Harvesting Innovation: Putting AI to Work in the Vineyard

Among the benefits AI innovators are offering viticulturists right now
are advances in robots, machine learning and predictive analytics.

By Laurie Wachter

The speed of new technology adoption is accelerating, especially for Artificial Intelligence (AI). News stories feature ChatGPT almost daily, and grape growers can see AI’s impact in vineyard management software offering precision agriculture, sensors and drones measuring vineyard activities and vine health, and robots and tractors performing tasks independently.

Despite challenges that come with any new technology, from the degree of trust people have in new solutions to the cost of replacing existing technology, a recent meta-analysis on innovation in the wine industry found that adopting new technology correlated positively with financial performance. At the recent Wine Industry Sales Symposium, Justin Noland, head of DTC at Treasury Wine Estates and a successful adopter of AI, recommended a practical approach to getting started.  

“The first thing to do is look at tasks you can delegate to give yourself more time,” he suggested. “Then get in and try stuff. There’s no substitute for just doing it. Since the goal of most AI is to learn from you, the more feedback you give it, the better.”

Among the benefits AI innovators are offering viticulturists right now are advances in robots, machine learning and predictive analytics.

Robots 

Mechanization has become increasingly important to farmers of more than 600,000 acres of winegrapes in California, who are searching for solutions to labor and climate change impacts. The shift began with the now mainstream use of machine harvesting. Today, viticulturists are evaluating remote-control and autonomous tractors and robots for other tasks, such as weed control, soil amendments and pruning. 

One hidden but highly consequential benefit of using self-propelled machines is that they can capture sensor readings and images of vines, leaves and grapes while undertaking repetitive tasks. This perk significantly increases the volume of data fed into AI systems without tapping human resources. More data — especially more detailed data, such as continuous images or videos instead of occasional mobile device snapshots or notes jotted in a diary ― will dramatically expand the knowledge base for machine learning analysis and, thus, provide more accurate recommendations.

“Farmers don’t want data; they want analytics,” says Tim Bucher, CEO of Agtonomy, which makes a hybrid autonomy and tele-assist platform for agricultural vehicles. “In irrigation management, for example, one task is to drive a quad around the vineyard and check the irrigation. Doing this autonomously is a massive help, but if it analyzes leaves simultaneously and tells the farmer, ‘You have powdery mildew over here,’ that adds value.”

- Advertisement -

Machine learning 

Machine learning is a subset of AI that uses algorithms and large volumes of historical data to train its models to recognize patterns and make decisions based on current data. ChatGPT is an example of this outside of agriculture. In vineyards, autonomous tractors and robots use data collected by mounted sensors to help them navigate. “Perception stacks” on the vehicle analyze the sensor data to identify obstacles, recognize vines and shoots that need trimming, and differentiate weeds from trunks. It is, essentially, machine learning in action.

Machine learning can analyze drone and satellite imagery, localized weather station statistics and soil sensor data to build irrigation schedules, identify disease in vineyard sections or monitor ash particles in the air. One aspect of machine learning, called “deep learning,” uses neural networks to handle more complex patterns. For example, autonomous vehicles use neural networks with images and videos to identify “drivable” space or, for a tractor, the open vineyard row between vines. 

Predictive analytics

“Growers manage complex ecosystems and have built mental models for how each vineyard block works based on years of farming these ecosystems,” explains Tom Shapland, CEO of Tule (now part of CropX), which employs in-field sensors and predictive models. “Growers use these mental models to predict and take corrective actions on issues such as pest infestation and water stress. The new suite of sensing technologies and predictive analytics is moving these mental models to the cloud so that growers can use AI models to understand and make predictions about every block. Because sensors update these models in real-time ― far more frequently than farmers can visit each block ― they can automatically send notifications or revise irrigation and spraying schedules.” 

Predictive analytics can also estimate harvest yields and grape quality, forecast extreme weather, predict pest and disease outbreaks, and measure a winery’s progress on greenhouse gasses and soil carbon respiration and sequestration.

While some grape growers are fascinated by the potential of AI and experimenting with different solutions, many feel overwhelmed. Others are simply uncertain about which solutions will bring the most value for the time and money spent. No matter your personal comfort level, viticulture is changing quickly. It’s time to start exploring what AI can do for you — and your grapes.

Editor’s Note: Wine Industry Network’s upcoming Growing Forward Vineyard & Grower Virtual Conference (July 19) is designed to educate and help grape growers and vineyard professionals decide how best to integrate cutting-edge technologies such as AI, and navigate other issues impacting the ever-evolving practices of vineyard management, climate adaptation, soil moisture monitoring and irrigation. For more information on this free virtual event, go to www.wineindustryadvisor.com/growingforwardvineyardconference.

________________________________________________________________

Laurie Wachter

Laurie Wachter developed her love of analytics and fascination with automation while advising consumer packaged goods companies, including Kraft Foods, PepsiCo and the Altria Group, on their direct-to-consumer marketing. Today, she writes about innovation in the wine and food & beverages industry for a global client base.