Awesome robot (slightly on topic!)

I'd swear that those sony robots were CG.

AdrianT

I think it looks weird because the legs don't behave naturaly. They have set timing for the footsteps which maked it look almost like a comical enthusiastic dance. Like it wants to go pee or something, ANd it never changes rhythm so it seems unnatural. Particularly over uneven terrain like loose debris.

Kanati

#10

didn't sony drop their robot line entirely recently? Or was it just the aibo?

Skitchy

#11

Aaaaaaargh.
No, I mean AAAAAAAARGH!

There was something very 'wrong' about that - I think its the way it seems somewhere between alive and not alive. Why does it sound like a model plane when used outdoors, but not indoors?

I'm going to sell my 'anti-robotic-mule' device to the military. Its called "the rope". You tie it between 2 trees.

Neuro

#12

That mule thing reminded of two people ajoined at the head walking together....kinda funny how they kick it around and stuff....almost feel bad for it....

#13

What's wrong with it is that it looks like two guys carrying a heavy dresser, and neither one is sure of what the other plans to do, so when they get knocked off balance they try to rebalance, but they're doing their own thing rather than working together.

Their walking algorithm is probably making each leg try to keep it in balance by itself, unless it is providing forward motion, and none of the legs know what any of the others is doing.

Also it is like two midgets on stilts carrying a dresser, in that it only takes tiny steps and has to take lots of them to get anywhere.

Skitchy:
Indoors, it is powered by electricity. See the wire? Outside, it's powered by chainsaw motors or something, that are just generating the electrivity I guess.

#14

I was looking for another video of a small dinosaur robot I've seen, but I found this instead:
http://video.google.com/videoplay?docid=8795179308223101590&q=dinosaur+robot

And these robots which are somewhat more realistic:
http://video.google.com/videoplay?docid=-8038213278474581043&q=robot

big10p

#15

What's wrong with it is that it looks like two guys carrying a heavy dresser

Ah, yes, it's the Chuckle Brothers Robot MkII. "To me, to you, to me..." :P

didn't sony drop their robot line entirely recently? Or was it just the aibo?

They may have done - dunno. I just found the vids and assumed it was their new one, as I hadn't seen it before.

#16

I found the dinosaur robot I was looking for, but this is a much longer (and more boring) video of it than I had seen. The original page for it appears to be gone now.

http://www.ai.mit.edu/projects/leglab/mpeg_vcd/videos/troody.mpg

puki

#17

Heh, creepy robot.

I have my own robot at home - one of these:

My robot can talk - he can see things - he can hear things - he can sit down and get back up again - he can dance - he can pick things up - he can shake hands - he can throw things - he can recognise colours - he can guard my bedroom - plus other stuff.

Grey Alien

#18

I thought it look exactly like how a Donkey behaves when you kick, don't you ;-)

Kanati

#19

I was thinking it looked a lot like a young fawn (deer) as it was trying to walk. Very unsteady but it wasn't falling down either.

Dubious Drewski

#20

"i can't really figure out the point?"

The practicality of this robot is irrelevant. Right now it
can't do much. But neither can anything else when it is
still a brand new idea/technology. Wait 50 years and
you'll see robots that can jump, climb, swim, throw, etc.
Then you will see the purpose of these seemingly useless
experiments. I'm sure that in the end, legs will be more
practical than wheels in alot of situations. (Though not all!)

"and none of the legs know what any of the others is doing."

I'm almost certain that's not true!

"Why does it sound like a model plane when used outdoors, but not indoors?"

When outdoors, it generates electricity via gas motor.
Indoors, it's wired to the building.

Naughty Alien

#21

..I tell you..I dont like robots..

#22

>The practicality of this robot is irrelevant. Right now it
>can't do much. But neither can anything else when it is
>still a brand new idea/technology.

True.

>Wait 50 years and
>you'll see robots that can jump, climb, swim, throw, etc.
>Then you will see the purpose of these seemingly useless
>experiments.

But they already began appearing in the sixties... That's 40 years ago! I want my flying car NOW!

>I'm sure that in the end, legs will be more
>practical than wheels in alot of situations. (Though not
>all!)

Nature's had a 10 million year head start and I haven't seen any species being born with wheels.

As for the robot in question, it's really not a robot. It's just a big radio controlled toy. It has limited autonomy, limited movement etc.

Andy

#23

That's 40 years ago! I want my flying car NOW!

Supposedly, that's supposed to happen within this decade...
http://www.autoblog.com/2006/02/17/holy-mighty-morphin-suvs/

http://news.com.com/Flying+car+ready+for+takeoff/2100-11389_3-6040007.html

Neuro

#24

lol...now imagine terrorists getting their hand on several of those....

Dubious Drewski

#25

Alright. That particular flying car looks very dangerous. Look
at the shape of it and proportions of the wings! I bet you'd
have to travel at 120mph to take off or land.

