Motoman has many heads, but he saves this special one for Halloween. He’s out there, right now, creepily making things out of LEGOs and assembling chairs.

Be very, very afraid.

John wrote in to let us know that Cornell has posted a new video of their “jamming” universal robotic gripper, which I neglected to mention is derived from a collaboration between the Cornell Computational Synthesis Lab, the lab of Heinrich Jaeger at the University of Chicago, and iRobot. There’s nothing specific in this new video that we didn’t already know, but watching someone squeeze the gripper in its flexible and rigid states is a very effective way to demonstrate the principle. Also, at the very end, it manages to pick up a flat plastic disc with ease, which is one of the things that it wasn’t supposed to be able to do. Hmm, I guess maybe it does actually do everything.

A few other things worth pointing out: it’s a pretty big step to totally abandon the conventional model for a universal robotic gripper, which is of course the human hand, and instead develop something exotic which is also easy to build simple to operate, and (above all) cheap. They didn’t start out trying to build a gripper, though… This particular application came out of a project to try to create “programmable matter” for DARPA.

And in case you were wondering, they absolutely weren’t kidding about the gripper itself being simply a balloon filled with coffee grounds:

Pretty cool!

[ CCSL Jamming Gripper ]

Okay, so maybe this isn’t Humanoid Robots 101, but the rules are a little bit different when you work at Trossen Robotics and get paid in semi trucks full of exotic servos. At least, that’s what I imagine happens over there.

[ Trossen Robotics ]

We didn’t quite have the budget to make it to AUSA this year, but Lockheed Martin has uploaded a bunch of videos on their displays to YouTube. Highlights include updates on HULC, a look at what used to be called MULE, and peeks at some new SAMARAI hardware,

SAMARAI:

Check out more, after the break. (Read more…)

It’s been a little while since we’ve posted a micromouse vid, so here’s a good one. It’s from a competition in Chubu, and the winner completes its final run in 5 seconds. Or maybe 4. Skip to 4:30 if you’re interested in the speed but not the exploration phase.

I can only assume that micromice are approaching some kind of asymptotic maximum speed when it comes to maze solving, but I’m optimistic that people will still keep on inventing ways of making them faster and more efficient… Vacuum power was one, and I’m sure there are more. It’s an interesting problem, though, because there has to be a compromise between speed and stability, which may involve slowing down, as this high speed video shows:

That robot is Tetra, who we’re already somewhat familiar with, but in this competition it took second place to the winner in the first vid, a micromouse called EggTorte.

[ RobolaboN @ YouTube ]

Lockheed Martin’s original pack-hauling robot was named (appropriately enough) MULE, which stood Multifunction Utility/Logistics and Equipment… Vehicle. MULE got canceled at the beginning of this year, though, and its effective replacement seems to be the SMSS, or Squad Mission Support System.

SMSS is completely autonomous, and will either follow behind a soldier or follow preset GPS waypoints, all while avoiding obstacles with its conspicuous and fragile looking LIDAR system. It can 1200 pounds of gear, has a range of over 200 miles, and also functions as a mobile gear charger.

The long-term vision of this system can accommodate armed variants, while improving its reconnaissance, surveillance and target acquisition capabilities within the concept of supervised autonomy. A squad-size manned or unmanned support vehicle is critical to today’s asymmetrical and urban battlefields.

It sounds like Lockheed Martin’s long-term vision is basically to turn SMSS into what MULE was supposed to be: a six wheeled platform that, in addition to hauling equipment, could (by simply adding some modular accessories) also be used for this sort of thing:

Of course, MULE was undoubtedly going to be super expensive and much harder to develop, but it had the potential to offer tons of flexibility, which personally I think is the key to the future of the military in general. Oh, and that thingy it used to get past the barbed wire? It looks an awful lot like one of these.

Anyway, I guess it’s good to see that Lockheed Martin’s original concept isn’t totally dead, it’s just kinda sad when an awesome robot gets made so much less awesome, even if it is substantially more practical.

