Hmmm,,,Robots?

[Gant]

I have an idea for a long lasting gyro-scope,,,,but not sure if it would be practical for those who are building,,or may have built,,,Robots. It seems that it is very hard to get robots to move as we do because they are unable to balance themselves as we do. The closest robots have come to be like us was made by Honda. But it still does not move as fluent as people do.

Here's the questions? Is it because of the gyro-scopes? Or is it more than one problem? And why?

[jordan]

Pretty much, there is the whole mimicking the movement of human muscles thing. I haven't seen one that can make a smooth motion like humans. They tend to move very unsteadily and consequently have to move rather slow to ensure they don't fall.

[Cap'n Refsmmat]

I saw in a Popular Science an article about how scientists designed a robotic insect that, surprisingly, needed no gyros. The way it walked gave it stability. Somehow it just loped and bounced back and forth and got right over objects without altering it's gait.

[Tesseract]

I saw in a Popular Science an article about how scientists designed a robotic insect that, surprisingly, needed no gyros. The way it walked gave it stability. Somehow it just loped and bounced back and forth and got right over objects without altering it's gait.

 

I suppose that could be used for small tasks, but would it work in a larger robot?

[Cap'n Refsmmat]

Yes. You can take it down from the original 6 legs to 2, and very large. It works.

[bdunshee]

As a big fan of giant biped robots.....

It's not really about the Gyros although they are important for detecting changes in balace as are accelerometers. Bipedal walking isn't easy. Your whole body reacts to keep you stable. That's pretty hard to build into a robot. Also think how much control motion a human foot has. If a robot is to walk naturally one would need a good bit of articulation in the joint proceeding the foot. People are working on it.

I'll edit in some links when I get to my computer to remember them.

(This is a work system)

 

As a side note I recomend the Japanese show Patlabor. It's a reasonably realistic depiction of what giant human controled robots could be like in the future.

With the rise of giant robots so comes the rise of giant robot crime. (Drunk operation, Bank Robbing, Terrorism ect....) Giant robot crime leads to special police forces with giant robots. (It's a comedy but has a amazingly possible depiction of the future.)

[Cap'n Refsmmat]

Why giant? Why not just any size, so they can get in tight places and such?

[Tesseract]

Why giant? Why not just any size, so they can get in tight places and such?

 

Maybe theyre made of a liquid metal like in "The Terminator 2"

[Ubermensch]

The robot-builders should really get with the biologists if they want to build robots that emulate life. It doesn't strike me as a very difficult conclusion. The conclusion being "I should ask the people that study living things if I want to design a machine that copy-cats living things."

 

For instance, when it comes to your gyro-less walking robot, the robot-builders would confront a question that has been studied in anthropology and other sciences: "how to humans stay upright when they walk?"

 

The answer has to do with the lower brain and the inner ear. The lower brain has been keeping primates erect for the past four million years. It contains the "logic" that allows the muscles to respond in such a way as to keep us erect. "Respond to what" is the next question, which is answered by "sensory data from the inner ear."

 

Now, humans have a metacognitive ability your machines cannot emulate. That is: even if a human has an inner-ear problem, he can compensate, and can still walk upright. But that is a bit much to ask of a robot.

 

The robot-builders must construct #1) a sensory device like the inner ear, #2) a set of logic circuits like the lower brain, and #3) a dizzying array of muscle-equivalents that can react to balance the robot.

 

As they haven't yet done these things, it either means A) they have not asked these questions (ie, they're not too clever) or B) it's hard to manufacture those three things. The latter is most likely.

[Gant]

LMFAO,,,ok,,you guys are way tooo funny,,lol,,but i thank you for your input. I will have to look into the robotic insect and learn more about it. Thank you cap'n for the information.

[Gant]

Oh and ubermensch has a good point. I do agree with you. If your an engineer, your expertise is in egineering, so to build with and manipulate materials to do what your design requires is what your education in that field has taught you to do. But to apply your trade in an area of expertise outside your scope makes it that much harder to achieve the goal you set out to achieve. So you do have to seek help from other fields in order to abtain said goal.

If it seems like i'm rambling on jibberish,,please forgive me as i am exhuasted. Long day.

[Ubermensch]

NP. It's been a trend for a very long time, in education. People over-specialize. I don't just mean "engineers don't talk about engineering with biologists," as in your robot situation. But it's got to the point where certain kinds of engineers don't talk to other kinds of engineers; certain kinds of doctors don't talk to other kinds; etc. It's the balkanization of academia, and is a major problem when it comes to cross-disciplinary stuff. After all, life on earth is essentially "cross disciplinary."

