Android Actuators, Little by Little

As I mentioned in a previous post, the Nanotech Institute of the University of Texas at Dallas, has been working on solving the problem of "robot muscle"--strong and efficient linear actuators for androids and robots. Ray Baughman is the director of the Institute, and principal overseer of the robot muscle projects at UTD.

In Baughman's latest work, done in collaboration with John Madden at the University of British Columbia, the researchers made actuators out of carbon-nanotube yarns. The yarns are created by first growing densely packed nanotubes, each about 100 micrometers long. The carbon nanotubes are then gathered from a portion of this field and spun together into long, thin threads. The nanotube yarn can be just 2 percent of the width of a hair--not even visible--but upwards of a meter long. According to Baughman, spinning these threads was "like hauling in a fish with an invisible line." In his conference presentation, he described yarns that could support loads 150 times greater than nanotube papers could.


However, much work remains to be done in developing the materials. For one thing, as greater loads are applied to actuators, they can start to exhibit "creep"--that is, they do not completely return to their original state with successive cycles. Baughman says that before these actuators can be useful, creep must be eliminated. "Under load, the cycle is not reversible--you've got a little creep. In most actuator applications, you don't want any creep."

Another key issue is scaling up from thin individual threads. Although the carbon-nanotube muscles can outperform natural muscles on a per-area basis, exerting 100 times the force, natural muscles are much larger, making them stronger. This scale-up issue has been a challenge for artificial muscles in general, which is why they still can't beat human muscles in such functions as arm wrestling, Bar-Cohen says.

Source.

Strong and efficient self-contained linear actuators are one of the main obstacles to making a free moving, unplugged android. The balance problem has been as good as solved by engineers in Japan, and by technology developed by college dropout inventor Dean Kamen.

Rodney Brooks and other progressive roboticists and workers in machine intelligence have made progress in making "robot brains" produce behaviour that is more animal-like, and eventually human-like.

But androids need to be able to ambulate without pneumatic, hydraulic, or electrical cables attached. They should be able to work for several hours without being re-fueled--just as humans can go for several hours between meals. For androids to be able to move in a human-like fashion, they need linear actuators that work in an analogous manner to muscle. And even though the workers at UTD still have far to go to accomplish all that, they are making significant progress.

So tonight when you are gathered with your friends in the local pub, raise a glass to Ray Baughman and his group. They are helping to change the world.