Skip to main content

Tyler Hamilton works with cleantech companies from across Canada as an adviser with the non-profit MaRS Discovery District in Toronto.

There's something growing within the operations of precision-agriculture company Farmers Edge Inc., and it's not your typical prairie cash crop.

The Winnipeg-based company isn't harvesting wheat, corn or lentils. It's farming data.

"We're ingesting hundreds of terabytes of satellite imagery data, and we've got thousands of sensors in fields, all providing us information on a per-second, per-minute or daily basis," says Kevin Grant, the company's chief technology officer.

That information, he says, is gathered from thousands of farms across Canada, the United States and overseas, and includes everything from soil and weather conditions to the performance of farm equipment and evidence of pests and rot. Sometimes it comes from Farmers Edge's own weather stations and in-field sensors; sometimes from third-party databases.

The end game is to increase crop yield and lower costs by helping farmers make better decisions about when and exactly where to plant, irrigate, fertilize and harvest, and by allowing for the pro-active maintenance of equipment.

As the amount of data piles up, however, companies such as Farmers Edge are increasingly turning to machine-learning technology and other flavours of artificial intelligence (AI) to help spot patterns or trends that might otherwise go unnoticed using conventional agronomics and human labour.

"This is about disrupting what's been done in agronomy for the past few decades. There's a perfect approach you can take to every field, and what we're doing is trying to tease that perfection out from the data we collect," Mr. Grant says. "In essence, we're asking our machine learning algorithms to do what simply wouldn't be feasible for us to do on our own."

It wouldn't be possible, really, without the Internet of Things and its rapidly expanding network of low-cost, data-gathering sensors, which touch everything from dairy farming to aquaculture.

Fredericton-based SomaDetect, for example, uses optical sensors to monitor fat, protein and antibiotic levels (and other biochemical markers) in milk as it flows from a cow's udder. As more data is collected, machine-learning algorithms are trained to instantly spot abnormalities that could ruin a large batch or indicate an infectious disease. Armed with this information, farmers can take action to protect the rest of the herd and preserve batch quality.

XpertSea, a precision aquaculture venture in Quebec City, has developed a "smart bucket" that uses computer vision technology and machine learning to count, categorize and determine the size of species sampled from massive shrimp and fish hatcheries. It's accurate and superfast, helping fish farms do more with less labour, prevent disease from overfeeding and look at historical trends that can reveal how to improve operations and grow bigger, healthier – and more profitable – fish inventories.

Another company, Deveron UAS of Toronto, uses sensor-equipped drones and machine learning technology to count and analyze the size distribution of pumpkins, allowing farmers to accurately estimate yield per acre before negotiating sales agreements with retailers.

The agriculture industry isn't the only resource sector in Canada that has been shaken awake by the potential of AI. In both mining and energy, there's a recognition that to stay globally competitive industry will need to embrace the future that AI represents.

"It will help us do two things faster," says Michelle Ash, chief innovation officer at Barrick Gold. "One is to understand solutions to our current questions and, more importantly, know the answers to questions we haven't yet thought to ask."

In Vancouver, a company called MineSense is already using machine learning to analyze data collected from sensors installed on shovels, buckets and conveyor systems that move ore – in this case, copper, nickel, zinc or iron. The technology scans for electromagnetic signatures in the ore and uses X-rays to instantly detect metal concentrations in each shovel load or as ore moves along a belt.

"The system uses machine learning to look for patterns in the data we collect and gets smarter as it predicts the grade and condition of ore," CEO Jeff More says, adding that the technology allows his customers – currently large copper mines in British Columbia and Chile – to find pockets of ore that can't be seen in a general mine plan. "The more precise we are, the more pockets of good ore we can find."

Higher precision and the ability to adapt to changing ore conditions in real time results in reduced use of energy, water and chemicals and less waste from ore processing, Mr. More explains. "The idea is to identify and eliminate waste as early as possible."

The bottom line is that efficiencies enabled through AI means we can get more by using less. We can grow and process more food using less energy, water and fertilizer; we can extract more minerals and metals from the ground using less fuel and producing less waste.

Expand this to the energy sector and the impact gets even more compelling. AI is helping monitor the integrity of pipelines, speed up the development of nuclear fusion power, and make it possible for smart homes and intelligent infrastructure to "negotiate" with the power grid. It's allowing us to squeeze the most we can out of solar and wind farms, and predict when power plant equipment needs maintenance so repairs can happen before failures.

Companies in these sectors that don't yet have a strategy to leverage the power of AI need to get moving. The change won't happen overnight, but the planning needs to start today.

Follow us on Twitter: @globeandmailOpens in a new window

Report an error

Editorial code of conduct

Tickers mentioned in this story