Artificial intelligence is the shiny new toy in agriculture, but specialty crop growers in the West – organic and otherwise – are still struggling to find practical applications for the data or even adapt to the new technology, a panel of industry representatives said.

While machine learning has been helpful for major commodities in the Midwest, effective data management in the West’s myriad produce and orchard crops has been elusive, a panel of insiders said during the Organic Grower Summit this week in Monterey, Calif.

For one thing, data points are so disparate that it’s difficult to put all the information together in a way that growers can easily use. Derek Azevedo, executive vice president of Los Banos, Calif.-based Bowles Farming, said the operation at one point had to make sense of 42 different data dashboards.

“It’s like having a whole bunch of pagers attached to your belt,” Azevedo said, adding that growers have sometimes had to spend a couple of hours at the computer just to schedule irrigation. “Ag is open to advancing technology. The VC’s (venture capitalists) wanted to sell the sizzle, not the steak. Farmers want the steak.”

Lesson learned

Bowles Farming hired a technology director in 2016 and quickly “learned a valuable lesson,” Azevedo said: “We started with the sizzle,” he said, but soon found that technology can end up costing you more money if it’s not focused.

“The pain point for farmers was data collection,” he said. “The user experience was awful.” Some operations had to hire more employees to compile “reports that were just nice to have for the CEO,” he said.

Better Food Ventures founder Rob Trice said more growers are interested in adopting AI systems, but the biggest barrier still “is the farmer’s notebook and three-ring binder.

“The thing I see being in Silicon Valley I there is a digital divide,” he said. “By and large, the rest of the world … is moving faster because they’ve got more data available to them.”

Most farms still have back-office employees who enter data onto Excel spreadsheets manually, said Megan Nunes, founder and CEO of the farm data firm Bountiful. Data comes in all sorts of forms, she said.

“It is true that agriculture works with lots of data, but it’s also disparate,” Nunes said. For some crops, such as almonds, it can also take a long time to develop.

“With almonds, it can take eight months,” she said. “Imagine harvest, and you don’t know for eight months what your yields are. You’re already into the next crop year.”

Nunes said she’d like to see more real-time production data from the field in specialty crops, down to row-by-row readouts. Azevedo agreed, noting that he’d like to receive more “passive feeds of information” so that he doesn’t have to constantly enter a username and password and pore over data.]

Growers also need a better way to compile all the disparate information, he said.