PDA

View Full Version : Robotic Animals: Can we make them?



Andrew D
2010-Jul-10, 01:55 AM
AOL NEWS EXTERNAL LINK (http://www.aolnews.com/article/air-force-aims-to-launch-spy-pigeon-drone-by-2015/19546975)



Air Force Aims to Launch 'Spy Pigeon' Drone by 2015
Updated: 3 hours 26 minutes ago
Print Text Size
EmailMore
Sharon Weinberger

Sharon Weinberger Contributor
AOL News
(July 8) -- In the run-up to the invasion of Iraq, U.S. officials often had to rely on grainy satellite photos to decide whether facilities on the ground were intended for producing weapons of mass destruction. Now imagine that instead of overhead satellite imagery -- or even high-flying unmanned aircraft -- they could send in a flock of microdrones that could actually fly right over, or even inside, such facilities.

Even better, these drones -- equipped with chemical sensors that could pick up possible weapons work with near certainty -- would resemble typical birds, like pigeons, making them nearly impossible to spot.
A prototype of the Air Force Research Laboratory's bird-like micro air vehicle is shown. Researchers say the would-be spy pigeon will flap its wings like a real bird, and even be able to land on power lines.
Sharon Weinberger for AOL News
A prototype of the Air Force Research Laboratory's bird-like micro air vehicle is shown. Researchers say the so-called spy pigeon will flap its wings like a real bird, and even be able to land on power lines.

This high-tech spy vision is precisely what Air Force researchers are trying to build, and they believe such a microdrone is not only possible, but could be ready to fly in just five years.

"Ideally, it'll be a bird-sized UAV [unmanned aerial vehicle], with the current goal being a pigeon," Dr. Leslie Perkins, the lead for micro air vehicles at the Air Force Research Laboratory, told AOL News. The "birdlike" UAV would also be able to operate with minimal pilot intervention for up to a week at a time, she said.

The scenario that the Air Force envisions for its would-be spy pigeon is a cross between the high-tech military thrillers of Tom Clancy and the science-fiction novels of Isaac Asimov. It would fly with almost no human interaction and be equipped with advanced sensors capable of detecting nuclear, biological or chemical weapons.

To date, the prototype hasn't progressed much beyond the world of RadioShack, which sells a variety of animal-inspired toy planes, such as a flying bat. The Air Force Research Lab recently displayed a prototype of its bird, built by Ohio-based Theiss Aviation, at a trade show in Florida.

Though the bird is supposed to eventually have flapping wings, the current model has fixed wings with a push propeller in back; it has flown for about half an hour in a test facility.

Air Force researchers hope that in a few years, however, the bird will look and fly like a real pigeon -- and even perch inconspicuously on power lines to recharge. While there are a number of universities and companies working on micro air vehicles, the Air Force Research Lab has laid out a specific goal to field a bird-inspired drone by 2015, and then one based on an insect by 2030.

As part of its research, the Air Force Research Lab in May officially opened a $1.5 million testing facility, called the micro-aviary, dedicated specifically for micro air vehicles. Perkins, who helps coordinate researchers' work across the lab's various locations, says everyone knows that they are pushing the envelope on technology.

Among the challenges faced with micro air vehicles, for example, is finding a power source that is small yet powerful enough to provide the drones with endurance. Though a number of private companies build micro air vehicles, like AeroVironment's WASP drone, they can typically only fly for less than an hour.

The Air Force Research Lab, by comparison, wants its micro air vehicle to operate for a week at a stretch. "When you talk with private firms, they don't necessarily laugh, but they do realize that's a holy grail," Perkins said.

The pigeon drone builds on a growing interest in biomimetics, which draws on nature to inspire technology. In the case of robots and drones, this means studying everything from spiders to hummingbirds to understand how they move.

At Brown University, aerospace engineer Kenneth Breuer has been studying bats, which are able to fly flawlessly through complex obstacles and tight spaces, such as in a cave. Breuer is looking at ways to better understand the physics and dynamics of the way bats fly, including their ability to flip themselves upside down to land.

"They do pretty remarkable things and we're interested in understanding how they do it," Breuer told AOL News.

Though Breuer thinks bat (or bird) flight may hold lessons for engineering, that doesn't mean that animal- or insect-inspired flight would necessarily always make for the best sort of drone. "[Living things] are optimized for evolutionary survival, not for a particular engineering mission," he said.

Mark Lewis, an aerospace engineer at the University of Maryland and former Air Force chief scientist, agrees that biomimetics, though good science, may not always trump traditional engineering and says it's wrong to assume that just copying nature is always the right answer.

"Nature," Lewis said, "never evolved a spinning rotor."

