How AI can help farming become more efficient and sustainable

Content of the article

At Nature Fresh Farms in Leamington, Ont., there’s something new among the rows of tomatoes, cucumbers, peppers and strawberries.

Using thousands of sensors in each greenhouse, artificial intelligence technology helps the farm optimize aspects like lighting, irrigation and harvest time.

“We wanted to use technology for us to grow more, have better-tasting vegetables, and just do more in general,” said Keith Bradley, vice president of information technology and security at Nature Fresh Farms.

Content of the article

Technologies from Intel and Dell help the farm to be proactive instead of reactive, he said, increasing the yield of crops and reducing the use of power and water. It even helps employees have a better work-life balance, he added.

Amid research into the potential benefits of AI for agriculture, farms like Nature Fresh are at the forefront of adoption.

Farmers have used a variety of technologies, with some having adopted high-tech tools such as drones to survey farms and find information about weeds, pests and diseases, said Jacqueline Keena, managing director of the industry-led non-profit Emili. The organization operates Innovation Farms, a “smart farm” where new technologies are tested and demonstrated near Winnipeg.

The next phase of the technology involves AI models using that data to make inferences, predictions and even decisions, Keena said _ and AI allows agriculture to be “hyper-optimized” to a much more specific level than before.

The technology is becoming more sophisticated, moving from simple rule-based systems to large language models, said Rozita Dara, assistant professor at the University of Guelph’s School of Computer Science and director of the Artificial Intelligence for Food initiative.

Content of the article

Content of the article

This has applications for precision agriculture, he said, which involves analyzing data from sensors to make decisions about how much water or fertilizer to use. AI can be used to make more complex decisions that humans have been making for a long time, he said.

AI can help solve problems like labor shortages and climate challenges, said Darrell Petras, CEO of the Canadian Agri-Food Automation and Intelligence Network.

For example, his group invested in a company called Croptimistic, Petras said, which collects data from the field to detect pests, crop color changes, and other potential stresses on crops.

AI “can identify when a stressor is happening earlier than … the human eye can pick it up and then management intervention can happen faster,” he said.

AI also has the potential to use grain grading in the field, which can help farmers know when to harvest their crops and what to expect when they sell them, Petras added.

It can also be used to reduce the effects of climate change, he said.

Much research on AI and agriculture is done at post-secondary institutions, Petras said, but then needs to be tested in the field. This is often done through a “commercialization vehicle,” he says, whether it’s a startup or an existing company.

Content of the article

There is a network of so-called smart farms in Canada, led by the Olds College of Agriculture & Technology in Alberta, which aims to test and demonstrate emerging agricultural technologies.

One of the farms in the network is Emili’s Innovation Farms.

“We really show how it works in a commercial setting, and in a way it’s a risk mitigation when testing the technology… the full adoption and integration of the new technology,” said Keena, from Emili.

Another smart farm is at Olds College, where Felippe Karp is conducting research on how to develop standards for data collection and processing to build AI models.

AI models are only as good as their datasets, explains Karp, who is a research associate at the college and a PhD candidate in bioresource engineering at McGill University. His current focus is measuring and predicting variations in soil nutrients.

“With this data set, we trained an artificial intelligence model … and used it to predict the availability of nutrients in the soil.”

Content of the article

It takes time to understand whether a new technology or a new approach has affected the crop, said Dara, and this can be a barrier to adoption for farmers.

“Sometimes … it is in the year, in the winter or in several years,” he said.

Farmers often get only “one shot” at a crop each year, Keena said.

“And we can’t ask them to take huge risks to integrate new technologies at scale as part of an operation on something that hasn’t been proven yet.”

“Innovation Farms … overcomes one of the hurdles of people having to be able to see the technology rolled out at full scale and commercially before they can use it themselves.”

Farmers’ trust level is also a barrier, Dara said, especially since AI sometimes makes the decision-making process unclear.

Data is paramount to AI models, he added, but farmers need more incentives to share data to make the technology better.

Farmers can be resistant to sharing their own data, says Karp: “That’s one of the challenges when we talk about developing more complex models.”

But over time, Petras said he’s seen an increase in farmers’ involvement.

“Farmer involvement is critical” to developing AI tools for agriculture, he said, which could include days of field demonstrations, conferences and workshops, he said.

“If they’ve demonstrated that, especially in our backyards through smart farms, then we’ll be a step closer to adoption.”

This report by The Canadian Press was first published June 16, 2024.

Content of the article