That should give us some food for thought on that subject: of course it is going to take a very long time before we can create a robot that can remotely compare to our fine motor control capabilities, in a complete "body" package.
We can get relatively simple machines to reproduce fine motor control, but only for a small section of our repertoire. Arm and shoulder motions in one package, leg and hip motions in another, and a few in-between. The challenge is incorporating a functioning "core", with multiple sections that can work both autonomously as well as synchronized. It doesn't have to look pretty, but as we have seen thus far, systems that incorporate such ideals are very slow, and nowhere near as stable when facing anything other than the most simple of tasks.
Doing it in a bipedal package is even more insanely difficult, but we are looking at ourselves as "nature's best", when in reality, we are the only creature that is 100% bipedal, with a body plagued with mechanical problems thanks to a spinal column that is designed for a mixed approach.. and yet we have limbs of the wrong length to mix up our movements between bipedal and quadrupedal.
In fact, I say we start demanding six-limb androids. I mean, quadrupedal is inherently more stable, faster, and there are many examples of quadrupeds showcasing greater strength in both rear- and forward- limbs. Take a cue from nature.
All quadrupeds suffer from comparatively less-fine motor control capability (some primates offer a challenge, but they aren't 100% quadruped or biped - and that's likely a greater challenge on the android front). So, throw on a third pair of limbs, and make these the ones that feature hands and arms comparable to our upper limbs.
Be an unsightly beast, most assuredly, but come on - it would be awesome.
On the bipedal versus quadrupedal approach, look to that one "mule" robot in steady development for the military: it can trot at a decent speed, with a heavy load (or no load), and climb and descend at fair rates of elevation change. Oh, and during all of that, it can remain upright even when kicked, or regain balance after losing its footing (rocks, holes, ice, mud, snow, etc etc). It's actually
creepy to watch, often looking quite life-like.
I honestly think it will be quite a long while before we will see that amount of stability in a bipedal machine, and one that is life-size.. and can burden a great deal of weight while juggling the unpredictable environment.
It will take a great deal of processing equipment to allow such a robot to actually "see" and prepare for stability challenges at the mental level, prior to actually encountering the obstacle... without that, I think the bipedal android ideal is not going to be in our near-future.
There's a lot going on to keep us balanced, calculations and muscle activations at very rapid rates, that most can barely imagine - and it'll take a similar approach to get a robot that can remotely compare to our physical potential.
/Ramble On, you
Ramblin' Man
Facebook
Twitter