Robustness without Wrinkles: Parallel Simulation and Robust MPC for Certified Deformable Manipulation
arXiv:2606.14188v1 Announce Type: cross Abstract: We present CORD-SLS, a real-time control method for safe deformable object manipulation, with a focus on ropes and cloth. At its core is a GPU-parallel differentiable simulator with contact smoothing which enables efficient gradient-based planning through intermittent contact. To robustly satisfy constraints under model and sensing uncertainty, we develop a real-time, GPU-parallel output-feedback robust model predictive control (MPC) algorithm that plans with this simulator. We further show that the simulator accelerates model-based RL for trai
Advances in GPU-parallel computing and differentiable simulation techniques are converging to enable new capabilities in robotic manipulation of complex objects.
Achieving robust and safe manipulation of deformable objects like cloth and ropes is a critical bottleneck for general-purpose robotic autonomy in unstructured environments.
This research provides a real-time, certifiable method for robots to handle previously challenging deformable materials, expanding their potential applications beyond rigid object manipulation.
- · Robotics companies
- · Logistics and manufacturing sectors
- · AI hardware developers (GPUs)
- · Human labor in highly repetitive handling of deformable objects
Robots can more capably perform tasks involving textiles, cables, food, and other flexible items, improving automation in diverse industries.
The increased dexterity of robots will open new market opportunities for automated services and products in areas like elder care or clothing production.
As robotic manipulation capabilities mature, they will contribute to the broader deployment of humanoid robots and advanced manufacturing processes, impacting global supply chains.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.AI