Robots have a tendency to move rather a lot like, um, robots. How dare they. The smooth and natural motions that we humans are so proud of comes from a combination of many different motions all at once: if you pick something up, you're generally not just using your arm like a robot does, but rather, subtly moving your arms, wrists, hands, torso, and even your head. With a new movement algorithm, iCub is learning to move in a much more human-like manner, even with complex motions.
Yes, at least, iCub has learned how to put things on its head. Brilliant.
You may have noticed that the beginning of the video that it was the recipient of the AAAI 2013 Shakey Award for Best Student Video. What's a Shakey? Why, I'm glad you asked! Meet Shakey, who was pioneering task-planning and route-finding at the Stanford Research Institute back in the 1960s:
Shakey is now retired, but you can find him chillin' at the Computer History Museum in Mountain View, Calif.
[ IM CLEVER ]
Evan Ackerman is a senior editor at IEEE Spectrum. Since 2007, he has written over 6,000 articles on robotics and technology. He has a degree in Martian geology and is excellent at playing bagpipes.
