Robotics & AI

Simulation enables people with a personal computer to develop very interesting robots, cars, spaceship, and an enormous range of scientific effects with the main limiting factors becoming time and imagination. A novice user with little to no coding experience can use simulation; developing interesting applications in a game-like environment.

Commenter Cynox was browsing through the 137 years of Popular Science magazine which are now available online, and he noticed this robot in the September 1984 issue. Called Odex I, it was developed by a (now apparently defunct) company called Odetics. Odex was six and a half feet tall, had six legs, and was fully capable of walking. Although it only weighed 370 pounds, each of its legs could lift 400 pounds. It could dead lift some 2100 pounds, and carry 900 pounds while walking at normal speed (which was about 18 inches per second). Odex used a tripod gait, and the fishbowl thing on top contained sensors that helped it avoid obstacles. It was one of the first robots with an onboard computer that helped coordinate all of its limbs. Since the limbs could articulate themselves in several directions independently, Odex was able to rapidly change its limb configuration to squeeze through tight spaces, move quickly, or lift stuff. It was able to climb into the back of a truck through a combination of automated step behaviors and teleoperation, which was pretty damn good for 1984.

Space robotics may appear to be a purely scientific endeavor -- brave little rovers exploring planets in search of life -- but it turns out there's a multi-million dollar market in space just waiting for the right kind of robot. This market is satellite servicing. Geostationary communication satellites fire small thrusters to stay in orbit. When they run out of fuel (typically helium or hydrazine), or when a battery or gyroscope fails, these expensive satellites often have to be abandoned, becoming just another piece of space junk, even though their mechanical systems and electronics work fine.

In a recent video, Willow Garage researcher Eitan Marder-Eppstein describes the open-source navigation stack they've released as version 1.0. The code, available at http://ros.org/wiki/navigation, was designed to be flexible and cross-platform, he says, and could be used in anything from a small iRobot Create-based bot to a large multi-sensor robot like Willow's own PR2 (which Spectrum has covered in detail here and here).

Willow Garage is organizing a workshop at the IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR) 2010 in San Francisco to discuss the intersection of computer vision with human-robot interaction. Willow Garage is the hardware and open source software organization behind the Robot Operating System (ROS) and the PR robot development platform. Here’s a recent video from Willow Garage of work done at the University of Illinois on how robots can be taught to perceive images:

I remember when, a decade ago, Kevin Warwick, a professor at the University of Reading, in the U.K., implanted a radio chip in his own arm. The feat caused quite a stir. The implant allowed him to operate doors, lights, and computers without touching anything. On a second version of the project he could even control an electric wheelchair and produce artificial sensations in his brain using the implanted chip. Warwick had become, in his own words, a cyborg. The idea of a cyborg -- a human-machine hybrid -- is common in science fiction and although the term dates back to the 1960s it still generates a lot of curiosity. I often hear people asking, When will we become cyborgs? When will humans and machines merge? Although some researchers might have specific time frames in mind, I think a better answer is: It's already happening. When we look back at the history of technology, we tend to see distinct periods -- before the PC and after the PC, before the Internet and after the Internet, and so forth -- but in reality most technological advances unfold slowly and gradually. That's particularly true with the technologies that are allowing us to modify and enhance our bodies.

Walking, swallowing, respiration and many other key functions in humans and other animals are controlled by Central Pattern Generators (CPGs). In essence, CPGs are small, autonomous neural networks that produce rhythmic outputs, usually found in animal's spinal cords rather than their brains. Their relative simplicity and obvious success in biological systems has led to some success in using CPGs in robotics. However, current systems are restricted to very simple CPGs (e.g., restricted to a single walking gait). A recent breakthrough at the BCCN at the University of Göttingen, Germany has now allowed to achieve 11 basic behavioral patterns (various gaits, orienting, taxis, self-protection) from a single CPG, closing in on the 10–20 different basic behavioral patterns found in a typical cockroach. The trick: Work with a chaotic, rather than a stable periodic CPG regime. For more on CPGs, listen to the latest episode of the Robots podcast on Chaos Control, which interviews Poramate Manoonpong, one of the lead researchers in Göttingen, and Alex Pitti from the University of Tokyo who uses chaos controllers that can synchronize to the dynamics of the body they are controlling.

