Reproduce “End-to-End Reinforcement Learning for Quadrupedal Jumping in Lunar Gravity” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Levin Müller, Oliver Fischer, Marco Trentini, Yusuke Tanaka, Marco Hutter
The authors present a single policy for Magnecko’s lunar quadruped to execute commanded point-to-point jumps. It converts a requested landing point and apex height into a launch velocity, then manages launch, flight, and landing. In simulation, jumps with launch elevations of at least 60° generally land within 0.3 m; hardware tests demonstrate only in-place vertical jumping under gravity offloading.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.