Reproduce “Synchronized Online Friction Estimation and Adaptive Grasp Control for Robust Gentle Grasp” 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
Zhenwei Niu, Xiaoyi Chen, Jiayu Hu, Zhaoyang Liu, Tang Jian, Xiaozu Ju
The authors present a robot system that estimates friction from tactile feedback while adjusting grip force. Its controller uses each estimate to adjust the grip, while resulting tactile feedback refines later estimates. In a real-robot cup experiment, friction stabilized within one time step after contact as force rose during pouring; no baseline comparison was reported, and contact detection requires a preset threshold.
Key insights
- Tightly integrating real-time friction estimation with adaptive grasp control in a synchronized, closed-loop system significantly enhances responsiveness and robustness for gentle object manipulation.
- Modeling the friction coefficient as a stochastic variable and using a particle filter allows for rapid, robust online estimation even under dynamic variations, surpassing deterministic or offline-calibrated models.
- A contact factor derived from micro-element resultant forces, coupled with a proportional control law, effectively drives gripper position to maintain grasp stability and gentleness in dynamic environments.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.