Hasta La Gadget!

Continue reading Mapping robots equipped with visual vocab filters for more accurate mapping

Filed under: Robots

Mapping robots equipped with visual vocab filters for more accurate mapping originally appeared on Engadget on Thu, 26 Feb 2009 15:27:00 EST. Please see our terms for use of feeds.

Read | Permalink | Email this | Comments

If you were freaked out by the spooky Big Dog quadruped robot, you are going to love seeing it in action alongside US Army soldiers, following them like a smart giant dog.

And that’s what it is: A really smart giant dog. Big Dog—which is being developed by robotics company Boston Dynamics—has some of the most advanced artificial intelligence and navigation systems in the planet. In fact, US Army officials are stunned by its programmed behaviors, which make Big Dog extremely helpful in the battlefield.

Big Dog can run along soldiers, walk slowly, or lay down to be loaded or unloaded, all while being aware of the terrain around it. No matter what happens, or how hard it’s hit, the robot maintains its course like Captain America, but without falling on the ground under any circumstances. The only way to get him off its path is by a major direct hit, which could mean a rocket or President Obama’s defense cuts—although, since it is low cost and not a Cold War-era weapon, it probably will survive the crisis.

Even while they are not yet ready for actual battlefield action, the US Army has high hopes for this beast. They see it as the most effective way to carry all kinds of material, reducing the weight that soldiers have to tug along, freeing them to move faster and be safer.

Right now, they only need to make Big Dogs quieter—they still buzz like a billion angry bees—and increase their autonomy. If there are no unforeseen problems, they will get into the battlefields in just a few years.

With drones now taking over the skies, it’s only a matter of time before a Big Dog 2.0 arrives. In fact, many analysts are already pointing at the possibility of bi-pedal robots fighting alongside soldiers or replacing them entirely. Science fiction authors have imagined this scenario many times, from Terminator to Metal Gear Solid 4, in which “geckos”—the two-leg Velociraptor-like fighting robots shown in this video—are clearly inspired by Big Dog’s biological design.

For all the technological terror they inspire among anyone who sees them, however, they can’t match the hilarious spoof that some troupe did a while ago. Sadly, the idea of Big Dogs, Terminators, or “geckos” taking over the battlefield—or the world—is not funny at all.

NEDO, an administrative institute in Japan, has been working on what it calls the “Project for Strategic Development of Advanced Robotics Elemental Technologies” since 2006. The project has now entered its second phase, and boasts some pretty impressive looking bots. Murata Machinery’s robotic delivery system (pictured above) which is designed to help in places like hospitals, delivering medications late at night so that nurses and aids don’t have to spend a lot of time on such tasks. The company plans to test it and monitor the bot in use at hospitals in order to verify its effectiveness. We don’t know about you, but the idea of this guy visiting us late at night when we’re feeling low in a hospital is either really awesome or terrifically creepy. Either way, we fully expect this guy to have a starring role in The Phantom of the Opera any day now. Hit the read link to check out the other participants in the project.

Filed under: Robots

NEDO’s Advanced Robotics project enters second phase, boasts totally sweet bots originally appeared on Engadget on Mon, 23 Feb 2009 21:37:00 EST. Please see our terms for use of feeds.

Permalink | Email this | Comments

The tilt sensors or the accelerometers on the HTC Android G1 phone aren’t just for playing Super Monkey Ball. A team of do-it-yourself drone enthusiasts are channeling it to remotely control and access live video feeds from a robotic blimp.

"I am a techie and this is techie heaven," says Howard Gordon, founder of Surveyor, a company that makes robotic microcontrollers. "My interest was in creating a 3D vision development platform and this is a really good tool for that."

For hobbyists, the Android-controlled robotic blimp is a relatively inexpensive fun project. With a fairly easy guide, replicating this blimp costs under $600. "Everything we have done with the robotic blimp is open source,"
says Gordon. "We have the details right from where you can get the
blimp to how you can get the code on your G1 phone."

The first step involved getting a 66-inch helium blimp kit, which is available for for under $300.The blimp was modified to add an ultrasonic ranging module a Wi-Fi antenna, a compass and a camera module called the SRV-1 Blackfin camera that Surveyor makes.

Then there’s the Android app available for download through a Google code site that once installed can be used to control the blimp. To install it on the G1, Gordon and his team used the developer kit environment.

The accelerometers of the G1 control the motion of the blimp. "Those are our primary control signals and can also use the scrollball to tilt it." The app also has buttons to control the thrust vector of the blimp’s propellers adjusting it to say a 45 degrees angle.

The robotic blimp project took nearly a year. But the detailed
documentation of the project makes it much easier for others to
attempt it, says Gordon.

Check out the awesome video demonstrating the G1 controlled robotic blimp:

More pictures

Photos: Howard Gordon

A graduate student in the famed MIT Media Lab has created a robot that uses circular OLED displays as eyes, enabling new ways to research human-robot social behavior.

The new bot, called TOFU due to its stretchy and bouncy actions, was based on animation techniques from Disney’s 2D cartoons from the 1950s. If you remember the cartoons of the time, you’ll notice that the "squash and stretch," "secondary motions," and eye rolls shown by the robot are quite accurate. According to the robot’s creator, this type of motion is usually not used in other bots.

TOFU is made out of a combination of elastic materials and the vibrant colors and brightness available in the OLED displays will enable the creators to offer more lifelike expressions. It would be especially interesting if the MIT students add an angry look to such a seemingly nice character.

TOFU’s jumbly, bumbly personality also resembles that of the Keepon child-development dancing robot. Keepon gained wide notoriety two years ago when its creator, Marek Michalowski from Carnegie Mellon University, released a video of the bot dancing to Spoon’s I Turn My Camera On. The Keepon is used to study human-bot relations and behaviors including eye-contact, joint attention, touching, emotion, and imitation.

Check out the video after the jump.

—

Photos: ryan.wistort/flickr