Pulse Beat Individual Articles

Protein Industries Canada Utilizing Artificial Intelligence for Variety Development

Miranda Burski, Marketing and Communications Consultant, Protein Industries Canada – Spring 2022 Pulse Beat

 PEA BREEDING ISN’T a new venture for DL Seeds, a company situated in southern Manitoba. They have been selecting and breeding yellow peas for the North American market for more than a decade, with the goal of achieving improvements that benefit both farmers and consumers.

With the increasing demand for plant-based protein products, however, the company felt it could take its breeding program a step further. DL Seeds saw the need for a yellow pea variety that had a higher protein content than that in varieties already on the market, and felt the best way to develop this benefit would be through strategic collaborations that push the boundaries of tech and challenge the status quo.

“Genetic breeding challenges are at the forefront of our thinking,” General Manager of DL Seeds Kevin McCallum said. “If we are going to develop new pea varieties through the Protein Industries Canada (PIC) project, DL Seeds is going to have to develop germplasm that pushes beyond traditional breeding techniques and open our minds to new ways of thinking how we can benefit peoples’ lives.”

In summer 2020, it was announced that DL Seeds was partnering with SeedNet and Sightline Innovation, with a co-investment from PIC, on a project that would use artificial intelligence and data-trust tools to develop new high-protein yellow pea varieties. The new varieties will be based on parent breeds chosen by DL Seeds; their higher-protein content and adaptation to Canada’s climate will be achieved using Sightline’s artificial intelligence and data trust tools, and SeedNet’s network of partners will be utilized to produce and distribute the seeds.

The combined skill set makes for an improved variety development process that DL Seeds feels speeds up its usual process.

“Any tool that can be used successfully in selecting new varieties is valuable to a breeding program,” McCallum said. “It could take up to 10 years to develop a new pea variety from start to finish. With the new AI learning, we are hoping to decrease that time by 20 to 30 percent, at least. That will allow new varieties and traits in those varieties to hit the marketplace faster than before.”

New varieties on the market, he added, means good things for everyone along the value chain, from farmers, to processors, to consumers.

“We (want to) develop a high protein pea variety that the pea processing companies and our distributor SeedNet can contract directly with farmers in a closed-loop production program,” McCallum said. “That way everybody in that value chain can benefit from each other based upon the extra value the higher protein content brings to the equation.” ■