balancing pole robot

--Check out the MIT Leg Lab.

Try looking for the "inverted pendulum". Two variants (two links each) are sometimes called the acrobot and pendubot.

Here are a couple relevant papers:

Daniel

Very interesting, but not quite what I wanted - I'll keep looking.

Mark

Yes. Marc Raibert's 1, 2, and 4-legged robots from the late 1980s and early 90s were all based on the basic deisgn of a 1-legged balancing bot. Used 2-D gymbals for x-y, plus an explosive pneumatic thruster for z. Check out the following for lots of the theory ....

- dan michaels

Most of those hop. I just want it to stand there and stay balanced. But they are very interesting.

A pole in a coffee can filled with cement balances pretty well.

Seriously, *something* has to move or rotate in order to achieve dynamic balance. Why don't you provide a little more of the design concept so that the answers aren't so hit or miss. A "1 legged pole" sounds like you wanted a hopping leg. A pole is a pole, and a leg is a leg.

A stationary Segway or a two-wheeled balancing robot does exactly what you want, if you think about it. It's a "pole," and though it happens to be on two wheels, when it's not being driven, it "stands there and stays balanced." However, looke closely and the wheels are moving ever so slowly back and forth to provide the balance.

-- Gordon

Mark wrote:

There is no question that trying to convey ideas about physical objects can be quite a challenge. The search engines don't help much if you don't know the exact words to search for. But, in this case, there may not be any good search terms.

Perhaps a new way to word what he is looking for would be "pole sitter" robot. I think there is a circus trick with that title. However in that case the are on a flexable pole which is rigidly connected to the ground at its base.

I have seen people balance on a pole not connected to the ground. In 2 dimensions, I've done it myself when I was on crutches for a short period of time. Obviously using a gyroscope to exert force is one method, but I don't think it is necessary because people can balance themselves on a pole without one. I have a hard time picturing what kind of physics were involved.

Joe Dunfee

How about a helium balloon attached to a balsa stick such that the balloon can hold the stick straight up, but not pick it up totally from the ground. Paint a robot face on the balloon.

I was thinks in terms of say a 16 inch vertical rod. Attached to the top are 2 servos that can extend weights in 4 directions. Appropriate sensors and a control algorithm sense that the robot is falling, and extend the weight to counterbalance appropriately to maintain balance. Obviously, slamming the side of the robot would make it fall, but I think it could be made to maintain balance that way. Or perhaps not?? It would be like a bipedal robot standing on one leg.

Mark

The center of mass of the robot would probably stay the same even when the weights are moved (action=reaction), causing it to continue to fall.

Balance needs to come from below, not above. Haven't seen too many tightrope walkers use a balancing pole *over* their head.

In Japan they used to have contests on balancing pole robots. They did use wheels as the concept was similar to a juggler who balances a long stick on his fingertips. By moving the robot laterally the pole was kept upright.

-- Gordon

True, but they don't have them at their feet, either. Oh well, just a vague thought for an interesting project.

Mark

UK Technogames a couple of years ago had a bicycle steered by the use of thrusters. Essentially propellors on the front wheel.

Moving masses about would be problematic for the reasons given by another poster unless you had a relatively large foot - that would defeat (hey a pun :) the purpose.

If you could make a set of thrusters work that would be neat. Especially if you used three just for the extra complexity.

Pete Harrison

When you think about it they use their entire body (some don't even use a pole). In any case, the pole is usually at about the center of mass on a human, which differs somewhat by sex and body shape, but is roughly the mid-to-lower abdomen. If a Polebot where to use a similar technique, it would probably have its balancers at least at its center of mass.

-- Gordon

I'm pretty sure you can balance from above with the right configuration. It works because there would be lateral resistance between the pole and the floor. The system would push against that to move its center of gravity opposite the direction of fall.

A weight that moves in the lateral directions at the top of the pole wouldn't work because the pole would just move and the center of gravity wouldn't. But if you had a pole in two sections with a pivot in the center, and a weight at the top, I think that might work. But it might take three sections.

Inversely, if you had a rigid pendulum hanging from the top of the pole with weight at the end it might work.

I don't think walking on a tight rope would be possible if you couldn't balance from above (though some of that works because you can push the rope left and right with your actions if there is enough slack in the rope).

You could do it for sure if you used air resistance to push against (like when a tight rope walker cheats by using an umbrella). But I think it's physically possible by just pushing against the floor in the later directions if you have the right configuration system.

I've never heard of a robot doing it. Sounds like a fun challenge. Might have to spend some time on the physics to see if it's really possible.

I thought long and hard about exactly this kind of balancer a few years ago, and came to the conclusion that you could do it if you had a bi-directional knee joint with the motors, and at the top of a further shaft, the battery pack. That way, moving the knee in one direction displaces the motor mass in one direction and the battery mass in the other, with perhaps similar masses but different angular moment of inertia.