Automaton iconAutomaton

How Squishy Would You Like Your Robot?

Most robots are rigid. Rigid is easy to design, easy to construct, easy to calibrate, and more reliable for all of those dull, dirty, and dangerous tasks that robots excel at. When robots make fundamental structural compromises to rigidity, they do it in complicated ways, like with series elastic actuators or hydraulics. It's worth it, though, because adding squishiness can make robots both more capable and safer to be around through passive compliance.

Taking this concept to the extreme has resulted in some incredibly squishy robots, including soft robots that can walk, and other soft robots that can roll. But in both of these cases, embracing squishy properties means giving up rigidity. MIT has been working on a structure for a robot that offers both: squishy when you want it, and rigidity when you don't.

Read More

NASA Testing New Robots in Microgravity Before Sending Them to Space

At ICRA in 2012, researchers from JPL presented a paper on a new type of robotic gripper that uses microspines to adhere to rough surfaces in microgravity. In a follow-up paper at IROS 2013, the gripper made an appearance on a JPL robot, which was really cool to see. And now that it's 2014, it's time for the next step: seeing how a gripper designed for microgravity actually works in microgravity. To do that, NASA is sending it up on some parabolic test flights, along with several other robotic systems destined for space.

Read More

'Gobble Hawk' Wins NASA High-Altitude UAV Design Competition

NASA is very proud of the Global Hawk UAVs that it uses for environmental monitoring missions like keeping track of hurricanes, among other things. We should know: we visited them last year. But Global Hawks are super expensive (between $130 and $220 million each, depending on whether or not you factor R&D cost into the mix), and while they have relatively long range and endurance, they can only stay up for about a day (28 hours) at a stretch.

So NASA wants more options, and it has turned to students for ideas. In a press release today, they've announced the winners of a competition to design high endurance uncrewed aerial systems for hurricane tracking. Coming out on top: Virginia Tech's "Gobble Hawk." Heh.

Read More

SupraPed Robots Will Use Trekking Poles to Hike Across Rough Terrain

Last year at the Stanford-Berkeley Robotics Symposium, we saw some tantalizing slides from Oussama Khatib about a humanoid robot that used trekking poles to balance itself. We were promised more details later, and the Stanford researchers delivered at the IEEE International Conference on Robotics and Automation (ICRA) this year, where they presented the concept of SupraPed robots.

The idea is equipping robots with a pair of special trekking poles packed with sensors that, according to the researchers, "transforms biped humanoids into tripeds or quadrupeds or more generally, SupraPeds." By using these smart poles to steady themselves, the robots would be able to navigate through "cluttered and unstructured environments such as disaster sites."

Read More

'MacGyver' Robots Use Their Environment to Solve Problems

MacGyver could make a rocket-powered harpoon gun/zip line out of some moth balls, cleaning fluid, rope, pulley, and an old telescope. Robots can't do that. But hey, neither can you. What you can do is use objects in your environment to help you complete tasks that you wouldn't otherwise be able to do on your own, and robots are learning how to do the same thing.

Yes, this means that soon they will be unstoppable. And there's video to prove it.

Read More

Can a 'Dropship Quadcopter' Deploy Rovers on Mars?

The most ludicrous way of getting a robot to the surface of Mars is maybe stuffing it inside a giant inflatable bouncy ball and dropping it from a parachute. And that is only slightly more ludicrous than attaching it to a rocket-powered hovercrane (a rocket-powered hovercrane!!!) and then lowering it to the ground with some sort of ridiculous cable contraption.

NASA has done both of these things. Successfully.

So, the bar is very high for finding ludicrous ways of getting robots to the surface of Mars, and the European Space Agency (ESA) has taken on the challenge with a quadcopter that can safely drop a rover down onto the Martian surface while hovering.

Read More

Tiny Walking Robots Powered by Muscle Cells

Most robots are powered by electrical motors that are big, bulky, heavy, and if they break, you have to replace them. Animals, on the other hand, use a biological motor—a muscle—that also requires electricity, but is far more efficient and, given a chance, can repair itself. We're just starting to be able to manipulate biological structures like these in clever enough ways to let us harness their awesomeness, and engineers at the University of Illinois at Urbana-Champaign have worked them into a tiny little "bio-bot" that uses muscle cells to walk.

Read More

North Carolina Set to Outlaw Hunting and Fishing Drones

Restrictions on small drones, model aircraft, sUAS (small unmanned aircraft systems), or whatever you want to call them are being promulgated almost too quickly to chart these days. Late last month the U.S. Federal Aviation Administration forbade modelers from engaging in first-person view piloting, and the National Park Service announced its intention of banning essentially all model aircraft from its 84 million acres—to name just a couple of prominent federal clamp-downs. In addition, 36 states have been busy formulating their own drone regulations this year, four of which have recently enacted laws, mostly focused on protecting people’s privacy. But get ready for a novel variation on this theme working its way through the North Carolina State Assembly.

Read More

Please Tell This Robot What a Turtle Looks Like

Pretend you're a robot. (I do this all the time, it's great!). Okay, are you pretending? Awesome! Now, take these colored blocks and BUILD ME A TURTLE, ROBOT!

If you're panicking right now, that's understandable. A general purpose robot would probably have no idea what a turtle was, much less how to build one out of blocks. There are ways that you could teach the robot about turtles and blocks, but you, being a human, are hopelessly flawed and would only be able to teach it your conception of what a turtle should look like and how to use blocks to make one. What the robot really needs is to be able to examine a bunch of different examples of a bunch of different turtles, and then use machine learning to choose the best, most reliable, and most efficient one to build. And rather than have you try and do that all on your own, researchers at the University of Washington are paying strangers to do it as part of a crowdsourced effort.

Read More


IEEE Spectrum's award-winning robotics blog, featuring news, articles, and videos on robots, humanoids, automation, artificial intelligence, and more.
Contact us:

Erico Guizzo
New York, N.Y.
Senior Writer
Evan Ackerman
Berkeley, Calif.
Jason Falconer
Angelica Lim
Tokyo, Japan

Newsletter Sign Up

Sign up for the Automaton newsletter and get biweekly updates about robotics, automation, and AI, all delivered directly to your inbox.

Load More