Lodos-v2: An Efficient Primal Formulation of the Exact Frictional Contact Problem
Lodos-v2: An Efficient Primal Formulation of the Exact Frictional Contact Problem
This paper introduces Lodos-v2, a GPU-based simulation framework designed for high-throughput rigid-body dynamics with contact. It resolves limitations in its predecessor, Lodos-v1, by solving the exact frictional contact problem through sequential convex relaxations in primal form and extending the contact-group reparameterization scheme to integrate with full simulator constraints like joint limits. Experiments show improved throughput over established frameworks, with source code available on GitHub.
The authors present Lodos-v2, a GPU-based simulation framework that addresses physical artifacts like lift-off and creep found in its predecessor by solving the exact frictional contact problem via primal form convex relaxations.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- Environment
- SimulationOpen source S2
- code: Available Artifact link Open source S5
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
arXiv:2611.08868v1 Announce Type: new Abstract: Lodos-v2 is a GPU-based simulation framework for massively parallel rigid-body dynamics with contact. It is specifically designed to achieve high throughput in scenes involving collisions among large numbers of detailed meshes. Its predecessor, Lodos-v1, had two main limitations. First, it solved the standard convex relaxation of cone-complementarity-based contact dynam
Open source S2
cs, enabling its integration with the other constraints of a full-fledged simulator. Experiments demonstrate that these contributions result in throughput improvements over established rigid-body simulation frameworks. The source code is available at https://github.com/martius-lab/lodos.
Open source S5
Source：arXiv Robotics — research abstracts · arxiv.org