Like a multi-dimensional Wikipedia for the bot mind, the “Robo-brain” project, led by robotics researchers and funded by tech giants and the US military, aims to give machines the ability to access and learn from public information on the Web.
The system is currently chewing through a mountain of public data — 1 billion images, 120,000 YouTube videos, and 100 million how-to documents and appliance manuals according to a release.
Robo Brain’s task is to synthesize all that information into a format that robots can access on demand.
“If a robot encounters a situation it hasn’t seen before it can query Robo Brain in the cloud,” said Ashutosh Saxena, assistant professor of computer science at Cornell University, in a release.
As a demonstration, Popular Science reports, the researchers made a request for affogato:
The bot, a two-armed, highly-dextrous PR2, queried the system, and discovered that affogato was an Italian dessert composed of ice cream and coffee. Without any human nudging or intervention, the robot located the coffee, figured out how to get it out of a dispenser, and poured it over the scooped ice cream.
In addition to Saxena at Cornell, researchers at UC Berkeley, Stanford University, and Brown University are part of the collaboration, due to include researchers from ten other universities in the fall.
Google, Microsoft, and Qualcomm, as well as the Office of Naval Research and the National Science Foundation are among the organizers funding this project.