DARPA's Newest Robot, Cheetah, can run up to 18 miles(29 kilometers)
57 replies, posted
artificial muscles would be awesome for robotic prostheses
[editline]5th March 2012[/editline]
and more animal-like robots.
[QUOTE=GunFox;35011084]Okay from a technical standpoint I am aware that this is neat.
But from a personal standpoint, really? THAT is the best we have for land based legged animals? I recognize the extreme difficulties with designing a legged animal, but is this awkward fucker running on nubby feet the best we can manage?
[/quote]
I think this is a really good example of a 4 legged robot that already exists.
[video=youtube;cNZPRsrwumQ]http://www.youtube.com/watch?v=cNZPRsrwumQ[/video]
Coolest bit is at 0:35 when the robot is shoved from the side and reacts like you'd expect a real animal to. It also slips and slides around on ice which is pretty nice too.
[quote=GunFox;35011084]
Is it, perhaps, because we haven't yet developed artificial muscles? It would seem to me, and I am certainly not versed in the subject, that this may be a tremendous waste of time if we can manage really solid artificial muscles in the next decade. Not that I feel comfortable in the slightest questioning the wisdom of DARPA.
[editline]5th March 2012[/editline]
Oh, that is much better, though it also seems to be using hydraulics to mimic muscles more closely. The use of feet that actually articulate is nice.[/QUOTE]
I don't see a massive problem with using hydraulics rather than artificial muscles, they are quite similar when you think about it and really hydraulics are just a way of modelling muscles. Hydraulic systems also have the advantage of being well established, so they have been improved and polished over many years - always a good thing. A major problem with all of these systems is power and getting enough of it on a mobile platform, which is going to be an issue with or without artificial muscles.
I'd guess this cheetah robot is specifically meant to prototype and develop tech to make robots 4 legged robots move quickly, while they have other robots designed to develop other tech areas. Hopefully they will bring these different developments together in multi-purpose robots.
Looks more like a rabbit than a cheetah.
All it needs now are better joints, synthetic muscles, and a brain. Also, a huge gun on its back.
[QUOTE=GunFox;35011084]Is it, perhaps, because we haven't yet developed artificial muscles? It would seem to me, and I am certainly not versed in the subject, that this may be a tremendous waste of time if we can manage really solid artificial muscles in the next decade. Not that I feel comfortable in the slightest questioning the wisdom of DARPA.[/QUOTE]
Artificial muscle sounds like a cool idea, and would be useful, but right now, it's not something we can assume we will eventually have nor necessarily need.
In terms of efficiency, organic muscle is fucking terrible. 25% is about the best you could hope for, compared to conventional mechanical stuff, which can easily pull 90%. Muscle's not particularly impressive regarding volume or mass to output ratios either- take a bodybuilder doing a backlift and you'll beat the amount of his body engaged in the lift with a basic vehicle jack, on the small scale, I'm not sure about anything the weight of the masseter, but if you go for two I can find you pistons that'll ruin it. There's also the whole "uses energy at rest" thing. Shit's pretty awful. Granted, the idea of growing actual muscle or an organic analog and using it in machinery isn't the only "artificial muscle" option, but the next one, electroactive polymers, aren't much better, since they're still rather weak, have lag, temperature sensitivity and durability issues, and plain old manufacturing challenges. Kinda a dealbreaker when your option for emulating the an organic limb can't actually keep pace with its equivalent, reacts too slowly, and is fairly fragile. There's also the nanoscale work like that being done by John Madden (heh) and crew, but it's, you know, nanoscale, not really of any use in a large robot, and not particularly useful at all yet.
The whole concept is neat, but that's really all it is yet, a concept. There's still a lot of basic nuts-and-bolts research that needs doing for at least another decade before they become useful for much. Meanwhile, using regular mechanical gear, we can [I]already[/I] build something that approximates or outperforms organic limbs and bodies well enough. The issue is getting them to do what you want- controlling them, finding out if they're properly suited to the task at hand, junking or tweaking them, and streamlining the whole process. Which is where this thing came from.
[QUOTE=DARPA]The Maximum Mobility and Manipulation (M3) program is striving to create and demonstrate significant scientific and engineering advances in robotics that will:
-Create a significantly improved scientific framework for the rapid design and fabrication of robot systems and greatly enhance robot mobility and manipulation in natural environments.
-Significantly improve robot capabilities through fundamentally new approaches to the engineering of better design tools, fabrication methods, and control algorithms. The M3 program covers scientific advancement across four tracks: design tools, fabrication methodologies, control methods, and technology demonstration prototypes.[/QUOTE]
[QUOTE=BadderSanta;35011029][img]http://www.transformers2.net/wp-content/uploads/2009/04/ravage1.jpg[/img][/QUOTE]
Ravage says to fuck off.
[img]http://media.comicvine.com/uploads/1/11002/227149-195872-ravage_super.jpg[/img]
[QUOTE=UberMunchkin;35008215]Jesus Christ next thing we know Darpa are gonna make a T-800.[/QUOTE]
[QUOTE=erik802;35011020]Example of their work for the people who dont know.
