>>15339452>But yeah, practically with existing technology, the aplication of this on robotics wasn't that viable, because it requires a complicated coordination and calculation, or else, the platform will flip out of control.
That's not the case, though. momentum, including angular momentum, is conserved. If you put a humanoid, perfectly still, in freefall, in a vacuum, all possible movements they can peform will only cause the rest of their body to move for the duration that they are moving their limbs. And when they stop moving, they will stop translating. and when they move back to their original stance, their entire body will be exactly where they started.
The only exception to this is joints, or compound systems of joints, that can be spun full circle, passing their original position and repeating the movement without stopping. In a humanoid, that's probably just windmilling the arms around, and, more usefully, sort of hula-hooping the hips. The latter method is how cats instinctively right themselves when falling. I'm pretty sure, however, that if you already have a spin, you cannot eliminate it via an acrobatic maneuver and then come to a complete stop. It only lets you change your facing, not your angular momentum. So maneuvering thrusters are quite necessary still, if you pick up unwanted angular momentum somehow.
Satellites often carry reaction wheels or gyroscopes, which are nothing more than a flywheel stuck to a motor, and distributed across the craft in such a way as to give it full 3 axis directional control without the use of thrusters. There's a number of advantages to this approach. The math is easy. The mechanics are simple, one moving part per axis, in simple rotation. rotating all the way around in any direction any number of times is easy, and all the moving bits can be completely contained internally, allowing the outside of the craft to have whatever you want mounted wherever you want.