Korea To Build Front-line Combat Robot

christchurch wrote to mention the story of a plucky Korean robot that has been built for combat. From the article: "According to design blueprints released during a meeting of science-related ministers, the robot will have six or eight extendable legs with wheels allowing it to move like an insect over uneven terrain. The robot will be armed with various weapons and will operate both by remote control and its own artificial intelligence system"

Yes

[Solr_Flare]

Yes, the USA, if I recall properly, has variations of their spy drone planes that are equipped with missle launchers. Likewise, there was an article a few months back about the US getting ready to deploy a heavily armed remote control tank-bot for "testing" in Iraq.

But, to date we have not yet equipped, to public knowledge, a robot with weaponry that is not purely remote controlled. Armed AI robots make people nervous, and for a variety of good reasons given our state of "AI".

Of course, we aren't talking a Skynet situation here(although some day that will likely be technically possible). Its more like not wanting a blue screen of death to literally kill you.

Fragging nerds

[aussie_a]

You put a gun in an overweight nerds hand and shove him onto the battlefield and he's going to get killed very quickly. You put him behind the remote controls for one of these babies, and you'll have a lean-mean killing machine. Will nerds be the nest people to be drafted by the government? After all, all of those years training in Quake and Doom should make them experts wielding these babies.

I can see it now, Korea is at war with someone else using these on the battlefield. Kim and his friends want a LAN party, so they PAY the military to for an hours worth of time renting out 5 of these. They get behind their computers, and are suddenly transported to a battlefield and they go for it. Just make sure it's programmed so that the thing can't shoot allies (perhaps the allies emit a beacon) and the kids can go for their life, trying to frag as many people as they can. It'll be all the rage!

What about the Asimov rules?

[Bananatree3]

Is it just me, or is the world forgetting Isaac Asimov's laws of robotics?:

Isaac Asimov's "Three Laws of Robotics" asimov

1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.

2. A robot must obey orders given it by human beings except where such orders would conflict with the First Law.

3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

It seems to me that this is a casebook example of such discrase for these laws. We have forgotten what these rules, layed down by father of robotics over half a decade ago. It is sad to see how we have used something like the robot to simply continue the cycle of ever-more expensive and bloody cycle of militery technology, and now with AI to go with it

Depends on your definition of robot

[Solr_Flare]

Most people call various expensive remote controlled devices "robots". If I make remote controlled spider vehicle is that any more a robot than a remote controlled plane? Does it have to walk on the ground to be a robot?

Or, is a robot defined by it's AI? If so, how much control does the AI need to have to make it a robot? How sophisticated does it need to be? Depending on how loosely you define AI, you could call some modern cars robots.

Then, after you define "robot" the next question is does the article writer using the same definition as you when they say robot?

It's not that I'm arguing with you so much as I'm just saying the term robot is very "fuzzy" these days, especially in the military arena.

Re:Constantly hearing about combat-bots

[globalar]

Well, we can certainly mount guns on things and shoot remotely. The current, long-developing trend in military warfare is towards smaller, versatile units that are hard to pin down but are extremely coordinated. Hive-like would be a wet-dream, but I'm sure beaurocrat is having such a dream as we speak. Bottom line, robots don't yet play into this dream. They are really only good as disposable long-range swiss-army knives. This will improve over time, of course. And as you read, the border patrol part is a focus as well.

This particular focus from SK seems to be a politically-pushed idea. Perhaps similar to the U.S. Star Wars idea in the 1980's (i.e. makes people happy, makes some contractors rich). North Korea weighs heavily on SK citizen's minds, partly because they hear so much crazy, horrible stuff about what happens in the country. Remember there is a huge infrantry deployment (for U.S. particularly) along the North Korean border, so any press about being able to one day replace these forces with robots is good press (who will hopefully just be sentrys really, because no one south of the meridian wants war).

Also, the South Koreans are fairly keen on technology in general, similar to the Japanese. Like all of us, they love to find excuses to play with robots and grant government contracts. The SK government has invested heavily in certain industries in the past and now the market seems to be sensing the technological shift. Yesterday (way back) the hot things were conventional transport like shipbuilding. Today its robots. So the story goes, if you can get a copy of the student roster at MIT, you can find out just what the South Koreans want to become leaders in. Though, I would add a few Japanese and Chinese schools to the list now as well.

