Having not used nitinol before, I've got a few questions. First of all, how long does it take to contract? Walkers are slow as is. Secondly, how do you know if it's contracted all the way? I imagine you would have to use some sort of optical sensor for that. Isn't if pretty important that the walker know the position of each leg? Seems to me like Nitinol is pretty "gee whiz" cool, but isn't real practical for this type of project.

It would be really cool if they could make it mobile and have it run off of an iPaq.

The robot looks like a Stiquito robot. The Stiquito robot failed miserably because it just wasn't robust at all and was a very slow robot. I think the Robot Store started their business by selling Nitinol and Stiquito robots, but I believe they finally discontinued the Stiquito because it was so frail. I believe they still sell Nitinol wire, though.

Nitinol for a robot is not ready for prime time for a number of reasons:

1. It's very slow. It takes like 1 to 2 seconds to contract. If it's power cooled it could go faster, but really it's very slow and who wants to have a power sucking peltier junction on every nitinol wire they have.
2. It takes gobs of power. Any serious robot using nitinol would need mega watts of power to keep their robot alive for any length of time. Nitinol contracts as it heats up. Anything that heats up, you know is converting electricity into heat and that takes gobs of power to do. So basically Nitinol sucks batteries dead in minutes.
3. It has no strength. You'll notice that any robot made with nitinol as the propulsion, is very light (made of styrofoam or foam core) and almost always has the CPU and the Batteries off of the base tethered from somewhere else because Nitinol just can't handle the weight. Sure you can parallel a bunch of Nitinol wires for added strength, but you'd need a car battery to power it for just a short time..

What's Nitinol good for?

1. Probably Nitinol is not good for robot propulsion.
2. It might be good for a robot in the refridgerator! The NASA Mars rover had a nitinol wire on it (it's cold on mars) to move some antenna or sensor thingy (I don't remember which) but it basically was a light object to move and was a limited space.
3. It might be good for a light robot that is plugged into a wall AC (not batteries) and a CPU tether (not onboard cpu).

What should people use for propulsion on a robot rather than Nitinol?

1. DC motor
2. Hacked Servo
3. Servo
4. Just about anything is better than Nitininol

I've never used Nitinol, but it seems to me that it has one big win: it's tiny, and can fit into ridiculously small spaces. You can also hook up many individually controlled segments in a very compact package. Servos have a very bulky form factor, and they're noisy and expensive, to boot.

The only other solutions analagous to Nitinol that I know of are pneumatic or hyrdraulic actuators (such as air muscles), which require heavy compressors or tanks. I suppose solenoids could be used in a similar capacity?

True, nitinol could fit in little spaces, but pneumatics wouldn't be too bulky. It would be bulky for a little stick bot like this, but not for anything a little larger. Have you seen Lego or FischerTechnik pneumatics? They are very compact. Also seems like you would get more force out of them. One problem with pneumatics is that you assume that the air is clean, so it might not be good for something like a fire fighting robot.

Came accross this link when doing some research into lightweight linear actuators. First time I have seen niti used in something that wasn't either a toy or a replacement for an explosive bolt.

The way they are using it seems to have negated several of the disadvantages. Of course underwater weight is less ...