$199.5 million

A research group at the University of Toronto called the Acceleration Consortium is getting that much federal money to create “self-driving labs” to devise new materials, from plastics to pharmaceuticals. The approach combines chemistry, automation and artificial intelligence.

O

Circular experiments are at the heart of the idea.

Like the occupant of a self-driving car setting a destination, humans with a self-driving lab can define what qualities they’re after—a conductive material for a wearable sensor, say, or a biodegradable plastic made from specific inputs—then let an AI guiding automated equipment figure out how to make it.

One experiment after another can vary the inputs, sequences of reactions, and conditions like time and temperature, while the AI “learns” from each iteration, working the new findings into simulations, deciding what to try next and doing it.

The Acceleration Consortium’s director Alán Aspuru-Guzik outlined the thinking in a paper he co-published in 2020: “The traditional paradigm of design, synthesis, characterization and testing is changing into a more integrated pipeline under a closed-loop approach.”

$1 million

Traditional research and experimentation can be intensely time-consuming. Self-driving labs might cut the time and cost of devising a new material from 20 years and $100 million to one year and $1 million, the University of Toronto said.

7 compounds, 30 days

A recent Aspuru-Guzik project used elements of a self-driving lab to develop a molecule with early potential for combating a common form of liver cancer. It took just under a month and only a handful of experimental chemicals. 

“We’ve already established proof of principle,” said Leah Cowen, U of T’s vice-president of research, innovation and strategic initiatives. “We can do this, we’re doing this, and here, we’re going to bring it to scale.”

The money for the Acceleration Consortium is one of several grants Innovation Minister François-Philippe Champagne announced today from the Canada First Research Excellence Fund. Two previous rounds have provided big money to Canadian universities, tens of millions of dollars per project.

Until now, though, the biggest was a previous grant to U of T, $114 million for work on regenerative medicine.

“Talent is a huge piece of this,” said Cowen, herself a professor of molecular genetics specializing in fungal diseases. “It takes brilliant minds. It takes great people.”

The university will add more than $100 million to expand a building for multiple new laboratories, which will be outfitted for different types of research, she said.

Less than 3 years

“Setting up a robotic laboratory is costly,” Aspuru-Guzik and his fellow scientists wrote in that 2020 paper, and for results and products to spread widely the labs would need “institutional support across academia, industry and government, along with financial incentives.”

The Acceleration Consortium already has an array of academic, industry and government partners, and today, not three years later, it has a lot of financial incentive.