 

I think it's a shame that they --the brightest, most intelligent, the scientists and scholars-- can't look at a sunrise and enjoy it as an astronomer, a physicist, a chemist, a geologist, an artist, a historian, a poet and a philosopher. I can. It comes highly recommended.

 

But not only the appreciation of beauty --but errors are caused simply because one particular boffin has been so sequestered from both reality and other scientists.

[Sayonara]

NP. It's been a trend for a very long time, in education. People over-specialize. I don't just mean "engineers don't talk about engineering with biologists," as in your robot situation. But it's got to the point where certain kinds of engineers don't talk to other kinds of engineers; certain kinds of doctors don't talk to other kinds; etc. It's the balkanization of academia, and is a major problem when it comes to cross-disciplinary stuff. After all, life on earth is essentially "cross disciplinary."

It has nothing to do with balkanisation - that implies that the root cause is inter-discipline hostility. I'm not saying it never happens, but it is not ubiquitous and certainly not the cause of increasing specialisation.

 

It is purely because as time progresses we know about more things, and we know more about those things. Every person simply can't learn every piece of information.

 

 

I think it's a shame that they --the brightest, most intelligent, the scientists and scholars-- can't look at a sunrise and enjoy it as an astronomer, a physicist, a chemist, a geologist, an artist, a historian, a poet and a philosopher.

Where do you get this stuff?

 

 

But not only the appreciation of beauty --but errors are caused simply because one particular boffin has been so sequestered from both reality and other scientists.

Some scientists in "made a couple of mistakes just like other people do" shocker.

[Cap'n Refsmmat]

I'm glad all the spam in this thread was deleted.

 

But it's got to the point where certain kinds of engineers don't talk to other kinds of engineers; certain kinds of doctors don't talk to other kinds; etc

Oh really? They HAVE to. It's necessary to cooperate between scientists to get things to work.

[ed84c]

the knowlage of one's 'self' and 'being', are embedded deep into the human concieness; try touching your nose with your eyes closed, that is something a robot cannot do unless programmed with the correct parameters.

[Skye]

The semi-circular canals are pretty similar to a gyroscope in function, and we can replicate that reasonably well, I imagine. A better gyroscope would of course be useful, and not only for robots. Reliable gyroscopes are very important for everything that buzzes around in the air. But anyway, we probably need more than good gyroscopes to replicate our motion.

 

We also have a very complicated sensory system that tells us roughly where all the parts of our body are, and of course mechanical pressure on them, though. This is why you can touch your nose with your eyes closed, and if you happen miss, you can then feel it when you poke your eye. This is also important for balance, as mechanosensors in the joints allow us to sense the pressure on them, and determine very accurately how the loads on our body are distributed. Using this information the brain can then signal the right amount of force on the joint to do whatever we want, say walking, or throwing a ball.

 

We also relate our motion to visual cues in the environment. And we integrate all the information very quickly in response to changes.

[Gant]

Cap'n,,,isn't that what ubermensch said?

[Cap'n Refsmmat]

Yes.

[Ubermensch]

the knowlage of one's 'self' and 'being', are embedded deep into the human concieness...

This is demonstratably incorrect. The philosophy of tabla rasa states that people are born as a blank slate, and they get their knowledge from living on planet Earth. This can be proven. And by proof, I do not mean this silliness:

...try touching your nose with your eyes closed...

This is not proof. A person can touch their nose with their eyes closed because #1) they know that they have a nose, #2) they have rudimentary hand-eye coordination, and #3) they know that reality exists whether their eyes are open or closed. The point at which an infant learns each of these three ideas is different for every infant but nevertheless is the subject of childhood psychololgy. That is: this "tabla rasa" business can be proven in a lab.

[LucidDreamer]

The robot-builders should really get with the biologists if they want to build robots that emulate life. It doesn't strike me as a very difficult conclusion. The conclusion being "I should ask the people that study living things if I want to design a machine that copy-cats living things."

 

It’s called biomimicry and its really cool. I'm not sure a biologist would be the best candidate for a robot team. I think a good team might consist of a bioengineer, a biophysicist, an electrical engineer, a mechanical engineer, a robot engineer (if there is such a thing), and me. I don't really have any qualifications, but I really want to build a robot.