Flying cars are not practical.

Driving down the highway in a car, if your engine fails, you pull over.

If your engine fails on one of these, you likely kill you and your family.

Ross C

#26

Very good point.

DampeS8N

#27

Nature's had a 10 million year head start and I haven't seen any species being born with wheels.

4 billion, and a wheel would be almost impossible to evolve because it requires free motion independant of it's location and therefore couldn't be connected with any strand-based organs like viens, nerves, and other connections... a wheel simply couldn't evolve unless that limitation were worked around.

Driving down the highway in a car, if your engine fails, you pull over.

If your engine fails on one of these, you likely kill you and your family.

People used to say that about air planes.

The thing is... your car has 1 engine, if it fails you car stops. The flying car has like 8, and can fly on 2-4 of them... it is very unlikely that 4 engines will crap out at the same time... you would obviously want to stop flying if even 1 died.

cars are far more deadly than planes, mainly because average people don't fly the plane, but these flying cars will have to damn near fly themselves to be legal. They fly along computer controled GPS flying highways and know where all the other flying cars near them are at all times and will avoid them. They will be able to sense when things aren't right and won't take off if the condition of the craft is too poor to fly.

wedoe

#28

Hmm, strange stuff.
I kinda feel sorry for the "creature" when they kick to tilt it...

markcw

#29

I have dial-up but i still looked at this.

Cool! Is my first remark. Wayyy cool! If this is what they're doing today just think what they'll have in say 10 years from now! They'll have proper walking robots which could probably go into battle instead of humans! Terminators! I really liked seeing this, thanks Tom.

#30

>4 billion, and a wheel would be almost impossible to
>evolve because it requires free motion independant of it's
>location and therefore couldn't be connected with any
>strand-based organs like viens, nerves, and other
>connections... a wheel simply couldn't evolve unless that
>limitation were worked around.

eyes were long considered impossible to evolve. The point I was trying to make, was that nature has had millions of years to work out the optimal solution for a given problem. animals have only just begun needing to move faster than 100km/h in last 100 years, and compared to the population as a whole, if you do move faster than 100km/h, you have a higher likelyhood of not fostering any off-spring, which is a fundamental part of evolution.

>The thing is... your car has 1 engine, if it fails you car
>stops. The flying car has like 8, and can fly on 2-4 of
>them...

The Moller project(Which isn't really a car at all and has evolved into a small aircraft with lifting surfaces everywhere) has 8 engines. Other projects use 0.25-8 engines.

>Cool! Is my first remark. Wayyy cool! If this is what
>they're doing today just think what they'll have in say 10
>years from now!

It's just a research study, research studies rarely advance to actual prototypes.

>They'll have proper walking robots which could probably go
>into battle instead of humans! Terminators! I really liked
>seeing this, thanks Tom.

Unlikely, 1, 2, 3 and 4 legged automatons have been built since the 60's, even being able to upright themselves if kicked.

The key is to fit balance, vision, intelligence and communication into a single package. With every new system that needs to plug in, the complexity and computing power needed increases exponentially. Humanity just isn't there yet.

You can read more about the project at BDI's website.
http://www.bdi.com/

EDIT: Some small changes to grammar and presentation.

Andy

xlsior

#31

The key is to fit balance, vision, intelligence and communication into a single package

An even bigger problem is a self-contained power source.

Not much sense in having a terminator robot that requires an extension cord.

#32

>Not much sense in having a terminator robot that requires
>an extension cord.

Oh please, where's your sense of adventure. :)

Andy

#33

Extension cord?

If I was an artifically intelligent computer, and I needed to power my robot army, and I didn't need to worry about the saftey or well being of those pesky humans, then I'd just go with an unshielded nuclear power source. Like interstellar spacecraft, you just stick a bit of plutonium in a little metal container and use the radiation from it to heat water and generate power via steam.

Voila, a battery with a half life of a thousand years. :-)

xlsior

#34

Of course, it's not like it grows on trees. :->

#35

>Voila, a battery with a half life of a thousand years. :-)

Well, it sounds simple, but in reality it's slightly more complicated. Radiation actually degrades stuff like metals, electronics and concret as well. So that invincible terminator with plutonium power pack may soon find its legs buckling and it's circuits frying.

