Robot Balance

[ysabetwordsmith]

This post claims that designers made a robot with better balance by modeling human motion.  I don't think it's the software; I think it's the hardware.  Almost everyone seems to make the mistake of top-heavy robots with the powerpack on the back, and they fall over because the build is inherently unbalanced.  But look at this one.  The powerpack is slimmer and the knees are bent.  This lowers the center of gravity and places the powerpack properly over it.  Me, I prefer to put them in the feet, but this is much better than other extant models.  \o/

Comments

[siliconshaman]

If you're building humanoid robots, why not place the bulk of the mass where the bulk of the mass in a human is? Just above the pelvis.

Well ...

[ysabetwordsmith]

It makes the center of gravity higher, but that's not a bad idea. Maybe put the batteries in the boots and the motor in the pelvis. Just for fucksake don't put heavy parts on the poor robot's shoulders, or it will fall down and go boom.

Re: Well ...

[siliconshaman]

Human gait relies on an 'interrupted pendulum' motion, which means you actually need a fairly high CoG. The motors should be distributed, if you're talking just the legs, that the bulk of the actual 'muscle' mass is in the thighs. So motors in the upper legs, batteries in the gut just above the pelvis...

Speaking as someone who's actually done this... [albeit, the robot was only 18" tall though...and mostly built of lego parts.]

The difficulty lies when you try to make your robot autonomous... the 'brain' makes up a significant amount of the 'bots mass, enough that it tends to throw off any mass distribution calculations, so you end up shifting something else, liek teh power pack, to someplace less optimal.

Re: Well ...

[ysabetwordsmith]

>>Human gait relies on an 'interrupted pendulum' motion, which means you actually need a fairly high CoG. The motors should be distributed, if you're talking just the legs, that the bulk of the actual 'muscle' mass is in the thighs. So motors in the upper legs, batteries in the gut just above the pelvis...<<

Interesting point, and you also need arms heavy enough to counterswing effectively.

>>Speaking as someone who's actually done this... [albeit, the robot was only 18" tall though...and mostly built of lego parts.]<<

Go you!

>>The difficulty lies when you try to make your robot autonomous... the 'brain' makes up a significant amount of the 'bots mass, enough that it tends to throw off any mass distribution calculations, so you end up shifting something else, liek teh power pack, to someplace less optimal.<<

To make it work you need good materials. That includes a small, light, powerful brain and batteries that are also small, light, and powerful enough to run the robot.

I will say I'm damn impressed by the cheetah. Even though it lacks a flexible spine, the legs work quite well, and its jumping algorithm is epic.

Re: Well ...

[kengr]

Most batteries won't take well to the constant jolting. Plus the longer cable runs waste power. And make it more likely that the cables will break.

Shorts of cables attached to modern high power density batteries are a *bad* thing.

Look at all the e-cig accidents.

Re: Well ...

[peoriapeoriawhereart]

What are the health ramifications of Tesla's distributed current (that might not be its actual name)?

Re: Well ...

[kengr]

No idea. I assume you mean Tesla's scheme for some sort of "broadcast power".

We don't know how he intended to do it. Well, there are a lot of theories, but none of them are workable.

It depends *majorly* on how it would be done...

Re: Well ...

[peoriapeoriawhereart]

Yes, that's the thing. I didn't know that we didn't have his practical thoughts/enough direction.

Re: Well ...

[zeeth_kyrah]

Natural harmonic resonance between similar antenna structures, I imagine.

Re: Well ...

[ysabetwordsmith]

Those are technology problems, not physics problems. The whole field of robotics is about how to make our technology do new things that are hard to do.

Making physics do something new requires getting into a new area of it, which is much much harder. *ponder* But we haven't done much with zero-G robotics in the sense of things that move around, rather than a flying computer. If we didn't have to worry about gravity making a back-heavy robot fall over, it would be a lot easier to make them work.

[zeeth_kyrah]

I tend not to put batteries in the feet, they take too much impact, so instead I tend to go for articulated feet. Batteries go just above the hip joints most of the time.

[we_are_spc]

Yeah, an' den (Cause mah world go this technology) ya can use the friction from the articulations help keep it goin'.

Batteries in eithah the hips oh the thighs. Problem is, how ya make it walk like we do. If ya don't mind it not, could work well enough.

(WHy yes, this be one of mah interests. Much o'my knowledge don't transfuh heah, but some of the basic shit do.)

--Isaiah~