Though just a design exercise at the moment, "Foldy" is offering its makers some valuable tips on how humans, robots and computers can interact in the most efficient manner. See, not every futuristic scenario involving robots has to end up resembling Terminator!
Foldy - you gotta love that name - was cobbled together by Professor Masahiko Inami and a team of students working out of Keio University in Tokyo, Japan. The team is developing Foldy in association with the Design UI Project, ERATO, Japan Science and Technology Agency (JST for short).
Foldy wasn't really designed to be a retail-ready, laundry-folding robot though the JST team's experiments may some day lead to just that. For the time being, Foldy flunks at his primary function - sure it can fold laundry but by the time it's done folding, you've run out of clothes to wear and need to wash another load. Here's a video of Foldy doing what Foldy's do:
As you can see, you'd have to be pret-ty, pret-ty lazy if Foldy can beat you in a laundry-folding contest. Then again, we all know at least one person who'd lose such a competition.
In a nutshell, a ceiling-mounted camera feeds an image of a flattened item of laundry centered beneath it. You're already halfway there, if you ask me - but I digress. A computer reads the camera's feed, plans out the best way to fold the item, and sends the info to Foldy who then rolls up and does his thing. As JST states in summary, "The interface consists of a simple garment folding simulation mechanism for detecting actions that would be impossible for the robot to perform, and it returns visual feedback to the user. The robot performs an actual garment-folding task by completing the instructions set by the user."
Naturally, Foldy is rather small and uncomplicated, looking somewhat like a cross between Wall-E and my son's latest LEGO creation. The guys & gals at Keio U must be saving the complex stuff for Skynet. (via Robonable and Plastic Pals)