Human face of 'androgynoid'

K-Bot is the face of the future of robotics. It is a mask of elastic polymer placed over $400 worth of stereoscopic cameras, wiring and tiny motors. And because K-Bot can sneer, scoff, and smile warmly, it is the ultimate in automaton sophistication.

It is the first robot with 24 artificial facial muscles and 28 facial expressions, and eyes that could follow you round the room.
And it weighs little more than a real human brain.

K-Bot, for the moment, has no personality and no gender. “It is sort of an androgynoid,” said David Hanson, a doctoral student at the University of Texas, Dallas, as he showed his creation at the American Association for the Advancement of Science meeting in Denver. He built it out of his research budget with a labour force of one and a half. The half is Kristen Nelson, a laboratory assistant, also of Dallas, whose face was the model for the mask.

K-Bot may yet acquire a personality, and a proper name, because it is programmed to learn in a natural way, and to recognise and respond to people—and then perhaps to become a research tool in AI laboratories.

“This could become an extremely useful tool for medical therapies,” said Hanson. “Say you have somebody with cognitive dysfunction, like brain damage or autism, and you want to train them nonverbally in a natural way; this machine could present a simplified and codified version of nonverbal interaction.”

Making faces, or running and jumping, are easy for humans but difficult for robots, said Yoseph Bar-Cohen of the Nasa jet propulsion laboratory in Pasadena, Texas.

The ambition is to create a generation of robots that used artificial muscles as easily and as comfortably as do humans, that can sense pressure and respond. “I have a challenge for the world engineering community,” he said. “Make a robotic arm that could arm wrestle with humans—and win. An arm like that, connected to the brain, could make disabled people very capable.” - Guardian Unlimited Â

Client Media Releases

Teraco achieves global top 3 data centre ranking
PhD graduate tackles strike participation at Transnet port terminals