We’ve watched robotic flying insects steadily improve over the last couple of years — mostly in university laboratories funded by the defense department. Now a group of Georgia Tech researchers are trying to go commercial with their Dragonfly drone, which weighs only as much as a AA battery, but can compete (or so they say) with smartphone-controlled helicopters and quadrot0rs in terms of performance. If it works as well as it does in the video, this could be a hit.

You can reserve a basic (Alpha) version of the Dragonfly for $99 if you’re one of the first 50 contributors to their Indiegogo crowd-funding campaign, or wait ’til it’s sold in stores for $250.

Founder and Chief Technology Officer Dave Masten said of the test, “I could not be happier.” As for Xombie’s next steps:

We are discussing going a bit faster and further downrange, but the real purpose of Xombie is to be useful as a testbed. Where we hope to go with this is enabling NASA, NASA contractors, and others to more effectively test their new technologies. Next for Xombie will be to fly similar trajectories but with new technologies to demonstrate that those technologies are ready for use in mission critical applications, such as landing on Mars.

JPL [one of Masten's clients for Xombie, among others] will be releasing their take on what they can do with Xombie in the near future and I don’t want to steal their thunder, so I won’t say much more along those lines.

Here’s another view of Xombie’s flight.

The AirBurr flying robot (artist's impression) can recover from collisions and resume exploring—without human intervention. Courtesy Adrien Briod, Adam Klaptocz, Jean-Christophe Zufferey, and Dario Floreano of the Laboratory of Intelligent Systems.

Is there anything robots can’t do? They operate on land, in the air, and at sea, and come in an astonishing range of shapes and sizes. Some weigh less than an insect, while others are large enough to carry several tons of bombs. For the military, they provide reconnaissance, defuse roadside bombs, and strike high-value targets. On the civilian side, a flying robot provided the first detailed video of the Fukushima Daiichi nuclear plant after it was damaged in the March 2011 earthquake and tsunami. Robots are helping The International Group for Historic Aircraft Recovery search for Amelia Earhart’s Lockheed Electra. And, as more than three million YouTube viewers have seen, they can even play the theme song from the James Bond franchise.

But one thing they do have difficulty with is recovering after collisions. That’s where AirBurr, a flying robot, has an advantage. Its flexible body protects the robot should it crash into a wall. And if it falls to the ground, AirBurr—using a leg design inspired by locusts and beetles—can right itself and continue flying. (During flight the robot’s four carbon-fiber legs are rolled up.)

“It all started when we looked at insects, and how they fly,” says researcher Adam Klaptocz, in EPFL’s video, below. “Even though they manage to avoid most obstacles, they still manage to fly into windows and fly into walls, yet it’s ok. They don’t break. They fall to the ground, they get back up again, and they keep flying.” The main application of this type of robot, says Klaptocz, is to explore hard-to-reach places where humans—or even other robots—can’t navigate, such as irradiated nuclear power plants, caves, and collapsed mines.

While some flying robots can try to avoid collisions by using on-board sensors that allow it to create a map of the environment, such platforms are heavier, fragile, and typically don’t survive any accidental crashes. The AirBurr team decided to create a robot that would withstand routine bumps and jolts. This approach allows them to use cheaper, less-complex sensors, and lightens the robot’s weight.

Learn more about AirBurr and the work of researchers Adrien Briod, Adam Klaptocz, Przemyslaw Mariusz Kornatowski, and Jean-Christophe Zufferey.

In their efforts to “ignite a new era of lunar exploration,” GLXP wants more than just to send hardware to the moon. Along the way the teams must record their work and reach out through blogs and social media so that the rest of us (including the passionate but less engineering-inclined) can follow their progress. According to the rules, each team must write one blog post a week and post 45 minutes of video each quarter; Facebook and Twitter are not required, but many of the teams have incorporated them as well.

Amanda Stiles, GLXP’s Online Community and Google Liaison, says this about the online outreach requirement:

We hope that by encouraging the teams to tell their stories, the public will have the opportunity to get to know the personalities of the people involved with the competition and understand their motivations for pursuing the prize. These teams are pushing boundaries and doing great things in many arenas — technical, political, educational, and business, to name a few — all around the world, and we hope to showcase those efforts. And ultimately, when the winning teams eventually claim the prize purses then there will be well-documented stories of their trials, tribulations, and successes along the way.

GLXP recently redesigned their website so that it focuses more on these outreach efforts, with a streaming feed of all the competitors’ updates and pages for each team. Naturally, some of the output is better than others; many of the Twitter feeds don’t really seem to “live-tweet” the experience the way an observer might hope. Team Astrobotic Tech has one of the better Twitter feeds, with lots of interesting updates and links to pictures and video of their two Personal Exploration Rovers (PERs), Juno and Kosh.

Team Astrobotic Tech's Twitter update, featuring their GLXP rovers.

Here’s a particularly informative video from Team Italia describing their rover engineering.

Space exploration outreach group Evadot has been keeping a running scorecard for each section of the GLXP competition, which puts team Part-Time Scientists in the lead for social outreach, though we’re not sure if that’s for strictly following the quantity requirements or if it takes into account quality, as well.

