We've been incredibly excited to see the progress that Boston Dynamics has been making on ATLAS in preparation for the DARPA Robotics Challenge, but we had no idea what to expect from the challenge's Track A teams, each of whom will be designing and building their own robot with capabilities comparable to what we've seen ATLAS do. Today, October 24, is opening day for the DARPA Robotics Challenge, or DRC. The press release sums it up nicely: "over the next two years, teams will compete to develop and put to the test hardware and software designed to enable robots to assist humans in emergency response when a disaster strikes." The first half of this is the hardware: DARPA is promising that an "advanced variation" of ATLAS (which is what the above picture is showing) will be ready to go by June of 2013, and will be provided to the advancing Track B and C teams (see our previous post on the DRC for more details on the tracks). As for the simulation software (pictured above), OSRF has been working very, very hard, and the DRC Simulator is currently available in beta version 1.0.
Boston Dynamics unveiled yesterday a massively upgraded version of its ATLAS humanoid that is smaller, lighter, and more agile. In a video, the new robot is seen walking untethered in snow-covered woods, lifting and placing boxes on shelves, and even face-planting and immediately getting up unscathed after being pushed by an engineer. As one observer commented, "We expected [ATLAS] to turn around and blast that guy with a laser beam." What is perhaps most impressive about the "next generation" ATLAS is that it is just a huge technological leap forward compared to its predecessor, which was already a pretty incredible robot. The new ATLAS can do things we've never seen other robots doing before, making it one of the most advanced humanoids in existence.
It's always exciting when a new robot arrives in your lab. Usually, the more expensive the robot is, the more exciting it is. With the possible exception of Boston Dynamics' ATLAS, NASA's Valkyrie has got to be one of the most expensive humanoid robots ever made, and last year, NASA promised to give away (or, at least, lend) three of them to universities in the hope that Valkyrie will learn some new skills. Within the last few weeks, the University of Massachusetts Lowell, which teamed up with Northeastern University in Boston, Mass., took delivery of their fancy new robot, as did MIT and the University of Edinburgh in Scotland. We talked to Holly Yanco at UMass Lowell and Taskin Padir at Northeastern, along with Sethu Vijayakumar at Edinburgh and Russ Tedrake at MIT, about what it's like to have a smokin' hot space robot show up on your doorstep in a bunch of pieces.
Humanoid robots have a very distinctive walk. Even Boston Dynamics' own Atlas uses this crouching sort of squat-walk to get around, because those perpetually bent legs are how it keeps from falling over. This sort of gait is so common with humanoid robots that it's become the "normal" robot gait, but it's also not at all the way that humans walk. We walk with straight legs, locking our knees with each stride, because it's much easier to support our weight that way. You can try it for yourself: that bent knee "bipedal robot" walk gets tiring to keep up, because your leg muscles always have to be engaged.
Like humans and most terrestrial animals, legged robots need the ability to move over rugged terrain to be useful in applications such as disaster response and search and rescue. However, designing control algorithms that can handle discrete footholds (like rubble or stepping stones) is challenging, because there are strict constraints on foot placement that can't be violated and the motion of these systems are governed by complex dynamical equations. By leveraging recent advances in optimal and nonlinear control systems, our labs at the University of California, Berkeley, and Carnegie Mellon University have demonstrated dynamic walking on stepping stones on the ATRIAS robot, even when the distance between the stones as well as their height is varied randomly. Legged robots are incredible machines capable of navigating over unstructured and uneven terrain. They're much more versatile than their wheeled counterparts, which have a hard time navigating terrain with gaps or significant changes in height.