Quadruped robot parkours over gaps and walls with new agility framework

Korean researchers have developed a control framework allowing four-legged robots to perform parkour-like movements on complex terrain. The system uses a neural planner to find safe footholds and a tracker for precise steps, enabling a robot to sprint up walls, leap gaps, and navigate uneven ground at high speeds. This advancement could lead to robots operating in disaster zones or dangerous environments inaccessible to humans.