The team spent a week fast-tracking Cassie through a year’s worth of simulated training designed to determine the most effective gait. But it wasn’t simply a matter of speed. For the Guinness World Record to stand, Cassie had to start in a standing pose, and then return to that pose after crossing the finish line rather than simply tumble over. This meant Cassie had to use two neural networks, one for running fast and one for standing still, and gracefully transition between the two. Ultimately, Cassie completed the 100-meter sprint in 24.73 seconds, establishing a Guinness World Record for a bipedal robot. This is a great deal slower than the sub-10-second times run by the world’s best sprinters, but the researchers believe progress will only accelerate from here. You can watch Cassie’s record-setting dash here.
Read more of this story at Slashdot.