On: Birds make better bipedal bots than humans do

[GauntletWizard]

How will this scale to quadrupeds? It seems novel, but birds are lightweight and topheavy, with significant active balance redistributing their weight front-to-back. Look at how much head movement is involved in a bird walking.

For light loads, this may work very well, but most robotic walkers are being developed as pack animals and I think that's the clear niche for bipedal robots.

[raylad → GauntletWizard]

The head movement of birds is a mechanism for depth detection since they do not have binocular vision.

By moving their heads back and forth, they can determine distance based on the rate of size increase of an object.

See for example: https://www.loc.gov/everyday-mysteries/zoology/item/why-do-p...

[enasterosophes → GauntletWizard]

You seem to presume that this specific technology should scale to quadrupeds.

Other people are already working on quadrupeds. You may have heard of Boston Dynamics.

Not everything needs to do the same thing as everything else.

[tuatoru → GauntletWizard]

Have you heard of ostrich races and emu races? Birds can carry reasonably heavy loads. Quickly.

Edit: much more energy-efficient than a drone delivery service, and probably not that different in speed for medium distances.

The difficult problem, as with any mechanical replica of a biological mechanism (tensegrity energy recovery locomotion in this case), will be long term reliability.

[regularfry → tuatoru]

And safety for bystanders.

[makeitdouble → GauntletWizard]

Wouldn't quadrupeds be a separate problem space with pretty different solutions ?

In particular, most quadrupeds have a strategy to keep two legs touching as much as possible, which fundamentally changes the balance game.

[boomlinde → GauntletWizard]

How will a boat propeller scale to an airplane jet turbine?