Reproduce “Decentralized Adaptive Control for Collaborative Manipulation of Rigid Bodies” 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
Preston Culbertson, Jean-Jacques E. Slotine, Mac Schwager
This research introduces a decentralized adaptive control scheme that allows robot teams to collaboratively manipulate rigid objects without requiring prior knowledge of the object's physical properties or precise robot attachment locations. The method guarantees almost-global asymptotic trajectory tracking and boundedness of all system signals, validated through 3D simulations and planar hardware experiments.
Key insights
- Adaptation of both object inertial parameters and robot attachment points is critical for robust collaborative manipulation, significantly outperforming methods that adapt only inertial properties.
- The proposed decentralized adaptive controller provides theoretical guarantees for almost-global asymptotic trajectory tracking and boundedness of all closed-loop signals, even with substantial parameter uncertainty.
- Individual parameter estimates do not necessarily converge to their true values due to lack of persistent excitation in the multi-agent system, though the average parameter error converges.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.