Reproduce “Planning for Tabletop Object Rearrangement” 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
Jiaming Hu, Jan Szczekulski, Sudhansh Peddabomma, Henrik I. Christensen
An A*-based algorithm for tabletop object rearrangement, STRAP, enhances scalability and solution quality for stationary and mobile robots. It yields solutions with comparable or lower travel costs than state-of-the-art methods like ORLA* while requiring substantially less computation time.
Key insights
- A more complete state representation, explicitly including the robot's pose, significantly contributes to finding higher-quality (lower travel cost) plans, particularly for mobile robots.
- Aggressively attempting to construct a goal plan during the A* search efficiently prunes the search space, allowing STRAP to find good solutions much faster than traditional A* methods.
- Iterative refinement of buffer locations in the generated plans can further reduce overall travel costs without sacrificing computational efficiency.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.