backtop


Print 94 comment(s) - last by tatoruso.. on Aug 24 at 4:17 PM


Artist rendering of the X-47B in combat  (Source: Northrop Grumman)

  (Source: Northrop Grumman)

  (Source: Northrop Grumman)
Northtrop Grumman and the U.S. Navy will fly the X-47B in late 2009

The U.S. military is furthering its funding of unmanned vehicles for combat. Just last week, DailyTech reported on the U.S. Army's new SWORDS unmanned robots which roam the Iraqi battlefield carrying M249 machines guns -- and in turn put human soldiers out of harm's way. The military's latest unmanned project leaves the desert behind in order to take to the skies.

The U.S. Navy on Friday awarded Northrop Grumman a six-year, $635.8 million USD contract to further develop the X-47B fixed-wing unmanned air system (UAS). The funding for the Unmanned Combat Air System Carrier Demonstration (UCAS-D) program will allow Northrop Grumman to conduct take-offs and landings from the U.S. Navy's nuclear-powered aircraft carriers.

"We are proud of our legacy of innovation and creativity in developing new combat capabilities and are pleased to be selected to lead this revolutionary advancement in unmanned systems capabilities," said Northrop Grumman's Scott Seymour.

"The UCAS-D award is the culmination of several years of effort with the Navy to show the benefit of melding the capabilities of a survivable, persistent, long-range UCAS with those of the aircraft carrier," continued Northrop Grumman's Gary Ervin. "The UCAS-D program will reduce the risk of eventual integration of unmanned air systems into carrier environments."

Northrop Grumman will build two X47-B aircraft for the U.S. Navy -- the first of which will take flight during the closing months of 2009. The company expects to begin the first carrier landings in 2011.

The X-47B, a sister-ship to the X-47A, has a cruising altitude of 40,000+ feet and a combat radius of 1,500 nautical miles. The stealthy vehicle can carry an internal payload of 4,500 pounds and can travel at high subsonic speeds.



Comments     Threshold


This article is over a month old, voting and posting comments is disabled

RE: And then SkyNet
By mdogs444 on 8/6/2007 11:32:02 AM , Rating: 2
There is no true AI, and our people actually control the movements.

Maybe you've seen iRobot or something too many times.


RE: And then SkyNet
By FITCamaro on 8/6/2007 11:38:55 AM , Rating: 2
True AI is possible. It just requires such a large machine today that its impractical for anything other than research into AI.

We're still probably decades away from the technology and software of having a real artificial intelligence being able to fit into something the size of a fighter much less a humanoid robot. Even if we did solve the problem of the processing power and software, we'd still have the power source issue to solve. We're not going to have nuclear powered robots walking around.


RE: And then SkyNet
By bldckstark on 8/6/2007 1:25:41 PM , Rating: 2
How about a link to back up this claim?

I am aware of nothing above the insect level of AI that actually works as intended, let alone human.


RE: And then SkyNet
By FITCamaro on 8/6/2007 2:47:34 PM , Rating: 2
They've gotten AI up to the level of an infant. The issue though is programming efficiency, speed of the computers, and storage space.

If you think about it, what exactly is our ability to make decisions? It's what we've learned. We've learned as we've grown that certain things do and don't work. From those experiences we try new things and see if those work. Why should a computer be any different? The brain is nothing more than an organic computer capable of extremely fast processing and near infinite data storage. The only issue is how well do we remember everything.

Sorry I'm having trouble finding a link to the example I'm thinking of. But just because we haven't mastered it yet, don't think it isn't possible. All it takes is time, ingenuity, money, and creativity. I think in the next 50 years, as computer grow more powerful and ever smaller, we'll see the birth of a computerized brain. Maybe not as smart as us, but able to make decisions based on what its experienced.

I believe the experiment I'm referring to was called HAL though. I have a book on robots back home. But I'm not there. Otherwise I could look for it.


RE: And then SkyNet
By Scorpion on 8/6/2007 4:23:21 PM , Rating: 2
Boy you sure got it figured out. Too bad you haven't a clue what you're talking about. Do you know how to program such AI? Simple AI? Ever heard of a little thing called "Bayes Theorem", "Artificial Neural Networks", "Bayes Decision rule"? These are all very simple probabilistic tools which often used on only the simplest events, but first you have to build models for them. I haven't even scraped the surface on information theoretic research, but it is far more complex then you seem think it is. These models occur billions of times in parallel all the time in our brains. To model something like that would take such tremendous resources and time, and that's only after we understand the models better than we already do. See also "Turing Machine".

I see too many people here too swayed by science fiction rather than science fact. Do you have any proof to your claim of "infant like" AI capabilities? Do you understand the cognitive abilities of infants? Do you know that infants are not capable of understanding that when something disappears from view that it has not disappeared from existence? That you can repeatedly show them something and take it away and they will still cry every time it is hidden. And that's only a small example. That is so very very far from the cognitive abilities of a full grown adult.


RE: And then SkyNet
By Ringold on 8/6/2007 5:26:55 PM , Rating: 2
I don't know about any of this myself, but I can draw some conclusion from whats known:

1) Super-computers are working their way towards the raw processing power of the human brain

2) Super-computers are also working on increasingly vast amounts of RAM

3) Moores Law

4) The spurt of quantum computing news over the last year

All that together means to me that, sooner or later, it'll be possible. We don't need robots pondering how many angels can fit on the tip of pin; just "Is that a target?" and "Should I fire at it."

Now, if programmers can't keep up with the raw power, as you (and game developers) hint at, then that's one thing.


RE: And then SkyNet
By FITCamaro on 8/6/2007 6:06:59 PM , Rating: 2
Yes I realize it is far more complex than I am stating.

And you said it yourself. It requires building a model. Then it has to "occur" or be run on a computer. Eventually the models will be made and perfected and the computers will be fast enough to handle them in real time much like the brain.

It's all just a matter of time and people far more intelligent than I am to figure it all out.

You're focusing on the present. I'm looking to the future. I don't doubt the capabilities of mankind one bit. Mankind has shown over and over than where theres a will theres a way. Its just a matter of expending the effort to do it. And scientists and researchers are.

100 years ago if you told someone that you could talk to someone via a video conference in real time they'd put you in an asylum. Who's to say what'll happen in the next 100 years. Technology has exploded in the past 40 years. Give it another 40 and see where we are.


RE: And then SkyNet
By BiuTech on 8/6/2007 4:32:29 PM , Rating: 2
I agree that it's just a matter of time. The future is automation and robotics, and as quantum computing continues to advance, much more will be possible with true AI.

It's amazing to see the imaginations of people like Isaac Asimov progressively develop into reality.


RE: And then SkyNet
By Black69ta on 8/7/2007 11:57:00 AM , Rating: 2
Sorry as this is kind of off the topic and backs you up without providing any proof but, if you want proof that we have a real possibility to one day see a good AI then watch old Star Trek episodes. the hypospray exists today. there is a limited function "tri-corder", even the teleporter has been demonstrated over a short distance, I think I read that in either CPU or MaximumPC, it was at a major Univ. Even a optical computer is close to a reality. Only 12 years ago ('95) a Cray-4 cost $11 Million and only delivered 32GigiFLOPs now a Sony PS3 delivers ~6 TeraFLOPS for around $500. So, I don't think the raw processing power required for "real" AI is that far off.


"If you can find a PS3 anywhere in North America that's been on shelves for more than five minutes, I'll give you 1,200 bucks for it." -- SCEA President Jack Tretton

Related Articles













botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki