Robotics startup Mecka AI has acquired Docula, a Vancouver-based company, as part of an effort to build what it calls the "data layer" for robotics, according to BetaKit.
The deal brings a three-person Canadian AI startup into the majority-Canadian Mecka team, BetaKit reports. The stated goal is to help Mecka sift through what the report describes as "petabytes" of motion data used to train robots.
Why does motion data matter? Robots learn to move, grip, and navigate by drawing on enormous volumes of recorded movement. But raw data at that scale is only useful if it can be organized, filtered, and made searchable. By acquiring Docula, Mecka appears to be betting that the hard problem in robotics is not just building machines, but managing the vast streams of data that teach those machines how to act.
The specific financial terms of the acquisition were not detailed in the available reporting from BetaKit.
The move fits a broader pattern in the AI industry, where companies increasingly treat well-organized data infrastructure as a competitive advantage rather than an afterthought. It also marks another instance of consolidation among smaller Canadian AI firms, with talent and technology from one startup folding into another.
Why it matters: As robots move from labs toward real-world use, the companies that can efficiently turn massive amounts of motion data into usable training material may shape how quickly — and how capably — those machines learn.