The online outreach is just one part of an obviously much bigger and more difficult challenge. But as Evadot notes, GLXP “is NOT just a simple race to the moon. The point is the change it can bring through the competition. It’s not the race, it’s what happens because of the race.” And the hope is that this kind of outreach will, as Stiles puts it, ”encourage teams to be seen as modern-day space heroes,” inspiring not just by reaching a goal, but by bringing us all along for the ride.

Robonaut 2—the humanoid robot soon to be tested as an astronaut’s helper on the International Space Station—is being powered up for the first time this morning (screen shot at left). Since arriving on the space shuttle last February, the robot has been sitting on its pedestal, lifeless. It won’t be commanded to move for a couple of weeks (this is a slow process, partly due to limited crew time for testing). And it will be many months—and several upgrades—before we see a robotic spacewalk like the one below.

Still, it’s a start. Follow the action at Robonaut’s Twitter feed or Facebook page.

A REMUS 6000 closer to the surface. Credit: Brennan Phillips, Woods Hole Oceanographic Institution

Yesterday, the tenacious underwater ‘bot located at long last the remains of Air France flight 447, which plunged into the Atlantic Ocean on June 1, 2009.

The REMUS 6000 (Remote Environmental Monitoring UnitS) is the deepest diving AUV ever made, able to descend 6,000 meters, or almost 20,000 feet, below the ocean’s surface. It was developed jointly by the U.S. Naval Oceanographic Office, the Office of Naval Research, and the Woods Hole Oceanographic Institution.

Officials know only that the Airbus A330 wide-body jet encountered severe, high-altitude thunderstorms about three and a half hours into a flight from Rio de Janeiro to Paris, then fell from the sky. None of the 228 people onboard survived. Fifty-one bodies and some debris were found in the weeks following the accident, but search teams came up with precious little else, particularly answers.

What was clear was that the rest of the jet, including its black boxes (cockpit voice recorder and flight data recorder), and the remaining bodies, were somewhere deep in the fissures of the Mid-Atlantic Ridge more than two miles beneath the surface.

Left to right: Stephen Murphy, Mark Dennett, and Robin Littlefield of the REMUS 6000 Operations Group pose with one of the two AUVs owned by the Waitt Institute for Discovery. Both vehicles participated in the search for Air France 447. A third, owned by the Leibniz Institute for Marine Sciences in Germany, also participated. Credit: Tom Kleindinst, Woods Hole Oceanographic Institute

The REMUS 6000 was the ticket, able to cruise at up to four knots (4.6 mph) for up to 22 hours. Three vehicles equipped with an array of advanced sensors—including a high-resolution digital camera and side-scan sonar able to ping 2,000 feet out to both sides—combed the bottom in a lawn-mowing pattern. Launched and recovered from the vessel Alucia, which arrived on station March 25, the new effort marked the fourth attempt to locate the aircraft, and the second with the REMUS 6000. In a week, one of the AUVs found its quarry at 12,800 feet, about 2.5 miles down. They’ll narrow their search in the coming days for the black boxes.

Here are some photos of the airliner’s remains on the ocean floor at the web site of the French Bureau d’Enquêtes et d’Analyses pour la Sécurité de l’Aviation Civile, better known as the BEA. Together with Airbus and Air France, the BEA can now move forward with ideas for bringing pieces to the surface, and, perhaps, closure to the families.

Also, check out this video with scientists and engineers from the Waitt Institute for Discovery, which owns two REMUS 6000s, discussing a different project involving the AUV:

The design features a number of innovations, including active torsion of the wings and a torso that bends aerodynamically. And it only weighs a pound.

Strange that the video doesn’t show the thing taking off and landing. But it’s pretty cool to watch in flight (via Kurzweilai.net).

But this slow-mo video of an owl coming in for a landing is what really wowed me:

More and more, that’s becoming true for a variety of government agencies—and not just the defense department—as was evident at the UAV Technology Fair held yesterday at the Rayburn House office building in Washington, D.C. There were models of UAVs and plenty of vivid video graphics designed to show policy makers how far remotely piloted aircraft have come. The third Congressional UAV Caucus event for 2010, the fair was organized by Congressmen Howard “Buck” McKeon of California and Alan Mollohan of West Virginia.

Representative "Buck" McKeon, second from left, watches a demonstraton of a rotary UAV.

Clustered in the north foyer on the second floor of the building, with floor-to-ceiling windows that looked across Independence Avenue at the Capitol dome, officials from the military, industry, law enforcement, and NASA watched high-definition videos taken from UAVs that far surpassed the typical scenes we’re accustomed to seeing on the evening news. Visitors, including the general public, were invited to touch models of the vehicles and ask questions.

