California winemakers are using A.I. to combat climate change challenges
Drones hovering over grape vines—that’s the sight you may see when you visit Gamble Family Vineyards in the Oakville region of Napa Valley, California. There, proprietor Tom Gamble and his team send up drones regularly to collect data on the vineyards, health of the vines, and the effect of recent weather.
It’s not his only embracement of technology. Gamble believes that utilizing innovative tools, like artificial intelligence, allow his team to drop many tedious production tasks and focus on more value-added projects. That includes wireless soil sensor networks with recommendations on water needs to support his deficit irrigation practices and drones that identify grapevine diseases, gauge vine density, and even apply mildew prevention products in hard-to-reach vineyards.
The soil saturation technology alone has saved him tens of thousands of gallons of water per acre, simply by better understanding “when and when not to turn on the spigot,” Gamble says, explaining that it’s been critical to harness the power of these tools to support his goals for a long-term approach to precision farming and sustainable architecture.
About 80 miles away in the Mendocino area of Sonoma County wine country, Joanna Wells runs Signal Ridge Vineyards, where she farms grapes for well-known producers including Coil Wines by Chris and Josh Phelps as well has her own label, Model Farm. For her role as winemaker and viticulturist, access to data is integral to making informed decisions about her property and vines.
Not only is she fighting the same challenges around climate change as her neighbors, but the configuration of her property—a 40-minute drive up a narrow road in a redwood forest—adds its own complications. She farms both Chardonnay and Pinot Noir grapes along steep hillsides. Each row of grapevines sits at varying aspects and exposures, with sometimes vastly different requirements. It requires highly specific, block-by-block farming that is largely done by hand.
So, they both have turned to technology powered by one approach: artificial intelligence.
“We can offset this labor-intensive farming style using predictive modeling A.I., dial into the needs of the vineyard on a micro level and get ahead of potential problems before they arise,” Wells says. “The ability to predict our needs is paramount.”
Seeing so much tech in California wine country is both expected and yet not. Wine has often marketed itself as a natural product, one that combines art and science, with much more emphasis on the art side of the process. That’s been changing in the past decade, as younger generations have taken the reigns at wineries and naturally shift to more technological solutions.
Recently, it’s been enhanced by extreme and volatile weather patterns in California wine country: ongoing drought, rising temperatures, and wildfires that have led to smoke taint from wildfires, diseases in the vineyards, and lots of lost juice in a region that accounts for 81% of wine production in the United States. A report by Allied Grape Growers and the California Association of Winegrape Growers estimates that in 2020, the California wine industry lost more than $600 million due to climate-related impacts.
Yet its proximity to Silicon Valley may have a natural effect too: winemakers are increasingly adopting technological tools to navigate current challenges and look to the future amid climate change.
“Winemakers need a tool to future profile their knowledge, a tool which will use all their historical wisdom and create self-mutating playbooks to help them resolve challenges for production, distribution, capital, and sustainability,” explains Prateek Srivastava, cofounder and chief executive officer of Terraview, a software product that helps wine producers make more sustainable choices in the production process. “The only tool that can possibly come close to achieving this is artificial intelligence. It is not a magical tool, but with its ability to learn in a compounding manner, it is the only one which can play catchup with climate change and combat it.”
Srivastava utilizes three types of A.I. with his platform. The straightforward voice and text A.I. allows a winemaker to speak into the Terraview program, and the technology will catalog and analyze that data. He uses image A.I. through computer vision to detect, identify, and recommend solutions for various plant stresses, such as disease or lack of nutrients for a specific vine. Terraview also employs functional A.I., which integrates with other connected IoT devices to allow producers to run an intelligent winery, even remotely. It’s all showcased on a single dashboard.
Similarly, Wells uses ADCON Telemetry’s Decision Support System, which, like Terraview, collects and processes inputs from raw data to help her make smarter decisions around crop management, disease pressure, and natural elements such as frost. She credits the program with her success moving the vineyard entirely to a dry farming model, eliminating any reliance on irrigation, as well as backing away from non-organic chemicals and even some organic fungicides.
“Traditional viticulture is built off the idea of eradicating any and all potential threats for the sake of vine health—strip spraying under the row, toxic chemical fungicides, denuding canopies—that also kills everything beneficial to the soil,” Wells added. “A.I. helps growers work off evidence-based data rather than fear.”
A.I. is found beyond the vineyard too. Tastry founder Katerina Axelsson relies on A.I. to drive her entire company, which uses advanced chemistry and machine learning to predict consumer wine preferences. She collaborates with wine producers like Atlas Wine Company and retailers including Stop & Shop to recommend wine blends that will appeal to their audiences. This done by analyzing more than 1,000 data points about a bottle of wine combined with consumer preference data by zip code.
By determining which chemical compounds in wine are associated with what flavor and texture characteristics, Tastry’s team can predict how a wine will be perceived by the palate. By matching that palate with that of the customer, Tastry can advise brands and retailers on what to stock that will more likely sell and suggest new markets for distributors and importers to consider for their products. It’s highly technical, and without the help of technology, completely unimaginable, Axelsson says.
Her data has recently inspired her to close the loop back to the vineyard: her information on consumer flavor preferences has inspired wineries to test new varieties for commercial success. As a result of the wildfires in the past few years in northern California, Tastry has made recommendations to clients on specific oak treatments that will mask the effects of smoke taint on grapes, which often causes ashy and burnt flavors in the final wine. The A.I. is reducing waste, salvaging otherwise unusable wine, and exploring the potential of new varieties, while minimizing risk.
Srivastava says that there are more A.I. products being developed in the market around weather data. These products will drive insurance products, climate risk prediction products, and recommendations for better global supply chains, he says.
Gamble shared that the University of California, Davis, is working to release a mobile testing kit designed for non-technicians to use in the field to analyze vine health and detect disease before it spreads. Gamble is also following the development of driverless tractors, and Srivastava says that robotics will play a role in the future of AgTech for both production and shipping.
“A.I. is the future for those willing to invest,” Gamble says. “Those who do not will become less competitive. The needs of people, planet, and profit—the Venn diagram of sustainability—are all being addressed when A.I. is property used.”
Never miss a story: Follow your favorite topics and authors to get a personalized email with the journalism that matters most to you.