REVIEW 5 cited by
Extreme Parkour with Legged Robots
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
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/
Forward citations
Cited by 5 Pith papers
-
High-speed control and navigation for quadrupedal robots on complex and discrete terrain
A hierarchical planner-plus-tracker system enables a quadruped to run on walls, clear a 1.3 m gap, and navigate discrete terrain at up to 4 m/s using a competitive generative curriculum.
-
Weights or Skills? A Survey of Robot-Learning Techniques: from Action-Predicting Weights to Robots that Write their Own Skills
A taxonomy of robot learning on a weights-versus-skills axis, with a five-rung self-improvement ladder whose top cell (feedback plus memory plus search) holds only a few recent systems.
-
Scoop-and-Toss: Dynamic Object Collection for Quadrupedal Systems
A simulation study shows that a quadruped with a passive leg-mounted scoop and a back tray can learn to scoop objects and toss them into the tray, collecting multiple objects via a hierarchical policy.
-
Motion Priors Reimagined: Adapting Flat-Terrain Skills for Complex Quadruped Mobility
A two-level reinforcement learning system that first learns animal-like gaits from flat-ground motion data, then learns small joint corrections that let a quadruped robot traverse rough terrain and navigate to goals.
-
First Order Model-Based RL through Decoupled Backpropagation
By computing gradients through a learned dynamics model while unrolling trajectories in the real simulator, DMO achieves SHAC-level sample efficiency with standard simulators and deploys on a real quadruped.
Discussion (0). Continue with ORCID to comment.