Clever idea from Japanese firm SoftEther, whose press release is available in English only via machine translation as of this writing. The video pretty much conveys the idea, however. QUMA is rather like an artists figure-drawing mannequin with sensors in the joints that report all the articulations through a USB cable. Appropriate software can then position a characters rigging to match, which seems like it would be both faster and more intuitive than dragging bones around a screen with a mouse.
Replies
As for animations, the thing isn't really designed for that, primarily posing rigged models, although i'm guessing you could use various poses as keyframes in an animation and go from there.
I've been wondering when something like this would come out. It looks absolutely perfect for refining and keying in major keyframes, then moving to the curve editor and dope sheet to clean up primary inbetweens and poses.
Its cool, and novel, and I just don't see the point unless its retailing for $10.
Otherwise pick up a digital version of "the art of rigging" (going for $20) spend a few hours putting the knowledge into practice with something more robust.
There used to be one made available from a game company, a blue soldier of somehow cartoonesque proportions IIRC. But I want to find something a bit more robust with good controls and customizable proportions in order to easily "hack" rigs for Mudbox models using Skinwrap...
The workflow would be to export the lvl0 or lvl1 to max, deform a continuous rigged "mannequin" to fit the porportions of the mudbox asset, pose the mannequin, drive the mudbox basemesh from it with either a skinwrap or a weight transfer, and export the posed basemesh back to mudbox to apply the pose as a layer.
(Don't laugh, I am actually investigating the use of Poser to do that!!)