LEONARDO can walk, slackline, and skateboard, but it’s definitely slower than Oregon State’s bipedal robot. Cassie managed to achieve a 24.73-second run at OSU’s Whyte Track and Field Center, starting from a standing position and returning to that position after the 100-meter sprint, without any falls.
Cassie is also the first bipedal robot to utilize machine learning to control a running gait on outdoor terrain and complete a 5K on Oregon State’s campus untethered on just a single battery charge. How is this possible? The robot has knees that bend like an ostrich’s and does not use any cameras or external sensors. It was trained for the equivalent of 1-year in a simulation environment through a computing technique known as parallelization, which enabled it to go through a range of training experiences simultaneously.
- Segmented Dynamic Color: Featuring RGBICWW technology, this silver light bar offers tunable cool-to-warm white and vivid RGBIC colors for immersive...
- Music Sync Lighting: The built-in mic detects ambient sound to dynamically sync color and effects. Enjoy ultra-responsive, natural lighting...
- Matter-Supported Smart TV Lightbar: Voice-control colors, brightness & power via Alexa, Google Home. Create cozy scenes, set schedules, and sync with...
We have been building the understanding to achieve this world record over the past several years, running a 5K and also going up and down stairs. Machine learning approaches have long been used for pattern recognition, such as image recognition, but generating control behaviors for robots is new and different,” said Devin Crowley, graduate student.
