Montreal-based automation platform vendor Vention this week announced the debut of a physical AI and industrial robotics laboratory in Canada with the goal of advancing robotic manipulation from cutting-edge AI research to real-world, scalable manufacturing deployment.   

Research at the lab, also located in Montreal, will span applied manufacturing use cases and industrial data collection and post-training foundation models for robotics and will build on intellectual property already generated, according to material from Vention. 

See also: Has physical AI gone ‘mainstream’? One new survey says yes 

The main goal is scalable deployment through validating new physical AI capabilities that can withstand the reliability, cost, and variability requirements of real production lines.

And as the underlying models mature, Vention said it expects the complexity costs of deployment to drop, which it said could allow more manufacturers to use the technology on factory floors. 

According to the Vention release, the lab’s research agenda combines industrial data collection, robotics control, motion planning, classical computer vision, vision foundation models, learning from demonstration and reinforcement learning, aimed at manufacturing tasks that are complex and unstructured.  

See also: From labs to factories: Physical AI’s growth expected to explode 

"What makes this lab different is the loop we've built: academic research feeding directly into live production problems, and production feedback feeding back into the research," said Jimmy Li, Vention’s director of physical AI, which generally is defined as artificial intelligence embedded into manufacturing hardware or augmenting robotic factory processes. 

"Our clients aren't waiting for a finished product to test; they're in the room while we build it. That's unique in this field, and it's what lets us move faster from a research result to something that actually runs on a factory floor."