REINFORCEMENT LEARNING TRIPLES SPOT'S RUNNING SPEED

Summary from the TLDR newsletter

The research version of Boston Dynamics' Spot quadruped robot comes with a low-level application programming interface that allows direct control of Spot's joints - it also unlocks some significant performance improvements, including a much faster running speed. The Robotics and AI Institute has been applying reality-grounded reinforcement learning techniques to enable much higher performance from Spot. This article looks at the improvements these techniques have made and also a brand-new hardware platform - an autonomous bicycle that can jump.