[video=youtube;mclbVTIYG8E]http://www.youtube.com/watch?v=mclbVTIYG8E[/video][/QUOTE]
Looks like they're already on it.
That is pretty damn awesome.
[QUOTE=Xenocidebot;35012877]Artificial muscle sounds like a cool idea, and would be useful, but right now, it's not something we can assume we will eventually have nor necessarily need.
In terms of efficiency, organic muscle is fucking terrible. 25% is about the best you could hope for, compared to conventional mechanical stuff, which can easily pull 90%. Muscle's not particularly impressive regarding volume or mass to output ratios either- take a bodybuilder doing a backlift and you'll beat the amount of his body engaged in the lift with a basic vehicle jack, on the small scale, I'm not sure about anything the weight of the masseter, but if you go for two I can find you pistons that'll ruin it. There's also the whole "uses energy at rest" thing. Shit's pretty awful. Granted, the idea of growing actual muscle or an organic analog and using it in machinery isn't the only "artificial muscle" option, but the next one, electroactive polymers, aren't much better, since they're still rather weak, have lag, temperature sensitivity and durability issues, and plain old manufacturing challenges. Kinda a dealbreaker when your option for emulating the an organic limb can't actually keep pace with its equivalent, reacts too slowly, and is fairly fragile. There's also the nanoscale work like that being done by John Madden (heh) and crew, but it's, you know, nanoscale, not really of any use in a large robot, and not particularly useful at all yet.[/quote]
On paper, I agree.
In practice we have this:
[url]http://en.wikipedia.org/wiki/BigDog[/url]
It runs on a 15 hp engine and can carry a fair amount, but a single horse could do exactly the same thing using considerably less energy. A human being with moderate strain is going to be using anywhere from 75 to 100 watts. I have lightbulbs that use less energy and aren't also powering a biological super computer much less self repairing.
Something isn't adding up. The power these are putting out isn't anywhere NEAR human efficiency. Not even ball parked. If it was, then outfitting them with a battery pack would be easy and this tech would be on the field already. We'd have power armor everywhere.
This tech is a long ways away from being useful. Even when it does hit the field I'm not sure a hydraulics based system will ever quite match a system made using synthetic muscle. Or frankly real muscle for that matter. Hydraulics are powerful, but ultimately a suit of them is heavy. So while they may put out huge amounts of power, it doesn't do you a lot of good if they collapse floors or, for larger applications, crack concrete due to their weight.
It is similar to soldiers in remote regions of Afghanistan and Iraq being unable to haul even armored humvees into the region. Weight matters often far more than output.
[QUOTE=erik802;35011020]Example of their work for the people who dont know.
[video=youtube;mclbVTIYG8E]http://www.youtube.com/watch?v=mclbVTIYG8E[/video][/QUOTE]
[media]http://www.youtube.com/watch?v=roBWj9YPKPo[/media]
[QUOTE=GunFox;35013445]On paper, I agree.
In practice we have this:
[url]http://en.wikipedia.org/wiki/BigDog[/url]
It runs on a 15 hp engine and can carry a fair amount, but a single horse could do exactly the same thing using considerably less energy.[/QUOTE]
Except walk on similarly steep inclines, uneven terrain, low lighting (horses need to be able to see or they stumble), etc.
Also note BigDog's engine comes from the fact that it's built using existing generic parts because it was a proof of concept not a purpose-designed unit. You don't have an optimized machine on your first attempt.
[QUOTE=GunFox;35013445]A human being with moderate strain is going to be using anywhere from 75 to 100 watts. I have lightbulbs that use less energy and aren't also powering a biological super computer much less self repairing.[/QUOTE]
I'm not sure what point you're trying to make with this. That's A, wrong (define moderate stain, ~100 watts is what you burn keeping yourself [I]alive[/I]), and B, irrelevant. A lightbulb's only principle of operation is being electrically inefficient.
[QUOTE=GunFox;35013445]Something isn't adding up. The power these are putting out isn't anywhere NEAR human efficiency. Not even ball parked. If it was, then outfitting them with a battery pack would be easy and this tech would be on the field already. We'd have power armor everywhere.[/QUOTE]
You need to define a "these" and a thing they do that we are comparing efficiencies across.
And again, you seem to be making this mistake in assuming that prototypes using existing parts to hammer out control/design/etc. issues should be at perfect efficiency. If that was true no one would ever successfully design anything. You start with parts that are available, as you understand what you want, you start having purpose-designed variants.
[QUOTE=GunFox;35013445]This tech is a long ways away from being useful. Even when it does hit the field I'm not sure a hydraulics based system will ever quite match a system made using synthetic muscle. Or frankly real muscle for that matter. Hydraulics are powerful, but ultimately [B]a suit of them is heavy[/B]. So while they may put out huge amounts of power, it doesn't do you a lot of good if they collapse floors or, for larger applications, crack concrete due to their weight.[/QUOTE]
Uhhhhhhhh
Powered armor is a neat scifi idea and a goddamn awful one IRL, so I'm really confused. Are we talking about actual technology here or bullshitting about walkers?
[QUOTE=Doctor Zedacon;35008300]Actually it reminded me of Crying Wolf for some reason.[/QUOTE]
Needs railgun and crazy babe.
