aiqbench analysis
This hierarchical RL-MPC framework addresses data inefficiency and sim-to-real transfer gaps in dexterous manipulation by decoupling high-level contact geometry planning from low-level contact dynamics control. The approach achieves a tenfold reduction in training data requirements while maintaining robust performance across diverse non-prehensile tasks like pushing and reorientation. By leveraging an object-centric interface for contact intentions, the method offers a practical pathway for scaling robotic manipulation systems without relying on massive end-to-end datasets.
Where to Touch, How to Contact: Hierarchical RL-MPC Framework for Geometry-Aware Long-Horizon Dexterous Manipulation
arXiv:2601.10930v3 Announce Type: replace Abstract: A key challenge in contact-rich dexterous manipulation is the need to jointly reason over global geometry and nonsmooth contact dynamics. End-to-end policies bypass this complexity, but often require large amounts of data and transfer poorly from simulation to reality. We address the limitations with a simple insight: dexterous manipulation is inherently hierarchical--at a high level, a robot decides where to touch (geometry); at a low level it determines how to move the object through contact dynamics. Building on this insight, we propose a hierarchical RL--MPC framework in which a high-level reinforcement learning (RL) policy predicts a contact intention, a novel object-centric interface that specifies (i) an object-surface contact location and (ii) a post-contact object subgoal pose. Conditioned on the contact intention, a low-level contact-implicit model predictive control (MPC) optimizes local contact modes and real-time (re)plans through contact dynamics to generate robot actions that robustly move the object toward each subgoal. We evaluate the framework on non-prehensile tasks, including geometry-generalized pushing across diverse object shapes, pivoting/flipping-based object reorientation, and environment-assisted object repositioning. It achieves high success rate with substantially reduced data (10 times less than end-to-end baselines), highly robust performance, and zero-shot sim-to-real transfer.
Abstract:A key challenge in contact-rich dexterous manipulation is the need to jointly reason over global geometry and nonsmooth contact dynamics. End-to-end policies bypass this complexity, but often require large amounts of data and transfer poorly from simulation to reality. We address the limitations with a simple insight: dexterous manipulation is inherently hierarchical--at a high level, a robot decides where to touch (geometry); at a low level it determines how to move the object through contact dynamics. Building on this insight, we propose a hierarchical RL--MPC framework in which a high-level reinforcement learning (RL) policy predicts a contact intention, a novel object-centric interface that specifies (i) an object-surface contact location and (ii) a post-contact object subgoal pose. Conditioned on the contact intention, a low-level contact-implicit model predictive control (MPC) optimizes local contact modes and real-time (re)plans through contact dynamics to generate robot actions that robustly move the object toward each subgoal. We evaluate the framework on non-prehensile tasks, including geometry-generalized pushing across diverse object shapes, pivoting/flipping-based object reorientation, and environment-assisted object repositioning. It achieves high success rate with substantially reduced data (10 times less than end-to-end baselines), highly robust performance, and zero-shot sim-to-real transfer.
Robotics (cs.RO)
arXiv:2601.10930 [cs.RO]
(or arXiv:2601.10930v3 [cs.RO] for this version)
Full text under the publisher's open licence · 203 words
Related coverage
Where to Touch, How to Contact: A Hierarchical RL-MPC Framework for Geometry-Aware Sim-to-Real Manipulation
arXiv · Aug 21, 2026
PIRAMID: Progress and Plans
LessWrong · Jul 27, 2026
Gripper-Aware Automatic Dense Packing of Irregular Objects
arXiv · Sep 21, 2026
Vision-Language Grounded Task-Context-Aware Imitation Learning for Robotic Disassembly
arXiv · Sep 17, 2026