Re:Constantly hearing about combat-bots

[LnxAddct]

The U.S. already uses several robots in battle with them, mostly drones for mapping things. Some drones have missiles. There is also a smaller land robot that infantry can carry with them to inspect suspicious areas. There are easily about 500 other forms of robots on the drawing boards among various defense contractors. The idea is that robots can be made in any number and are dispensable, so eventually a soldier should be replaced by a robot. The robots we know exist are most likley nothing compared to what is behind closed doors. We often don't hear about the really good stuff until its about 50 to 75 years passed. With something like a robot that can fight as a soldier can, the U.S. probably would keep that under tight secrecy until the public needed to know(i.e. because we needed to use them for a large scale war). Something like that is not something you want your enemies to know about. As it is right now, if we ever had to battle a country like China where the government literally controls how many think and act and at the drop of a dime could send billions of people to war, we would need to scale up our forces, robots are the answer. One popular robot idea in the early research phase is obviously autonomous vehicles, as is shown by DARPA's annual grand challenge.
Regards,
Steve

Re:The AI must be really good.

[h4rm0ny]

Pretty simple. You declare a zone as "militarily active", or some other such pentagon-speak. In go the troops and the robots, and anything moving in there that doesn't carry a friendly radio tag gets shot at. If a "civilian" goes outside or rises off the floor or refuses to have radio-emitting "security restraining devices" hurriedly slapped on him by the soldiers, then he is "resisting" and will become a "active target".

Sorry for all the " ". That tends to happen when a politician is trying not to say they've just commisioned a machine to kill people.

Re:Constantly hearing about combat-bots

[ramblin billy]

"Hive-like would be a wet-dream"

Depending on your definition of robot, we may already have a "hive mind" controlled robot in operational use. The AEGIS Combat System [janes.com] is a combat control system that integrates a wide variety of sensors and weapon platforms into a single, computer controlled system. The central control computer receives data from linked sources such as radar arrays, satellites and aircraft located anywhere in the battle space. Using various algorithms, the system can track over 100 separate targets, assign priorities, choose the appropriate weapon from any of the assets under its command and target and fire the weapon remotely. An entire fleet of ships can thus act as one in presenting an integrated defense. Once the system is enabled the computer makes the individual decisions without human interaction. The humans just sit back and watch it go. That's pretty close to a hive mind. AEGIS has been around for about thirty years.

There are also examples of "fire and forget" weapons that may be classified as robots, although not in the traditional sense. Sea mines exist that can hover, scan the area with passive sonar and deploy only when certain conditions are met. These conditions can include the detection of the acoustic signature of a single ship or group of ships. Missiles and torpedoes can lock on to a target and independently proceed to an intercept, dealing with evasive behavior and counter-measures. Some munitions, upon losing target lock, can throttle down and cruise in search patterns in the attempt to reacquire the original target or aquire a new target of opportunity. Once any of these weapons is fired, they proceed without further human input.

The reason any of these weapons can be safely (well, safely in a battlefield sense) used is that they are confined to limited areas that are designated "free-fire zones" or are capable of differentiating between targets. That's fairly easy for ships and planes, not so easy for troops on the ground. It's easy to see how robot controlled guns could be used to repulse waves of attackers advancing through no-man's land. It's less clear how they would tell the difference between probing enemy scouts and a lost squad of their own men. Initial deployment might be on the basis of "go over HERE and let me know when you get there or detect anything on the way". Options might include "laser illuminate THIS target until follow-on munitions arrive then return to base" or "let me know if THIS target proceeds in THIS direction". It should also be remembered that acceptable levels of safety are quite different in times of war and on the battlefield. "Let loose the bots" may become the last ditch effort of desperate men.

billy - AEGIS is currently being migrated to Open Software...could it be...Linux?

I'm just waiting for the first virus/worm to hit

[xiando]

Building weapons like this is a bad idea. How can they ensure that the enemy, who ever they are, does not get access to the robots control system and thus turn them? By software, I assume. Software can have bugs. Bugs, on any kind of armed robot, can have disastrous effects. What if an enemy attempt at taking over the robots makes it lock it's control system from external access so even their owners can't control it anymore? What if the attack software bugs and starts viewing any biological life-form as hostile? Is it just me who gets way too many questions like that when I read about such robots? Am I the only one who gets a huge fear that the schedule for the inevitable Terminator-movies Judgment Day was just moved forward by a century? What if this robot, or the next generator, or the far more advanced version that will be produced in 10 years get infected with a computer virus, worm or just malfunctions because of a bug in the software?