next-gen servo-motors

[weed_eater_guy]

about the physical nature of robots, if there were to be humanoid-like robots, or robots that were designed to last for a long time without repair, better motors other than servo motors (with gears and everything that can easily errode or jam from extended use). One idea i heard about was a material that behaves like muscle fiber: give it current and it contracts. I couldn't help but notice how this looked like what they put on the guys in I,Robot, the bundles of fiber on thier arms and shoudlers. How realistic is this stuff? is it really better than servo or can it break down just as easily? just thinking it'd make sence to model motor systems off of nature's tired-and-true method rather than a simple motor and gearbox.

Dan

 

[planaria]

ive been tainted by the beautiful new ghost in the shell 2 movie..

in it the androids had circular joints that connect to the body somehow .. so i speculate as to how this works ..

for one thing i think a type of circular joint that uses magnetic fields might work best because there are no parts to wear down .. the limbs could be made of a magnetic material so when the joints are active the circular joint could alter its magnetic field to produce movement .. due to the fact that the limbs would be attracted to the magnetic field due to their material.

one problem might be these androids wouldnt be very strong.. so possibly superconducting materials might enhance the strength of the magnetic field maybe? im not sure..

im still thinking about this but the idea of using circular joints somehow with magnetic fields i think could work well .

 

[weed_eater_guy]

sounds like the motor on a hard drive, a disk with tiny magnets on it revolves inside a ring with alternating electromagnets on it, re-pulsing like a maglev train. very efficient, very fast, but like you said, very low torque, not very strong. The problem is the magnetic field isn't concentrated solely on where it needs to be, the energy goes all around the electro magnets so only part of it hits the magnets on the spining disk. Also a problem, if you're talking about fitting these motors in joints, the leverage is well against the motor, making it need more torque, thus more power. If we could find a way to concentrate the field, and concentrate it VERY efficiently, maybe this system would work, but if we couldn't do it, and we just used superconductors and upped the wattage, the poor android would have to go perfectly limp if he wanted to use a radio, or watch TV, or make a download off a wireless LAN or anything! if any of his joints were tensed, meaning an EM field was generated, it would be a VERY VERY powerful EM field, maybe enough to yank metal objects toward it (coins, paper clips, beer cans, keys, etc.). But that's if you wanted a strong android, like human-strength. maybe something less labor intensive, and smaller (sparing appendage leverage), would work. But then again it might make more sence to just put a solenoid in.

 

[craterchains (Norval]

Hydraulic bulb "muscles" are elongated when empty and constrict when filled with liquid.

Just a thought.

 

[Clockwood]

Craterchains: But then you need to have a series of pumps, a significant fluid store, and a system of valves and relays throughout the robot. Whatever you do, there will be tradeoffs.

 

[Pete]

The real advantage of biological muscles is that they can self-repair.
If I had to guess at what the future holds, I'd lean toward self-repairing servos rather than fibrous muscle type mechanisms... depending on the application. I guess we'll see both sooner or later, plus others yet unheard of.