Reproduce “ASAP: Automated Sequence Planning for Complex Robotic Assembly with Physical Feasibility” 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
Yunsheng Tian, Karl D.D. Willis, Bassel Al Omari, Jieliang Luo, Pingchuan Ma, Yichen Li, Farhad Javid, Edward Gu and 5 more
ASAP is a framework for automated sequence planning for complex robotic assembly that integrates comprehensive physical feasibility checks, including gravitational stability and collision-free paths. It achieved up to 82.08% success in generating physically feasible plans on a large dataset of complex 3D assemblies, significantly outperforming baseline methods, and was demonstrated on a real-world robotic setup.
Key insights
- Integrating comprehensive physical feasibility constraints, including gravitational stability and collision-free paths, is essential for generating truly executable robotic assembly sequences for complex, general-shaped products.
- Learning-based guidance via Graph Neural Networks can effectively navigate the combinatorial search space for assembly sequences, leading to substantially higher success rates compared to geometric heuristics or random approaches.
- Computational efficiency for complex physics checks can be achieved through greedy approximations (e.g., for stability), which provide an order of magnitude speed-up with approximately 90% accuracy.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.