Discussion Forums > Technology

Robots

<< < (5/7) > >>

mgz:
we actually understand quite a bit about the brain the only problem is interpreting and actually determining what is what.

Its as though for some of it we are trying to read a book in a language that nobody knows. We know its a book we can see there is words but for the most part its simply not feasible for us to interpret it.

Now back to what i was talking about how why its such a waste to focus on humanoid robots.
The primary reason i believe this is a waste is because its so inefficient. The reason it works for humans is because of our lack of a program to control us its the freedom that allows our bodies to work functionally.
The structural design is not one which is properly can maintain a large amount of weight IE thats why the morbidly obese have all kinds of issues with their bodies physically(spine and joints and such). Because the humanoid body/design is a terrible design when you deviate from its fairly tight restrictions on size and weight functionality.

There is next to zero benefit especially at this point in time to design a robot with legs vs a wheel or track system or combo of the 2. It is more efficient and more functional.
Arms as a basic function depending on their targeted design point can be VERY useful. But legs are kinda one of the important parts of a humanoid design and where a majority of the waste comes into play since there is alot more then that goes into creating whole systems to allow the thing to balance itself. Creating the systems to allow it to move freely in a human like way at anything above slow as fuck puts extreme stress on many design points and suspensions that would need to be made to allow it to do so.

The only plausible reason atm to design something more humanoid like is for aesthetic reasons of allowing them to assimilate with humans easier, but we are so far from this happening that its a nonexistent issue.

I would much rather see leaps and bounds made in the AI field then them making another asimo like robot. You know see that race that they have between the colleges with all AI controlled vehicles go at a pace faster then slow as fuck.

Because we can already make things then can climb nearly vertical cliffs better then any human, go faster etc etc.

kostya:
Big Dog is a cool artificial walking... thing. It can withstand jerks kicking it, rough and shifting footing, deep snow, ice, and can climb some pretty steep angles. This is where having legs is good:

http://www.youtube.com/watch?v=cHJJQ0zNNOM
http://www.youtube.com/watch?v=b2bExqhhWRI

blubart:
i kind of fail to see how having legs (like the big dog) is good in those examples.
basically over the whole time i watched those two videos i thought: come on guys, just put some inteligently designed tracks on that guy and it will be over in a second.
in really deep snow and heavy rumble legs would have a theoretical advantage. but looking at how the big dog works there is nothing of that advantage left and the disadvantages like crappy speed and the possibility to fall over in the first place outweigh by far.

carniska:
I think they have made awesome job developing Big Dog with quick reflexes regarding surface and robot's balance. As far as I know, it's equipped with 'targeting' sensors that it can follow squad leader in the war zone. And the original purpose is to carry wounded soldiers (imagine 4 legged stretcher) and anything that you can come up (explosives/food/equipment). Not mentioning it can be used in rescue missions ~ accidents/earthquakes etc. And it can go through narrow spaces.

But yeah, It needs 'further' development (or something): the walking speed is low, it's engine/motor is loud and annoying and the balancing system is dangerous for the 'patient'. And I think Crusher is more useful (w/o guns :P ):
http://www.youtube.com/watch?v=8qDo6ehxKds  // autonomous
http://www.youtube.com/watch?v=o2Kh7FVgDCU // remote controlled

Thought they have made in The Leg Laboratory interesting "balancing" solutions in aspect of "legged" robotics (old vid, but you get the point):
http://www.youtube.com/watch?v=M4Q9dRUeNU8

And Japanese people are funny :) Here's a joke regarding BigDog:
http://www.youtube.com/watch?v=X5BPrlOVX2o

darkjedi:
mgz,

I would think much of Japanese and Korean robotics focus is already placed on non-humanoid robots rather than humanoid robots since non-humanoid robots are the cornerstones of their high-tech industry. Without non-humanoid robots their industry can't survive.

Those academics who work on humanoid robots already understand that non-humanoid robots are better industrial platforms than humanoid robots. They already know that. They are not focusing an unnecessary amount their attention on humanoid robots. A greater pool of human resources is still dedicated to non-humanoid robotics research and AI than to the development of humanoid robots. These humanoid robots shown here were the outputs of a wisely scaled-down endeavor.

Navigation

[0] Message Index

[#] Next page

[*] Previous page