Reproduce “Transporting Unsecured Stacked Payloads with a Quadrupedal Robot via Multi-Objective Reinforcement Learning” 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
Nobuo Namura, Masayuki Hiromoto, Kento Uemura, Hironobu Sasaki, Kanata Suzuki
PAMORT teaches a quadruped controller to balance locomotion and stability when carrying unsecured boxes across terrain. It trains a policy across movement–stability preferences, then adjusts them online using robot proprioception, without payload sensors. On a Unitree Go2, it achieved 0.850 mean success versus 0.675 for task-adapted TAR* across eight real-world tasks, including an unseen three-box stack; evidence is limited to rigid boxes and two objectives.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.