Novel Algorithms for Smoothly Differentiable and Efficiently Vectorizable Contact Manifold Construction
Onur Beker, A. René Geist, Anselm Paulus, Georg Martius · 2026
Read Novel Algorithms for Smoothly Differentiable and Efficiently Vectorizable Contact Manifold Construction by Onur Beker, A. René Geist, Anselm Paulus, Georg Martius (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-17538.txt.
Description
Beker, Geist, Paulus, and Martius attack the collision-detection bottleneck in robot simulation for contact-rich tasks. Standard routines are tuned for CPU speed and small memory, but their branching logic blocks the smooth gradients and GPU vectorization that modern optimization needs. The authors design a fresh pipeline built for twice differentiability and massive parallelization, covering how contact surfaces between rigid bodies are constructed. The goal is faster, gradient-aware planning where robots touch, push, and grasp.
Catalog record
| CREATOR | Onur Beker, A. René Geist, Anselm Paulus, Georg Martius |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 24551 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.17538