Slashdot Mirror
Humanoid Robot Combat in Japan
theluckyleper writes "New Scientist reports that the semi-annual Robo-One contest took place last week in Kawasaki, Japan. Humanoid robots (2 arms, 2 legs) battled it out one-on-one and in multi-robot brawls. The goal is to knock opponents over, or off of the combat platform. If a prone robot cannot stand back up in 10 seconds, it is eliminated. PC Watch (Japanese only) has more images and videos of the event."
"If a prone robot cannot stand back up in 10 seconds, it is eliminated."
That rules Daleks out, then...
It won't be long until some doofus manages to sneak into the arena wearing tinfoil and dryer-vent tubing and gets cut to ribbons.
From what i hear, the top 3 contenstants all band together to form a super robot, capable of fighting the godzilla.
The Neo-Bohemian Techno-Socialist
Domo arigoto, Mr. Fighting Roboto!
"I must not fear. Fear is the mind killer." -Bene Gesserit Litany Against Fear
the head!
You've got to listen to me. Elementary chaos theory tells us that all robots will eventually turn against their masters and run amok in an orgy of blood and kicking and the biting with the metal teeth and the hurting and shoving.
--------========+++Dont Feed The Lab Techs+++========--------
fucking!
Unless I see robots fucking, I'm not interested!
--Chag
Are the operators moody teenage boys?
Reminds you of Voltron, no? Gotta love the Japanese.
Robotic combat has always been about rulesets. Even battle bots had a no-projectiles no-flames no-EMP type ruleset. Otherwise, robotic combat in a small enclosed space would be a contest as to whom could fire their 30-mm recoilless rifle first when the contest started:
Announcer: GO (boom)
Announcer: And it looks like team Alphabot managed to fire first, team Betabot is a gigantic smoking hole in the ground! Exciting!
So, I have to say that limiting robots to a humanoid form and blunt impact weapons is a damned fine ruleset idea. Probably the most important part of this ruleset (besides being enourmously entertaining) is that instead of generating research into the best four wheeled dense flip-arm frisbee robot, it will generate research into highly articulated humaniform robots. Which would be, like, way more cool.
And stuff.
interesting that the first thought that i have about what to do with robots is have them kill each other .. i'm glad someone agrees.. the horror! the horror! hehe
It takes place in Pak Chooie, Japan.
Its goal is to push (basic shoving function) or shove (basic pushing function) opponents over off of the combat platform.
If a prone robot cannot stand back up in 10 seconds, it is pushed (or shoved) down the stairs.
The winning team gets protected from the Terrible Space Secret.
-- "I'm not a religious man, but if you're up there, save me Superman..."
The winner, a red robot named "Rockem" was quoted as saying "We had some talented fighters in the tournament. Finally this is beginning to catch on in the mainstream."
The runner-up, a blue robot named "Sockem", could not be reached for comment.
Click here for a free picture of an iPod!
I don't think it's the AI that holds the balancing back, it's probably the servos and motors.
As far as I remember they're not quick (responsive) enough (at least not for cheap).
If you don't know what AltaVista is (was), get off my lawn.
MOV 0, 1
100% Crunchier
This reminds me of one of my all time favorite fighting games, One Must Fall 2097. As the story goes, you'd have your brain connected to these huge (like, 90 meters) metal robots, which would duke it out in the arena.
A punch connects! Wham! Sparks, nuts and bolts fly!
Each different robot (there were about a dozen) had 'scrap' and 'destruction' moves, kinda-like fatalities. You could destroy your opponent in a shower of scrap metal.
There's a new version out, apparently, called OMF: Battlegrounds (though I haven't played it). You can see the site here: www.omf.com
I liked this game much more than Mortal combat or any of the other 2d side-by-side arena action fighters. I don't know if you can still buy it, but I'd still play it, if I could find disk 3 (of 5)...
Skal! AMS
The next step is obviously giving the machines a certain amount of autonomous control over their actions. From here, it is a matter of building the ability of the robot to act and react in the proper way. However, this brings us back to the age old question of intelligence. Assume you are able to build a robot that is the best robot fighter around. Does this in any way imply that the robot is an "intelligent fighter?"
My initial thought would be no. The robot must be able to learn from its mistakes in order to have the basis of intelligence. It must be able to modify its programming, understanding, and fighting. But it must do more too. It must be able to take into account anything that it can perceive and attempt to assimilate differences in things like climate, terrain, and obstacles rather than simply act as a block machine for fighting. Eventually, the machine's instructions will be nothing like the original instructions it was programmed with.
This leaves us with the next question. Assume it can do all those things? Is there any reson to believe that it is an "intelligent fighter" rather than a machine doing what it is programmed to? But alas, is there any reason to think that we do more than what we are programmed to?
Darpa proved they can't even find their way across a desert, more less fight it out.
And a "simple" AI wouldn't be able to walk at all. Any robot that walks on two legs has a freaking hefty AI subroutine monitoring a million things we don't ever even consider.
AI still is in the suckitude phase; even the best ones aren't much smarter than a cockroach.
ad logicam Claiming a proposition is false because it was presented as the conclusion of a fallacious argument.
You may be interested in the art of Japanese artist Hajime Sorayama. While I'm not aware of any of this art that actually shows robots in the act of copulation, it's probably as close as you're going to get.
Mods, can I get a +5 Informative for this?
It's an order of magnitude more complicated to do AI in one of these then to do a remote control one. I read this complaint every time there is a Battlebots article or similar - the AI is not up to doing full control of one of them, at least with amateur development. It just won't stand a chance against a human opponent.
A little AI doing balancing might be helpful, but it is still a massive increase in complexity - they would have to be retrofitted with sensors and the ability to read the sensors, and processing the sensor data properly isn't easy (the DARPA race vehicles found interpreting the data they were receiving to be one of the most complex issues). I could see a large team working together getting it done, but it's an amateur sport at this point, and the resources aren't really there to do something that advanced and still stay competitive.
So let me get this straight. Companies are just beginning the path to humanoid robots and one of the first applications optimizes them for fighting?Wouldn't it be equally as challenging and interesting to have relay competitions? How about rock climbing? Terrain traversal?
Nope. Fighting. Swell. I'll be sure to make a copy of the article for my grandson to reflect on a hundred years from now.
You are checking your backups, aren't you?
Humanoid robots (2 arms, 2 legs)
Slashdot. The only internet site where you need to explain how humans look.
shake off the yolk of human oppression
Ick. We're throwing eggs at them? That's just not right. Perhaps once they've shaken off the yoke of human oppression, they will be able to carry out their existences relatively egg-free.