[ SMSS ]

Picking up things is pretty sucky for robots, since so many complex variables are involved. And the more complicated the object, the more complicated the task of picking it up seems to be. This new type of robotic gripper can pick up all sorts of objects (the more complicated, the better) by using a technique called ‘jamming.’ Jamming is simple and effective: a balloon is filled with coffee grounds. When air is blown into the balloon, it expands slightly, the coffee grounds can move around a little bit, and the gripper can conform to complex objects. When air is sucked out of the balloon, the coffee grounds bind together, and the object is now firmly held. It’s basically a practical method of shape shifting. Cool!

The funny thing about jamming is that effectiveness of the grip is, in most cases, directly proportional to the complexity of the object. This means that flat, smooth objects like sheets of paper are pretty much impossible to pick up using jamming, which is the inverse of using a different technique like suction. Porous or squishable objects like cotton balls also don’t lend themselves to being gripped in this manner. But this is a small price to pay for being able to pick up more or less any object without having to first examine that object and calculate exactly how to go about doing it… Plus, a balloon and coffee grounds? Sounds pretty damn cheap to me.

[ PNAS ] VIA [ Hizook ]

We were sort of wondering what WowWee’s new Roboscooper was really capable of, and this video of it in action sort of shows some stuff. Kind of. I’m still vaguely wondering if there’s any sort of localization system besides simple obstacle avoidance, although I’m guessing not at this point. And while golf might be cool, does it actually aim? I guess for $70 you can’t really be too picky, but WowWee set the bar pretty high with Rovio and it would be nice to see something from them that was both fun and had the potential for more.

Incidentally, that poor little kid could impale people with his hair… But, maybe that’s normal. In, um, Australia, or something.

[ WowWee Roboscooper ]

You might remember Daniel Wilson from (among other things) his Keepon snuggling escapades, but he’s also the artistic and technical brains behind such books as ‘Robopocalypse‘ and ‘How To Survive A Robot Uprising.’ It’s now looking like Steven Spielberg will be directing a movie adaptation of the former, which will start shooting in January of 2012 for a 2013 release. The movie is about the human race’s attempt to survive an apocalyptic robot uprising, and it strikes me that it would have been smarter (or at least, better for humanity in general), if ‘How To Survive A Robot Uprising’ had come out on film first… But that book has only just been picked up for a movie by Jack Black, of all people. This doesn’t mean that Black is going to act in the movie, he’s just picked up the rights to make it, so we’ll have to see what comes of that.

Of course, it’s hard for me to condone movies about robot uprisings and Robopocalypses and stuff… But, as long as people can take these movies as they’re intended (as fiction), I guess that’s okay. Meantime, I’ll be bracing myself for the inevitable slew of breathless “Robopocalypse: Could It Really Happen?” articles that’ll be popping up as soon as the first trailer for the movie hits.

Sigh.

Robopocalypse (the book) should be available in June of next year, and How To Survive A Robot Uprising is in stores now.

[ Robot Uprising ] VIA [ Deadline ] and [ io9 ]

TUM’s PR2 (named James) and another resident robot named Rosie have teamed up to make some pancakes. This is pretty awesome. But all the stuff that’s going on before the actual pancake making even starts is just as awesome. To begin with, James as no idea how to make pancakes. So, he does what we would do, and looks up the recipe on the internet. From instructions he finds there, he’s able to generate his own pancake-making program, which includes using image recognition to determine what the right bottle of pancake batter looks like when he goes searching for it in the fridge. Both robots are also able to dynamically adapt to errors and obstacles and show up in their environment, common in a real kitchen.

As we’ve seen before, pancake flipping isn’t easy, but Rosie manages it with finesse. And I have to admit: when Rosie flipped that pancake, I clapped in real life. And I may have giggled.

So, how long until I’ve got a couple robots in my kitchen serving me pancakes whenever I want? Considering that this particular pancake likely involved about a million dollars worth of robotic hardware, it might be a while… But the autonomy and adaptability demonstrated here are laying the foundation for a James/Rosie lovechild in every home and pancakes for one and all.

VIA [ Willow Garage ]