Andy

DampeS8N

#36

andy, a wheel can't evolve because there is no simple progression to it that would be helpful to an animal.

eyes went from ight sensitive spots, to immobile specs, to compound eyes, all the way to the human eye.

a wheel is very unlikely to evolve because it requires so many things that don't exist in animals to ALL happen in the same mutation. Eyes might roll, but they can not turn 360 degrees and then continue to turn in the same direction 1000s of times. They are connected via veins and arteries and nerves, muscles and tendons.

a wheel would require nothing less than:

a way to feed blood and impulses to the wheel for healing that doesn't work on direct connections, something that has never happened in the animal kingdom.

a freely mobil socketed connection to a body, something else that has never evolved.

some kind of shaft to drive the wheel, which would require a hard connection to the wheel, which in the generations leading up to the wheel would be a detriment.

and possibly the worst hole is the fact that there is only 1 kind of motion producing organ in the animal kingdom, aside from micro propulsion like flaggellum. The muscle, which produces linear motion only (and can be in rings to constrict, still linear) and require a connection with anything they manipulate. Meaning in order to make the wheel turn, you would require either a new kind of motion organ, or a cam shaft, with more than 4 connections that has more of these brand new swivel joints the wheel uses, which the muscles connect to so they can produce the rquired circlular motion.

Wheels didn't evolve in the animal kingdom because they aren't better... they didn't evolve because they require 3 brand new systems to evolve together that have never existed before. Animals are made up of basic parts, and very little changes from generation to generation... a wheel would be a detriment until it had fully evolved, and detriments are weened out long before they breed... even if the wheel evolved to that point, it would still be bone and be very weak, it might be able to go 10 miles per hour and it would need a lot more evolution to become useful... like a more eficient food delivery system since plants couldn't support all the wasted energy from linear to circular conversion and friction. If the wheels could evolve into metal, then maybe... but we have yet to see an animal do that... so thats another evolutionary leap that would be required to make an animal wheel work....

#37

andy, a wheel can't evolve because there is no simple progression to it that would be helpful to an animal.

Actually, you're wrong. Just sitting here reading your post, I thought of several ways that a wheel (actually probably 2, 3, or 4 wheels) could evolve on an animal and actually prove useful. You just have to think outside of the box...
The only possible limitation that I can think of would be the lubrication of the wheel bearings.

And, I think that the biggest detriment to using wheels wouldn't be the evolution, but the kind of terrain it would have to cross, much like our own vehicular wheels. But they might have actually evolved if there was enough flat open (and fairly hard) terrain.

Dubious Drewski

#38

No... Dampe seems right to me. Wheels would require
alot of evolutionary effort for very little payoff compared
to legs. It's just not a good choice.

Plus, like he said, how would bodily nutrients travel to a
wheel to allow it to grow or heal?

#39

>andy, a wheel can't evolve because there is no simple
>progression to it that would be helpful to an animal.

That's a conclusion based on an assumption.

The likely reason we don't see animals sporting wheels is not that wheels could not evolve, but rather that there has been no benefit so far in developing them.

>eyes went from ight sensitive spots, to immobile specs, to
>compound eyes, all the way to the human eye.

Exactly, but until recently science could not even clasify the 4 different types of eye developments that can be found in nature. There was heated debate on the subject until it was realized that these different types were in fact 4 different evolutionary tracks.

>a wheel is very unlikely to evolve because it requires so
>many things that don't exist in animals to ALL happen in
>the same mutation.

No, wheels could well be an off-spring of legs, and the development wouldn't need to happen in one go. Legs are wheels, except they don't rotate and feet does not extend into a circle.

>Eyes might roll, but they can not turn 360 degrees and
>then continue to turn in the same direction 1000s of
>times. They are connected via veins and arteries and
>nerves, muscles and tendons.

Yes, but wheels could be made from bone. The animal could also be genetically programmed to shed the horn after a while making it a detached wheel. Soe salamanders can shed their tails if they need a fast getaway.

A leg is one track for ground locomotion. Think about it, if an animal simply had the ability to curl it's legs into circles and unfurl them rapidly, this would be the equivalent to wheels, in that it allowed the animal to reach extreme speed for acceleration. If the 'wheel' was solid, it could be moved by grabbing it and moving it.

>a wheel would require nothing less than:
>a way to feed blood and impulses to the wheel for healing
>that doesn't work on direct connections, something that
>has never happened in the animal kingdom.
>a freely mobil socketed connection to a body, something
>else that has never evolved.
>some kind of shaft to drive the wheel, which would require
>a hard connection to the wheel, which in the generations
>leading up to the wheel would be a detriment.

That's how humans would do it, but does nature have to do it our way in order to create wheels?

>Meaning in order to make the wheel turn, you would require
>either a new kind of motion organ, or a cam shaft, with
>more than 4 connections that has more of these brand new
>swivel joints the wheel uses, which the muscles connect to
>so they can produce the rquired circlular motion.

You're looking at it from a mechanics standpoint. It's like asking Leonardi Davinchi to built a digital computer.Not knowing about electricity he would propably opt for animal power, not knowing about boolian logic, he would propably use rope, pulleys and gears. Eventually, given a good explanation of the subject, I'm sure he would get there, but it wouldn't be a PC.

>No... Andy seems right to me.

You mean 'No... DampeS8N seems right to me.'...

The whole point of evolution is that it develops application specific beings, and under different conditions, evolution might well have come out completely different.

Andy