We're used to seeing quadrocopters perform feats of aerial acrobatics or coordinated dances in a strictly controlled lab environment, where position sensors can give them precision location info. But swarming quadrocopters outside the lab is trickier, and also more of a hint at their future real-life use. That makes this clip of a performance of 49 AscTec Hummingbird quadrocopters at night in Linz in Austria even more amazing, especially when you add in the fact it's a record for coordinated flight of the vehicles. The son et lumiere show hints at a near future where swarming drones help with everything from surveillance to construction tasks.
Boston Dynamics is working on a robot modeled on a cheetah because of the stability of a four-legged robot platform combined with the extreme agility and speed a living cheetah can manage. Now, by tweaking the algorithm that controls its springy legs and back and boosting its power, BD has managed to increase its speed to set a new record of 28.3 miles an hour. This is a tiny bit faster than the official record for humans, set by Usain Bolt in 2009 with a 27.8 mph sprint during a 100m race. The lab prototype is the test bed for a future robot that'll be untethered and capable of maneuvering at speed on outdoors surfaces—for military purposes, among others.
While the ecological benefits of solar photovoltaic power are well known, one big barrier to adoption for PV systems is that they can be expensive if required to be efficient. That's because for maximum light-grabbing they have to be angled toward the sun both horizontally and vertically, and this requires motorized mounts on panels, adding to the cost. Robot maker QBotix has revealed a clever solution that fixes all this: An array of panels that are connected to a rail system and a single robot that zips up and down the array, constantly tweaking each panel's angle before moving on. It's zippy enough to adjust 200 panels, and the firm says it bumps up PV efficiency by 40% over static-plate installations.
NASA's Oil Bot. A Norwegian company, Robotic Drilling Systems, has acquired a rather unusual new parter for its future plans for oil discovery and drilling—NASA. Specifically, the company's signed an info-sharing agreement to see if there's any technology on NASA's remarkable Curiosity robot, even now roaming on Mars, that it could benefit from. Ultimately RDS foresees the intensive almost entirely mandraulic work on an oil rig being replaced with robotic systems, even to the point rigs will self-navigate to new drill sites.
Hurricane Isaac's Robo-Scientist.
Underwater Search Bot. Police in New Jersey have just been testing a new team member—the SeaBotix LBV-200-4. It's a robot—well, technically a tiny remote controlled submarine—and it's designed to help police in situations where divers would have a hard task locating underwater evidence or in dangerous situations. It can even see in otherwise zero-visibility situations via its sonar imaging system, and it has grasper tools so it can bring up items it finds underwater.
Robots Mess Up Hugo Awards. Robots of a different kind received really bad press this week: Copyright protection bots which monitor the internet for streaming video content that (potentially) violates rights-holders licenses. The online stream of the prestigious Hugo awards for science fiction was killed by a system—which may have been run by stream-provider Ustream—which decided that some TV clips used in the ceremony, completely legally, were infringing. Ustream then deemed it impossible to reactivate the feed.
We've covered the idea of highly robotized workforces recently, and it seems to have become something of a hot topic. Not long ago the New York Times tackled the issue in a long article that argued a "new wave of deft robots is changing global industry".
The Times article showcased two highly robotized Philips factories and also mentioned Tesla's production facility, perhaps one of the most heavily robot-populated car factories yet. But an article in Supply Chain Digest takes these ideas further and suggests a near future of "extreme use" of robots in factories. The publication notes that robots have already evolved dramatically beyond the failed systems tried by GM in the 1980s, and highlights the work of former MIT professor Rodney Brooks—now building robots for his own company Rethink. Brooks recently spoke at an MIT forum on the future of manufacturing in the U.S. and described his work in creating robots that are agile, flexible enough to have multiple uses in production jobs, and which are safe and smart enough to work right alongside humans.
Rethink, which is said to be planning its first commercial products by the end of this year, thinks a new generation of smart, adaptable industrial robots will completely revolutionize the manufacturing paradigm.
That's quite a claim, until you remember that the last time manufacturing's paradigm was really shattered was Henry Ford's assembly line thinking in around 1908. Even robotic car factories haven't moved far beyond Ford's original core idea.
But "extreme" robotizing of manufacturing, which promises huge efficiency boosts, faster production times and the attendant ecological benefits of these changes, is going to be a tricky matter. In recent months a lot of attention has been given to arguments that companies like Apple and Google shouldn't be producing their worldwide hot-selling mobile devices in China, but in the U.S. instead. There's undoubtedly a little nationalism and even jingoism behind some of this thinking, but the debate misses out on a key point: Assuming 21st century manufacturing facilities were set up in the U.S., leveraging the billions of dollars of a firm like Apple, would they actually create much wealth and, more importantly, employment?
Chinese manufacturing in firms like Foxconn relies on cheap Chinese labor costs, abundant resources and the infrastructure of nearby firms who make subcomponents (in factories that themselves leverage cheap labor). Even under these promising conditions, Foxconn's boss has vowed to improve efficiency and lower costs in the near future by employing a million robots.
Considering increased labor costs in the U.S. compared with Chinese workers, if the same sort of cost efficiencies were to be achieved in a Foxconn-like facility, one can imagine that a very highly robotized production line, perhaps an "extreme" one, with very low human intervention, would be a great place to start. Such a facility may require very few skilled human monitors, a small workforce for simplistic tasks that are even beneath robots. The only local economic boon a highly robotized factory may deliver may be in consuming water and power, in construction of the factory and in transport of components and finished goods. If you consider that the robots themselves may not even be U.S.-sourced, the matter gets even more thorny.
What all this tells us is one important fact: Robotic workforces really are on the way, and they're really going to upset the manufacturing applecart. They may even upset the entire manufacturing workforce, and thus the dynamics of dozens of other industries that are third parties in the manufacturing process. Before you know it, this issue may become highly political.
[Image: Flickr user jurvetson]