Cassandra Landry | 11.26.2014 | Issue #5
“If technology could change your life, from the soil to the bottle, what would you want to happen?”
For a winemaker, that’s a loaded question. For someone in the tech industry, that’s one hell of an opportunity.
The answer led to the latest disruption in the winemaking business.
A tech guy happily marooned in the wilds of Santa Barbara wine country, David Baeza spent most of the decade working as a VP at a software company. On the weekends, making the rounds at birthday parties and backyard barbecues, conversation centered less around APIs and SAAS, and more around topsoil, yield, and vigor.
“We’d hang out at parties and they’d talk about their problems, and I’d say, man, if you guys invested in just a little more technology, a lot of these problems could be solved,” he says of his vintner friends. It went on like this, casually, for a few months, until he realized there was a simple way to bridge the two worlds he cared about: drones.
Depending on who you’re talking to, the word “drone” either elicits a flinch reflex or a knowing nod. It’s both exotic and run-of-the-mill. Baeza says the reaction changes mostly by age—the winemaker in that 30-40 age range uses technology every day, and are thus more likely to be receptive—but as with most new applications of technology, it becomes a matter of show, not tell.
“Drones are the gateway drug. It’s really sexy and gets a lot of press, but it’s just a tool to gather data,” he says. “It definitely gets them interested. Once we start to present the data and they see what the drone can see, the conversation is about the promise to improve quality and yield.”
The drones—robustly dubbed Vine Rangers—drift over the hills and roll through the trenches of a vineyard, capturing near-infrared images of both the canopy level and the rows from above, gathering data about their grapes (mostly indicators of crop disease or water issues) that winemakers often can’t see until it’s too late. The data basically enables farmers to spot-reduce. No need to coat an entire area with psychotic honeybee-killing pesticides when you can locate a problem and act quickly. No finger-crossing, no watching your life’s work slip away at the hands of a sneaky disease.
All this jazz is, of course, accessible through an iOS and Android app, and Vine Rangers will even fly the wino-bots for you through a turn-key system. The drones, software storage, and data processing will only run you a mere $20 a month per acre, a major coup for an industry just beginning to scratch the surface of vine surveillance in the new battle for proactive farming.
“Our whole lives circle around 200 acres of grapes that we’re in all the time,” says Keith Saarloos, a Santa Barbara winemaker involved in Vine Rangers’s pilot program. “Everybody thinks they’re doing the right stuff, but when you start taking a really analytical look at it, from something outside of who you are, you start learning massive amounts of information.”
100 years from now, he says, these drones we think of as cutting-edge? Just another proverbial tractor, taking the place of some other horse and plow. It’s a tool, he emphasizes, echoing Baeza’s sentiments, and points out that farmers are usually among the first to embrace new technology.
“Food and agriculture today is not the picture of the farmer with hayseed in his mouth goin,’ ‘Aw, shucks.’ The only thing the farmer needs to see is whether this is making me more money, and making me more efficient,” he says. “Our whole goal is to future-proof our entire company, and to future-proof our wines by looking at things in ways that nobody’s looking. If you can be the first to get behind that, you get to carve the path of where it goes.”
This particular path is being gently worn into existence with the help of the computer science department at the University of California, Santa Barbara, whose Smart Farm Initiative program aligns pretty neatly with Baeza’s Vine Rangers. When you’re squaring up to save the world, there’s no harm in the buddy system.
“We’ve got some common issues we’re trying to solve,” he explains. Vine Rangers’s biggest stumble in the early days centered around data processing: an afternoon of flying over an average-sized vineyard garnered about 9GB of data, which was then pulled off, sent to Baeza’s partner in the United Kingdom, Oxford’s Dr. Ashutosh Natraj, for analyzing, and then sent back. It was a clunky, slow process. UCSB was interested, so the two start-ups struck a deal for a three-year program: UCSB helps process the data and churns out analytics, and in return, they get access to Vine Rangers’s anonymous data for their initiative. “They get it. They want to build a better planet too.”
“We spend so much time in the two feet mode. You’re in a row, looking at a vine from a few feet away, squatting down, which is kind of the opposite of how most people look at things. Farmers are very much about each individual vine, each individual plant,” Saarloos says. “By being up in the air, you get the ability to identify a problem that you’d be too close to if you were standing in it.”