Apperently government engineers don't have a continuing education requirement:

Youtube : ATP synthase (http://www.youtube.com/watch?v=uOoHKCMAUMc&feature=related)

gzhpcu
2010-Jul-10, 06:39 AM
The Japanese have come up with Paro, a therapeutic robot baby seal. It is used as a pet for patients with Alzheimer's, for example.
http://mentalhealthnews.org/robot-helps-comfort-dementia-and-alzheimers-sufferers/841432/

hahn1
2010-Jul-10, 09:32 AM
Can we make them?

I guess it depends on how you define animal. Are we waiting for a sufficiently sophisticated AI, or just an immensely complicated pre-recorded response program. If the latter we've (well Japan anyway) already made them.

I like the idea of a constantly flying robot. If they can make it efficiently aerodynamic, I image that it could be powered with solar radiation. I also imagine it would be possible to increase power efficiency by using multi-directional wind turbines to counteract cross winds (though I'm no engineer).

Japan's Paro mentioned above really reminds me of the Asimov short story about the sentient cars ("Sally"). Except, if it's the Airforce, these robots really will be capable of ... MURDER (music: den-don-den )

SkepticJ
2010-Jul-10, 10:23 PM
Well, if the engineering problems could be worked out, the act of perching on a power line could be the source of power for the drones.

If the perched drone could connect to both the + and - lines at the same time, it would have a source of power. But, dealing with that much voltage and amperage might not be possible; little robo-bird fries itself.


They'll make them eventually, but five years? yeah right.

Andrew D
2010-Jul-11, 05:27 PM
It seems to me the biggest problem is in materials and mechanics. I don't think a pigeon sized package would be able to be sufficiently articulated to achieve biomemetic flight, definately not with servos and pistons.

SkepticJ
2010-Jul-11, 06:34 PM
Well, I doubt they'd use pistons. They wouldn't have enough hertz. Artificial muscles are the jazz they use now.

Flying isn't the problem (http://www.wired.com/dangerroom/2009/07/video-pentagons-robo-hummingbird-flies-like-the-real-thing/) now, powering the thing for long periods is.

It needs something like thin, flexible photovoltaics that fold up for flight, but have a nice surface area when deployed.

It's a common misconception that man made materials are lacking compared to those of the natural world. That isn't necessarily the case anymore. Carbon fiber composites are stronger and lighter than any bone, artificial fibers exist that exceed the tensile strength of the strongest spider silk (heck, some steel alloys do that) metal alloys exist that barely corrode on the scale millennia, even if immersed in seawater. What the natural world does well is order poor materials into strong ones--think bird eggs, abalone shells, and bones.

Andrew D
2010-Jul-12, 12:00 AM
Artificial muscles are the jazz they use now.

After I wrote that last comment I looked for something along those lines, but only found research for implants and inflatable sacks some Japanese companies are using. Do you have any links?


Flying isn't the problem (http://www.wired.com/dangerroom/2009/07/video-pentagons-robo-hummingbird-flies-like-the-real-thing/) now, powering the thing for long periods is.

That's awesome, I had no Idea we were already doing this. What area are we talking about for the photovoltaics?

mugaliens
2010-Jul-12, 08:10 AM
I think the final hurdle is one of efficient power, and that's where biology excels. It's no secret the most energy-efficient per passenger mile vehicle on the planet is a bicycle (http://en.wikipedia.org/wiki/Bicycle_performance).

Murphy
2010-Jul-12, 11:45 AM
Isn't it rather self defeating to tell the world about this technology? Now any "Rouge State" trying to develop WMDs will just shoot any birds within a 10 km radius of their secret site. :lol:

EDG
2010-Jul-12, 06:30 PM
If we could make them, does that mean we could finally get this? :)
http://www.youtube.com/watch?v=-CVYOCMpJRY

HenrikOlsen
2010-Jul-12, 08:41 PM
Now any "Rouge State" trying to develop WMDs will just shoot any birds within a 10 km radius of their secret site. :lol:
Is that a state that uses too much makeup? Or did you mean "rogue"?

hahn1
2010-Jul-13, 05:56 AM
Is that a state that uses too much makeup? Or did you mean "rogue"?

North Korea is both. But they can't shoot anything accurately enough to make a difference.

Antice
2010-Jul-14, 02:06 PM
After I wrote that last comment I looked for something along those lines, but only found research for implants and inflatable sacks some Japanese companies are using. Do you have any links?


well... there are quite a lot of research going on out there with this stuff: Electroactive_polymers (http://en.wikipedia.org/wiki/Electroactive_polymers)
Google yields a plethora of results about that at least. it's good for making muscle fiber analogs.