Feds back $108-million push to automate farming

A consortium of groups, including three from Alberta, want to usher in a new era of advanced ag technology

A new consortium has been given nearly $50 million in federal funding to find ways to bring cutting-edge technology to Canadian farms. Among the tech being focused on is self-driving equipment (pictured is DOT Technology’s autonomous platform), robotics, artificial intelligence, data analytic, hyperspectral imaging, traceability systems, and “smart farms.”
Reading Time: 4 minutes

It may not seem a priority given the weather and trade challenges farmers are facing — but now is the time to launch a new era of smart farms, AI, and robotics, say backers of a new $108-million tech fund.

Ottawa is giving $49.5 million to the Canadian Agri-Food Automation and Intelligence Network (CAAIN), and the consortium expects to attract another $58 million from industry partners to “accelerate” the development and adoption of cutting-edge tech on farms.

The ag sector needs to move forward at the same time as it deals with the trials of today, said Cornelia Kreplin, an official with Alberta Innovates, a provincially funded corporation that is co-leading the CAAIN initiative.

“It’s the same old trope we always hear: If agriculture is going to feed a world population of 10 billion people by 2050, then we are going to have to increase agricultural production by 70 per cent,” said Kreplin.

But to make that happen, agricultural powerhouses such as Canada need to invest in what some may consider “blue sky” ideas such as artificial intelligence and robotics, she said.

“(In Canada) there is little agricultural land that can be brought into production,” said Kreplin. “We are going to have to increase yield on the same land base while we use less water and meet consumer demands for environmental sustainability.

“How are we going to break through that yield plateau? One of the answers is in automation and another is in utilizing data much more effectively than we have in the past.”

Improving the productivity and precision of agriculture through technology will play a big role in CAAIN’s research.

“We want to look at how you can apply seed, fertilizer and pesticides in a way that is more precise than on a quarter-section basis,” said Kreplin. “We want to manage that on a nine-square-metre basis and even smaller if we can.”

The consortium is targeting four areas to show what can be done with technology, including establishing “a network of Canadian smart farms,” in which data collected by GPS devices, soil testing, remote imaging and by other means is used to make decisions about crop and animal management.

The group’s other three goals are to use robotics and automation to reduce labour requirements; use data analytics to, for example, predict disease or pest outbreaks; and use traceability systems (and possibly blockchain) so farmers can be “paid fairly for premium products.”

Bringing products to market

CAAIN is a network of eight partners. Along with Alberta Innovates, the others are the University of Guelph’s Vineland Research and Innovation Centre (the other co-leader of the project), Olds College, Lakeland College, Edmonton traceability company TrustBIX, Saskatchewan autonomous platform maker DOT Technology, B.C. satellite and robotics company MDA Systems, and Guelph auto parts and industrial manufacturer Linamar Corp.

The consortium’s goal is to research, develop and improve on-farm applications of technology such as artificial intelligence, advanced sensors, hyperspectral imaging, blockchain and other cutting-edge technology.

But the focus is not just on technology but to find a “path to commercialization” for automated and digitized farm products, said Kreplin. The network occupies a space about halfway through this development pipeline, researching ways to iron the kinks out of projects in a grey zone between initial research and marketability.

“At one end of the pipeline you might have the university side where there are thousands of ‘blue sky’ research projects,” she said. “Some may move through this pipeline to the other end where you wind up with either one of two things: Information that producers can utilize to make a difference on their operations or a product in the marketplace producers could use in order to improve profitability on their operations.

“You’re not going to get continual improvement unless you invest all the way along that pipeline.”

Producer resistance remains

Marketing automation to producers can sometimes be a difficult task. While there are always early adopters, reactions to automated technology among farmers are not always favourable.

In a poll of 432 producers conducted by Glacier FarmMedia (the owner of this paper) in 2017, 70 per cent of respondents ranked investing in autonomous farm implements — such as the self-driving DOT platform — a low priority.

But a producer’s view on things such as data analytics, automation, and other cutting-edge tech often depends on the size and intensity of the operation, said Kreplin.

“You look at those guys who run 30,000 or 40,000 acres and they spend more time on a computer today than they do on a tractor. The reason they do is because they need to make better decisions so they’re utilizing data even in real time.”

On the other end of the spectrum, she said, would be an operation with 50 cows or a couple of sections supplemented by a job in town or the oilpatch.

“Those farms will continue to exist and are really important to the rural fabric in Western Canada. But those agricultural producers who consider farming their principal source of revenue are going to diversify.”

So if the future is self-driving implements planting and harvesting crops with machine-taught computers making precise input recommendations, where does the farmer fit?

Right where they’ve always been, said Kreplin.

“You need to have that knowledge about agricultural production,” she said. “We can improve decision-making by utilizing machine learning and artificial intelligence but you still need the farmer to produce the product at the end of the day.

“Whether it’s canola or livestock it doesn’t really matter; you need to have that wisdom in order to select the right seed and the right genetics. Machine learning might help you narrow down those decisions but they still end up being farmer-based decisions.”

About the author



Stories from our other publications