Ep#104: Light-Loco-Parkour: Versatile Perceptive Whole-Body Locomotion via Multi-Skill Distillation

Podcast: RoboPapers · Chris Paxton and Michael Cho

Opublikowano: 2026-09-14

One of the key advantages of legged robots like humanoids should be how effectively they can move across a wide variety of terrain types to accomplish their task. But Light-Loco-Parkour from the team at Light Origins aims to change that: using only onboard sensing, they show a policy which can decide when to walk, vault, climb, or otherwise traverse as it moves through a complex environment. Unlike many others, it uses sparse seeds instead of relying on a large motion corpus, learning when to use its skills to move around without specific sub-task labels. Xiaodao Chen and Yuntao Ma...