Abstract
We introduce a unified framework for gentle robotic grasping that synergistically couples real-time friction estimation with adaptive grasp control. We propose a new particle filter-based method for real-time estimation of the friction coefficient using vision-based tactile sensors. This estimate is seamlessly integrated into a reactive controller that dynamically modulates grasp force to maintain a stable grip. The two processes operate synchronously in a closed-loop: the controller uses the current best estimate to adjust the force, while new tactile feedback from this action continuously refines the estimation. This creates a highly responsive and robust sensorimotor cycle. The reliability and efficiency of the complete framework are validated through extensive robotic experiments.
Method Overview
We propose a new stochastic modeling approach for friction and introduce an online particle filter-based estimation algorithm that enables rapid, robust, and real-time inference of the friction coefficient during robotic manipulation, even under dynamic physical variations.
We develop a tightly coupled estimation-control architecture, where the particle filter-based estimator is synergistically integrated with a reactive grasp controller, while new tactile feedback from grasp continuously refines the estimation. This allows a closed loop for online friction estimation and continuous force modulation, achieving real-time adaptability and grasp stability.
We validate the proposed framework through extensive real-world robotic experiments, demonstrating significant performance in both grasp reliability and adaptability.
Experiments for Stable and Gentle Grasp
Stable and Gentle Grasp in Acceleration and Deceleration
BibTeX
@article{SofeagcGrasp2026,
title={Synchronized Online Friction Estimation and Adaptive Grasp Control for Robust Gentle Grasp},
author={Niu, Zhenwei and Chen, Xiaoyi and Hu, Jiayu and Liu, Zhaoyang and Ju, Xiaozhu and Tang, Jian},
journal={arxiv},
year={2026},
url={https://ethan-nzw.github.io/SofeagcGrasp/}
}