Icosatetraped does, in fact, mean “twenty-four legged.” I’m not sure how to inject “soft” into that word (icostatetrasquishaped?), but this robot does have 24 soft legs. Or rather, 8 legs are soft (and moving) at any one time, while the other 16 are pressurized to carry the weight of the bot. It can move at about 1 meter per minute, which isn’t especially fast, but who cares, look at all of those little legs go! Made from plastic medical tubing, particle board, a bunch of solenoids, a Mac Mini, and some 24 volt rotary vane compressors salvaged from Gulf War nerve gas detecting equipment, this is about as DIY as it gets, and it’s awesome.

VIA [ Boing Boing ]

Last time we heard about robot evolution, the bots were figuring out how to deceive each other. Now, researchers at EPFL in Switzerland have been using the same sort of genetic programming techniques to enable robots to teach themselves how to solve mazes, cooperate on tasks, and hunt each other (we’ll save that one for last).

The way genetic programming works is that the robots are only programmed on a very basic level, with simple information on their sensors and objectives. At first, the robots are clueless as to how to take the information from their sensors and apply it to completing their objectives, but after each test, random variations (mutations) are introduced into the code. Robots that demonstrate the most improvement have their code passed on to the next generation, and the process was repeated a bunch of times. In this experiment, after 100 generations, the robots taught themselves how to navigate a maze without running into a wall, and figured out that having their sensors pointed in the direction that they were going was the best way to be. It gets cooler:

“In another experiment they programmed groups of robots to push tokens along a wall to a marked area to win points. They selected the robots that gained the most points to pass their code on to the next generation. Over time altruistic behaviors were observed, in which robots sacrificed points if the entire group would benefit, and the robots cooperated to push larger tokens together to earn more points. As in nature, the robots followed the biological principle of kin selection, in which they only helped robots having the same code lineage.”

Code-based kin selection. Crazy, huh? It makes sense, though, if you think about it… Robots work together best if they’re using the same code, and robots who aren’t using that code won’t know how to cooperate with the robots that are. So, they’ll end up being less efficient, and won’t be as likely to make it into the next generation.

The predator-prey dynamic is perhaps the most interesting. One group of robots with several sensors were programmed to chase after another group of robots, who had fewer sensors but were faster. At first, the predator robots simply chased after the prey robots, who ran away. After 125 generations of evolution, the predator robots had figured out how to stalk the prey robots, hiding out and then sneaking up on the prey’s blind spots. The prey robots, on the other hand, learned where the predators liked to hide and made sure to keep their sensors facing them.

All this, in just a hundred or so generations in a lab. Kinda makes you wonder why we don’t just set a couple hundred Roombas loose in a dust covered wearhouse, let them fight it out and breed with each other, and after a couple years we’d get the world’s smartest, most efficient, and deadliest robot vacuum.

[ Evolution of Adaptive Behaviour in Robots by Means of Darwinian Selection ] VIA [ Physorg ]

The paper also refers to another robot evolution project called Golem@Home, which used distributed computing to design a moving robot from scratch. Different simulations were raced against each other, and the winners were actually created. We posted about it back in 2007, it’s pretty cool.

We first got wind of this back in October, but now it’s official, with a picture… DARPA’s LS3 (Legged Squad Support System) program has just awarded Boston Dynamics $32 million over 30 months to design and build a prototype for a new, larger walking quadruped platform that will probably not be called HugeDog. Here’s what DARPA specifies in the contract:

The program goal is to develop a walking quadruped platform that will augment squads by carrying traditional and new equipment autonomously. These platforms will be capable of managing complex terrain where tactical vehicles are not able to go, lightening the load of Marines and Soldiers and increasing their combat capability. LS3 will carry 400 pounds or more of payload, and provide 24 hours of self-sustained capability over as much as 20 miles of maneuver. LS3 will weigh no more than 1,250 pounds (including its base weight, fuel and payload of 400 pounds).

Key LS3 program themes are:

-Quadruped platform development: design of a deployable walking platform with sufficient payload capacity, range, endurance, and low noise signature for dismounted squad support, while keeping weight and volume scaled to the squad level.

-Walking control: develop control techniques that allow walking, trotting, and running/ bounding and capabilities to jump obstacles, cross ditches, recover from disturbances, and other discrete mobility features.

-User Interface (to include perception technologies): the ability for the vehicle to perceive and traverse its immediate terrain environment autonomously with simple methods of Marine/Soldier control.

The key differences from the existing hardware are increased range (20 miles from 12), increased payload (400 pounds from 340), and the ability to jump and trot. Oh, and the “low noise signature,” which I guess means that DARPA wants it sounding like something besides a giant swarm of bees. Or one giant bee. If everything stays on schedule, we should be seeing some awesome and hilarious videos sometime around 2012.

[ LS3 (PDF) ] VIA [ IEEE ]

We tend not to cover robotics toys that much around here, but if you know someone who’s interested in robotics, something fun and simple (and cheap) is a good way to get them introduced to the field. Trossen Robotics, which generally sells serious hobby robots, today opened a new online store for robot toys aimed at the younger and less experienced roboticist. There are over 40 items priced under 20 bucks (including walking robots and line followers), plus a “girl friendly” section, whatever that means. The first 50 orders today get a free Hexbug robot, which is a good reason to start your Valentine’s Day shopping… Trust me, this is pretty much the most romantic gift there is.

[ Robotics Toys ]

Hobby robots tend not to prioritize form over function, but Palro (pal-ro, robot pal, get it?) looks pretty slick with his 15″ tall molded body and LED head. He’s got 20 degrees of freedom, an integrated camera, and apparently can respond to voice control. A comprehensive library and open software package gives a lot of programming options, and an Atom processor inside ties it all together.

While Palro certainly looks cool, if you use him to do the kinds of stuff that hobby robots are good at (like backflips and somersaults and stuff), you’re bound to blow out a servo sooner or later. And from the look of things, Palro might not be the easiest (or cheapest) robot to fix. ‘Course, he’s not the cheapest robot to begin with, costing just under $3300. The first thousand units will be available March 15.

[ Press Release ] VIA [ Plastic Pals ] and [ Robonoble ]