U of M's Marlo robot uses algorithms to conquer uneven terrain

Robots walking unaided on flat ground is tough enough as it is -- just look at last year's class of DARPA Challenge failures -- so when one can handle uneven terrain in any direction (not just a straight line), we take notice. The latest example is Marlo, a joint project between University of Michigan's Jessy Grizzle and Oregon State University's Jonathan Hurst. The key difference here is how it achieves this feat: a bank of algorithms containing different instructions for different walking styles.

Analyzing data from sensors in the biped's knees, hips and torso, Marlo adjusts walking style on the fly, pulling from a library of 15 pre-programmed gaits and blending them based on ground-cover or inclination angle.

Marlo's speed and direction is determined by a user holding an Xbox controller, but anything other than that -- like movement speed -- is handled by the bot itself. What's more, the school says that this algorithm is general enough that other robots could use it as a baseline for movement. And more than just fueling your nightmares of the impending robocalypse, this has implications for us fleshy humans too: The team says that this tech could extend to robotic prosthetics that'd make walking easier for lower-limb amputees.