Extreme Parkour with Legged Robots
About
Humans can perform parkour by traversing obstacles in a highly dynamic fashion requiring precise eye-muscle coordination and movement. Getting robots to do the same task requires overcoming similar challenges. Classically, this is done by independently engineering perception, actuation, and control systems to very low tolerances. This restricts them to tightly controlled settings such as a predetermined obstacle course in labs. In contrast, humans are able to learn parkour through practice without significantly changing their underlying biology. In this paper, we take a similar approach to developing robot parkour on a small low-cost robot with imprecise actuation and a single front-facing depth camera for perception which is low-frequency, jittery, and prone to artifacts. We show how a single neural net policy operating directly from a camera image, trained in simulation with large-scale RL, can overcome imprecise sensing and actuation to output highly precise control behavior end-to-end. We show our robot can perform a high jump on obstacles 2x its height, long jump across gaps 2x its length, do a handstand and run across tilted ramps, and generalize to novel obstacle courses with different physical properties. Parkour videos at https://extreme-parkour.github.io/
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Robot Parkour | Surmounting Simulation | Traversal Rate68.2 | 18 | |
| Robot Parkour | Wall-assisted Gap (Simulation) | Success Rate77.2 | 18 | |
| Locomotion | Procedurally generated terrains (Offline evaluation) | TSR61.02 | 9 | |
| Robot Parkour | Stepping Stone Simulation | Success Rate (SR)81.1 | 9 | |
| Balancing Performance | Xi 10,000 trajectories (eval) | Collision Rate89.1 | 8 | |
| Parkour | Isaac Gym Hurdles | Success Rate18 | 4 | |
| Parkour | Isaac Gym (Steps) | Success Rate14 | 4 | |
| Parkour | Isaac Gym (Gaps) | Success Rate10 | 4 | |
| Parkour | Isaac Gym Overall Average | Success Rate16 | 4 |