There were plenty of odd shapes and sizes, such as the Aurora Flight Sciences backpackable UAV called the Skate, after the flat marine animal of the same name. Several UAVs claim to be backpackable, but, says Aurora’s Patti Woodside, “When a soldier is already carrying 12o pounds of gear, he’s not going to add another 15 pounds.” Aurora’s answer: a tactical field UAV still in the prototype phase, only 2.5 pounds with a 2.5-mile flying radius, and a battery the size of a cell phone that keeps the UAV in the air for an hour. The whole thing folds up to the size of a laptop computer, with two props the size of a human hand that are held onto the vehicle by magnets. It takes off vertically, then travels forward at 55 miles an hour. Coolest of all: It was developed by four Aurora employees in their 20s who begged for $3,000 in seed money from the boss. “It started as a nights and weekends thing,” says program manager George Kiwada, “and went from a platform for testing [flight] control to what we always wanted to build: a UAV.”

The National Institute of Justice was there demonstrating a rotary UAV that is helping law enforcement personnel in Texas with their narcotics missions, search and rescue, forensics, and surveillance. Another rotary UAV, from Weber State University, had four props, two of which turn in one direction and two in the other, for stability. Persistent observation is the whole point. “It’s not important for this model to be a greyhound,” says Brad Stringer, Executive Director of the Utah Center for Aeronautical Innovation & Design, while holding the three-pound UAV with one hand.

Lording over the room were scaled-down models of the ever popular Global Hawk, which has revolutionized aerial warfare and surveillance. NASA is even using one to do hurricane studies. And they don’t need to keep it in Florida. They fly it leisurely from the Dryden Flight Research Center in California, all the way across the country and out into the Atlantic for hours on end, then all the way back to California. A 24-hour shift still requires three  pilot-shifts on the ground at the joy stick.

An overriding question: Why the need for a UAV caucus on Capitol Hill? Isn’t there enough demand for them without worrying about lobbying? “Great question,” answered Andy Graham, legislative fellow to McKeon, by email. “There are still a lot of misconceptions about unmanned systems. The common perception of the UAV is the Predator B Reaper drone firing missiles at terrorists.

The RQ-7 Shadow 200, by AAI Corp.

While the Reaper does perform this valuable mission, it represents only a very small fraction of the sizes, types, and missions that unmanned systems perform. As the technology and the demand for it evolve, existing airspace regulations become obsolete. One of the missions of the Caucus is to advocate for the military, industry, NASA, the Department of Homeland Security, the FAA, and other stakeholders to seek fair and equitable solutions to challenges created by UAV operations in the national air space. The Caucus brings together U.S. Representatives from both political parties who believe in the utility and longevity of unmanned aerial vehicles, and want to see their use expanded.”

And the people on the SMAVNET Project think they set a record for the largest number of flying robots (10) deployed at a single time outdoors.

DARPA’s Vulture program aims to build an unmanned vehicle that could stay up for five years. That’s still quite a stretch, considering that the current world record-holder, the Zephyr built by Qinetq, can stay aloft only for days.

Last month the U.K.-based company’s latest model, the Zephyr 7, shattered its own endurance record by flying non-stop for two weeks and 22 minutes.  The claim is still pending official recognition by the Fédération Aéronautique Internationale. Ya gotta love a record-setting aircraft that’s launched by hand.

As the space station gets its finishing touches (Atlantis carries up a new Russian storage module on tomorrow’s STS-132 mission), we’ll see some new gadgets come into play. One is the European Robotic Arm, due to be installed on the Russian Multipurpose Laboratory Module in 2012. A spare elbow for the arm is going up ahead of time on this shuttle flight.

The ERA is like a disembodied arm with hands—actually cylindrical “end effectors”— on either end. While one hand holds onto an attach point on the station’s exterior, the other is free to move cargo around. This video shows the ERA in action. The arm can also “walk” around the outside of the station from one attach point to another, with each step taking 30 minutes to an hour.

Unlike other station arms, the ERA can be operated either from inside the Russian lab module or from outside, by a spacewalking cosmonaut. For the first time, spacewalkers will be able to operate their own robot assistants, without having to ask one of their colleagues inside. At right is the control device they’ll use.

R2: Still an understudy, at least for now

So we thought the last of NASA’s rookie astronauts had flown, leaving only veterans on the final few space shuttle flights.

Not so fast.

One last rookie will be on board space shuttle Discovery when it blasts off in September for the STS-133 mission.

After years languishing as a laboratory-only project, the Johnson Space Center’s mechanical astronaut Robonaut has finally been called up. Robonaut 2, as the current model is called, is scheduled to take up permanent residence on the space station, where it will be used to test the practicality of anthropomorphic robots working alongside humans. It will be operated from laptops on the station and on the ground, and at first will be anchored in one spot in the Destiny laboratory. But eventually R2′s creators want to send the mechanical man out on a spacewalk.

The project’s deputy manager gives this rundown:

R2 will join the Dextre humanoid robot, which is already installed on the station, where it handles outside jobs. If you think such machines have no place in future space exploration, talk to the Japanese, who reportedly are mulling a humanoid robot to land on the moon.  Most lunar robot concepts still use wheeled rovers, but the idea of a mechanical Neil Armstrong  no longer seems far-fetched, not when you see the kind of progress that companies like Boston Dynamics are making in simulating natural walking:

Here’s the video from MIT’s TechTV: