A Beijing company that helps robots learn by watching people has raised $100 million. GenRobot AI closed the Series A+ round in the third quarter and disclosed it on November 2. Its business is not building robots. It collects the examples that robots learn from.
What the company does
GenRobot AI makes equipment and software for recording human demonstrations. Its gear captures what a person sees, along with hand and body movements. The tasks come from homes, factories and warehouses.
A platform called Matrix then takes those recordings, lines them up, cleans them and labels them. Robot developers can use the result to train and test their machines.
Who is backing it
The investor list is long and mixes industries. It includes carmaker SAIC Motor, Malaysia's Eastworlds, Baidu, Shoucheng Holdings and Galaxy General Robotics. Financial backers include GL Ventures, CITIC Goldstone, CCB International and Huatai Zijin.
Why it matters
Robots learn from good examples, much as a new hire learns by watching a skilled coworker. Good examples are hard to gather. Companies that supply them may help decide how soon useful robots reach homes and workplaces.
What is not known yet
The material does not say how the company values itself, how many customers it has, or which robot makers use its data. It also does not say how the money will be spent.
The takeawayTeaching data may matter as much as the robot itself, and GenRobot AI is a $100 million bet on that idea.