A recent fly-over brought some vines that were growing a little too vigorously to Saarloos’s attention. The ability to go straight to the source revealed underground breakage in an irrigation duct, which was promptly fixed and the pocket of vines was brought back in line. Money, resources, water, and time saved.
“It’s the Air Force, right? You can fight on the ground all you want, but once you’re up in the air you see where everything is,” he says. “Looking into that is the next big step.”
This is the first time farmers can have a leg up on what’s happening on their vineyards before they even walk in, Baeza explains. “The more control you put back in the hands of the farmer, the better. They know their property better than everybody else. They know their microclimates, their topography.”
For those smaller vineyards, like the ones Chris Brockway of Broc Cellars in Berkeley works with, the only worry seems to be striking the delicate balance of owning an expert tool and making sure it didn’t become a crutch. The only reason you know that topography inside and out is purely because you spend so much time trying to figure out what’s going on at the “two-feet mode.”
“A lot of times I’m just dealing with a two-acre block or a four-acre block, where the best thing is really to get out there and poke around,” Brockway says. “That being said, it would be a great secondary tool for me, especially in areas where we do have some problems every year or every two years. Since I deal with a lot of older vineyards, there’s usually some kind of inherent vine disease.”
In Brockway’s mind, even though drones seem like a trapping of a much larger vineyard, he admits that in dire drought conditions like these, anything that could monitor and lessen water use deserves consideration.
“There’s never a downside to more information, especially when you’re talking about vineyards,” Brandon Moss, a winemaker with Gramercy Cellars in Washington, adds. “I certainly think more technology in the vineyard is better, because you need to keep track of those vines. Each vine is its own individual being, and there’s so much variation. Just to be able to tell what’s going on from up high would absolutely be something we could use.”
Winemakers are the ultimate cowboys, loyal traditionalists, staunch scientists, down-to-earth farmers. No matter how you slice it though, drones, due to their passive investigating, have the ability to appeal across the board.
“That traditional-side of winemaking comes out more in not wanting to do extreme filtering or things like that. That’s the technical side that we like to stay away from,” Moss says. “But this is a non-invasive procedure. You’re not cutting into the vines, you’re not doing anything that’s negatively affecting them.”
In preparation for his recent TEDx talk in Boston, Baeza and Natraj were going over their data one last time. As a last test, the team was doing cluster analysis at Firestone Vineyard, another local winery involved in the beta testing.
“We’d hold up a cluster, and [winemaker] Paul Warson would taste it and tell me if it was ripe, what the quality was, the flavor,” Baeza says, barely able to conceal the excitement bubbling up through his voice. “We would shoot it in infrared, and we would send it to [Natraj], and he had to determine what Paul was tasting, site unseen. And he got it right, ten for ten.”
To be able to tell, without a shadow of a doubt, that a cluster of grapes was healthy and at peak ripeness from 5,000 miles away was huge.
Before you freak out and visions of conscious, cunning machines start crashing through your mind, know that the data is directional, “but not absolute,” which guarantees the drones won’t be taking over the entire winemaking process a la the robotic bartender. To prove his respect for the wisdom of the vintner above all, Baeza decided to run an analysis and come to a conclusion based purely on the data, without consulting a winemaker, and see how far it would get them. “For shits and giggles,” he says. It, uh, went as you might expect.
One image showed very high canopy—high vigor and growth—and very low canopy on the other side. Water saturation was wildly different on either side. They banged out an analysis and logically advised allocating more resources to the weaker-seeming vines. The winemaker, gracious in his approach, told them that while the interpretation of the actual image was on-point, sure, but they were still wrong. The fruit on the side with low canopy was good, because the vine wasn’t wasting sugar and nutrients on the leaves in the canopy, it was pushing it to the grapes. On the other side, with the big, high, beautiful-looking canopy that looked gorgeous and healthy, the fruit was worthless.
“This can never replace the winemaker or the farmer,” Baeza assures. “It’s just going to help them make more informed decisions.”
He hopes to have Vine Rangers on the market by February or March of 2015, after the second half of the beta program—post-harvest, pre-bud break—wraps up this winter.
“Wine takes years. It’s slow. We only get one shot a year. So every time we take it, it’s like a moon landing,” Saarloos says. “If we can have more data and more information and go straight to where we need to go, we can make a better wine immediately.”