Actually, I wish we could see something like that in the near future. A "tank" in a similar shape as crying wolf's, with the same speed and strength. That would be really cool. Probably not really useful or cost efficient and easy to handle, but still cool.
[QUOTE=Jimpy;35008281]So next question, how does it turn?[/QUOTE]
It doesn't, everything else moves out of its way in fear :v:
Robot cheetahs. There's machines that can crush steel.
So this is how it ends.
[QUOTE=Overv;35008537]Host long can it actually run at that speed? Because I thought the title meant it could run up to 18 miles.[/QUOTE]
I hear they're still investigating the use of human flesh as a fuel source.
[QUOTE=Hivemind;35011451]I think this is a really good example of a 4 legged robot that already exists.
[video=youtube;cNZPRsrwumQ]http://www.youtube.com/watch?v=cNZPRsrwumQ[/video]
Coolest bit is at 0:35 when the robot is shoved from the side and reacts like you'd expect a real animal to. It also slips and slides around on ice which is pretty nice too.
I don't see a massive problem with using hydraulics rather than artificial muscles, they are quite similar when you think about it and really hydraulics are just a way of modelling muscles. Hydraulic systems also have the advantage of being well established, so they have been improved and polished over many years - always a good thing. A major problem with all of these systems is power and getting enough of it on a mobile platform, which is going to be an issue with or without artificial muscles.
I'd guess this cheetah robot is specifically meant to prototype and develop tech to make robots 4 legged robots move quickly, while they have other robots designed to develop other tech areas. Hopefully they will bring these different developments together in multi-purpose robots.[/QUOTE]
Seen this before; pretty cool. Also, let's see this Cheetah thing navigate rocky terrain and stuff like bigdog... pretty legit
All this time I've been playing the MGS series and being bothered by how ridiculous the robotic elements of the Metal Gears look. It just didn't seem like US agencies would actually make something like that.
And then IRL DARPA goes and does stuff like this. :v:
I may make myself out to be a pansy over this, but this thing seriously creeps me out....
[QUOTE=Xenocidebot;35013912]Except walk on similarly steep inclines, uneven terrain, low lighting (horses need to be able to see or they stumble), etc.
Also note BigDog's engine comes from the fact that it's built using existing generic parts because it was a proof of concept not a purpose-designed unit. You don't have an optimized machine on your first attempt.
I'm not sure what point you're trying to make with this. That's A, wrong (define moderate stain, ~100 watts is what you burn keeping yourself [I]alive[/I]), and B, irrelevant. A lightbulb's only principle of operation is being electrically inefficient.
You need to define a "these" and a thing they do that we are comparing efficiencies across.
And again, you seem to be making this mistake in assuming that prototypes using existing parts to hammer out control/design/etc. issues should be at perfect efficiency. If that was true no one would ever successfully design anything. You start with parts that are available, as you understand what you want, you start having purpose-designed variants.
Uhhhhhhhh
Powered armor is a neat scifi idea and a goddamn awful one IRL, so I'm really confused. Are we talking about actual technology here or bullshitting about walkers?[/QUOTE]
How is powered armor a bad idea?
Shame good things like this always end up used for warfare...
Every damn time.
Remeber the exo-skeleton, which was made to help partially paralyzed and old people?
[QUOTE=Hivemind;35010417]Average human running speed is 15mph, this thing can run at 18mph.
We are all doomed.[/QUOTE]
are you sure about that? 15mph is insanely fast. even then we're fucked though.
[editline]5th March 2012[/editline]
nvm, in short distance 15 mph is reasonable. i don't think anyone would last very long in a race against that thing though lol
[QUOTE=erik802;35011020]Example of their work for the people who dont know.
[video=youtube;mclbVTIYG8E]http://www.youtube.com/watch?v=mclbVTIYG8E[/video][/QUOTE]
God damn, check that strut. That is one confident motherfucker.
[QUOTE=Jetblack357;35017671]How is powered armor a bad idea?[/QUOTE]
For starters, it's really heavy. It requires specialist training just to ensure you don't break your arm doing basically anything. It costs a lot, plus the training, making it a waste of money unless you're invading a country literally made of oil. You don't really [I]need[/I] to carry anything particularly heavy when 90% of combat is just suppresive fire, and even if you did it's nothing that two or three extra soldiers helping can't do. And, even if it makes you a walking tank (which it wouldn't unless it cost more than an actual tank), you're still getting shot at - and bullets can have the force to knock someone to the ground, as well as break bones. How will medics treat people wrapped up in steel and wires?
It's just not necessary yet. When we come across clean, cheap, affordable fuel, have weapons and ammunition that weigh several tons, and basically have an amount of money that just isn't likely to happen with the current economy, then maybe we'll see.
[QUOTE=Jetblack357;35017671]How is powered armor a bad idea?[/QUOTE]
Maintaining it would be a bitch for starters, especially in the field. If electronics/whatever the fuck is powering it dies or is destroyed then you are rendered useless but atleast you have a fancy metal casket.
Does anyone else think ti looks like it's running backwards?
