that’s five years out,” said Rod Farrow to a packed crowd at the Washington State Tree Fruit Association annual meeting in December.
That was one of the most oft repeated bits of advice from Farrow and three fellow panelists discussing “game changers,” innovations likely to drastically impact the fruit industry. The presentation was one of the better attended at the Wenatchee, Washington, conference, with more audience members than chairs.
The other panelists were West Mathison, CEO of Stemilt Growers in Wenatchee; Sam Godwin, an independent grower in Tonasket, Washington; and Dan Plath, orchard manager and owning partner of Washington Fruit and Produce in Yakima, Washington.
Farrow, a partner in Lamont Farms in Albion, New York, considers a robot-ready orchard one with canopies less than 3 feet — ideally 2 feet — wide with flat fruiting walls that can accommodate automatic pickers, self-driving platforms, automated computer-integrated apple counters and other forms of mechanization.
He also envisions future orchards with between 8,000 to 10,000 leaders per acre, producing fruit at a breakneck volume of 170 to 200 bins per acre.
He advises growers to start thinking right away to plant enough robot-ready orchard land to make the investment worth it. Robots are coming faster than anyone would have guessed just 18 months ago.
“All of a sudden robotics went from ‘Yeah, not in my lifetime,’ to ‘Holy cow this is coming’,” he said.
Godwin, a relatively small, independent grower like Farrow, echoed the sense of urgency.
“It’s never too soon to start for small growers,” he said.
Godwin has invested in tall trellis poles well before he plants to be ready to install shade cloth when he wants. He also has purchased small, 15 to 20 acre farms, near Tonasket to expand to create a critical mass to justify the expense of operation.
For example, he and Farrow agreed on needing roughly 25 to 30 acres to make a platform worth the cost, more for a robotic picker.
Mathison also expected canopies to get narrower for robotic pickers and assumed they would start by picking 80 percent of the fruit automatically, leaving 20 percent for hand pickers. That’s not perfect, but the breakdown would improve over time, he said.
Besides, the industry needs something besides people, he said. Of the 1.35 million specialty crop employees in the country, 880,000, or 65 percent, work fewer than 150 days a year, according to the 2012 Ag Census by the U.S. Department of Agriculture.
He also expects Initiative 1433, Washington’s new minimum wage initiative passed in November, to boost the cost of labor by 16 percent to 20 percent in addition to cost-of-living increases. California and New York also have new minimum wage increases going into effect.
“We’re going to have to find a way to harvest fruits and vegetables other than the current way we’re doing it,” Mathison said.
The panelists also riffed on advice and prognostications about everything from family-work balance, banking and planning.
Farrow particularly emphasized planning — one of his favorite topics — urging growers to resist the temptation to stay so busy they don’t consider long-term ideas. Hire chores out if need be, he said. For instance, spend $10,000 to pay someone $20 per hour to spray for 500 hours per year, leaving enough time to devise ways to double that in increased revenue.
“If you can’t figure out how to make $20,000 dollars in your business with 500 extra hours to spare, maybe you should be considering whether you’re in the right business,” he said.
Planning requires knowledge. Most of the panelists suggested growers attend as many workshops as possible. Mathison even urged growers to visit the research and development workers at packing houses, something a few of his growers do at his Wenatchee campus.
Plath urged growers to look beyond saving money, pointing to studies that showed more financial gain in a Gala orchard from improvements than from an equal amount of expense cuts.
He called mechanization a way to attract workers, not just a cost-saving measure. For example, while platforms have improved efficiency by 10 percent to 50 percent at his orchards, employees get excited to work in an orchard without ladders, he said.
Technology is more than expensive tools, too, he said.
Growers should study up on the availability of rootstocks from tissue cultures, thinning using pollen tube models, stretching reflective ground cover to maximize light exposure and switching to new varieties that will be popular for the next 15 years.
Quoting the late Sunnyside, Washington, orchardist Jerry Haak, Plath said, “The cheapest way to do something is the right way the first time.” •
– by Ross Courtney
Please tell me how to get 170 bins per acre
Andrew price
Northern Ireland
We’re working on an open source robotic harvesting platform that should cost less than $10k per unit. It utilizes some of the same computer learning and image recognition tech that’s been deployed in self-driving cars. You’ll note that it wasn’t necessary to redesign the entire road system for self-driving cars to work, neither is it necessary to redesign orchard layouts in order for automatic harvesting robots to work. Harvesting robots can rely on the exact same computer learning methods and navigation technology as used in self-driving cars, with the addition of a picker arm. Any well designed robot can mimic the actions taken by a human, but faster and with more accuracy. If a human can navigate your orchard without difficulty a well designed robot can too.
Our design involves each robot being controlled by a central server that can be located anywhere in the world, provided network latency is low enough (< 250ms), this minimizes the need for expensive on-board processing power for image recognition and control. This also means the only major change that would be necessary for orchards to benefit from this technology would be the installation of some form of internet access in the orchard, this could be cellular or conventional internet. In later deployments we plan to offer a weatherized central unit for orchards that don't have internet availability.
The plan is to sell the robots at our cost, give the software away or change a very low licensing fee (depending on our development funding source) and only make a profit for customizing deployments or assisting with initial setup (if needed). We want to make this as affordable as possible, but we still need to eat. The objective is to keep orchards in business that otherwise couldn't afford to be. Let's face it, with most money making ventures the objective is to make as much money as possible and small to medium size orchards couldn't easily afford this tech if it were released in the normal fashion. We'd rather keep you around as independent customers if we can.
We've applied for a WSDA grant to help fund the initial development of this project. If we get that we anticipate prototypes being used in the Wenatchee area in the next 1-2 years with full release to growers in Washington state within 3 years. Otherwise it'll depend on how long it takes to raise the money needed for parts and prototyping equipment. In the absence of a grant we would be using our own funds for this, but we may also reach out to the small to medium size growers community for assistance.
Regardless of the funding source, the intent of this project is to create a robotic harvesting platform that is highly affordable, works in any orchards that humans can work in, and has no proprietary components. The plans will be made freely available, as will any advancements we make. If growers want to build their own robots from our plans, or adapt our plans to their needs that's fine, same goes for the software and its design. Ultimately, if we have data collaboration between small to medium size growers the accuracy of all robotic units will improve due to a greater mass of accurate field data and that will be a benefit to everyone using this technology.
We're doing this because we've got some friends in the orchard industry who won't be able to stay in business without a solution like this, because in the last two years the tech to do this has become widely available (and free) and because, as anyone who loves a challenge will tell you, its there to be done and we know we can do it.
We'll be launching a web site in the next week that provides the current state of our concepts/designs for public comment. This is necessary both to allow for feedback and design input from people in the Orchard industry as well as to ensure that anything we develop hits the public domain before anyone who might be unscrupulous tries to patent it and restrict its free use. I will provide the URL as a follow-up comment when its available.
So, while its up to you, you might want to wait on redesigning your orchard layouts for just a bit. Rather than adapting your orchards to the technology let's adapt the technology to your orchards.