This little helicopter is able to understand you when you tell it what to do. No pushing buttons, no using special commands, you just tell it where you want it to go and (eventually) it goes. Of course, I’m sure it required a bit of work to define where “door” and “elevator” and “window” are, but it’s a much more intuitive way to control a UAV that works when your hands are full, when you’re stressed (think military), or simply when you have no idea now to control a UAV. I don’t have much in the way of other details on this project, besides the fact that it probably comes from the Robust Robotics Group at MIT, and possibly from someone who lives in this dorm. How do I know? Well, one of the research goals of the RRG is “to build social robots that can quickly learn what people want without being annoying or intrusive,” and this video is on the same YouTube channel. ‘Nuff said.

This robot is named Cody, and he’s from the Georgia Tech’s Healthcare Robotics Lab. I don’t know why the robot is called Cody… It kinda seems like it should stand for something. You know, as in, C.O.D.Y. Anybody got anything? No? Okay then.

Rosie, the robot who kept house for the title family in "The Jetsons," a 1960s animated television show, has at last come alive—sort of. Before you'll see a robot slicing cucumbers in your kitchen, researchers will need to make these mechanical servants smarter. Here's how three teams are tackling this challenge.

We are entering the robotic age. All over the world, we see research projects and companies working on realistic, market driven robots, with impressive realizations ranging from intelligent vacuum cleaners to humanoid robots.   This is a very exciting time and some people see in the current situation many common points with the early days of the computer industry. However, like PCs in the early 80's, today's robots are still incompatible in term of software. There is yet no standard way to reuse one component from one robot to the other, which is needed to have a real software industry bootstraping. And most attempts have been failing to provide tools genuinely adapted to the complex need of robot programming.   Here at Gostai, we believe that the industry needs a powerful robotics software platform, ready to face the challenges of Artificial Intelligence and autonomous robots programming.

ALSOK, a security firm that specializes in robot guards, has sent their drones into shopping malls, office buildings, and museums.  This video shows one of them patrolling an art gallery.  Not surprisingly, even in Japan, the sight of a robot patrolling its beat is more than enough to distract some of the visitors from the actual works of art!

GetRobo has pointed out a new website by Francisco Paz, which focuses on his experience building an open source personal robot called Qbo.  From the few images on the site Qbo looks remarkably well made and quite similar to NEC’s PaPeRo, meaning it might be used to experiment with image processing, speech recognition, speech synthesis, and (assuming it has wheels) obstacle detection and SLAM.  He also mentions in his blog some of the open source software that’s out in the wild such as OpenCV, Festival, and Sphinx, which would allow you to do some of that.

NAMO (Novel Articulated MObile platform)  is a humanoid robot built by The Institute of Field Robotics (FIBO) at King Mongkut’s University of Technology Thonburi in Thailand. FIBO is active in the RoboCup scene and have developed a wide range of robot types, including an experimental biped.  NAMO was unveiled on March 29th 2010, serving as FIBO’s mascot as part of the university’s 50 year anniversary celebrations.  NAMO will be used to welcome people to the university and may be deployed at museums.  Given its friendly appearance and functionality, it could be used to research human robot interaction and communication. NAMO is 130cm (4′3″) tall and has 16 degrees of freedom.  It moves on a stable three-wheeled omnidirectional base, and is equipped with a Blackfin camera for its vision system.  It is capable of simple gesture recognition, visually tracks humans or objects of interest automatically, and can speak a few phrases in a child-like voice (in Thai).

Remember A-pod, the realistic ant-like hexapod from last year?  Well its creator Kare Halvorsen has uploaded a brand new video showcasing its improved capabilities, and it’s a stunner.  His last video, posted around this time last year, went viral due to the robot’s realistic movements. This year, he ups the ante by showing it walking around and manipulating objects. Some of his past robot projects can be seen in brief snippets, and they’re not too shabby either.  Imagine a horde of these guys with sophisticated A.I.!