Abstract

Commercially accessible dexterous robot hands are increasingly prevalent, but many remain difficult to use as scientific instruments. For example, the Inspire RH56DFX hand exposes only uncalibrated proprioceptive information and shows unreliable contact behavior at high speed (up to +1618% force limit overshoot). Furthermore, its underactuated, coupled finger linkages make antipodal grasps non-trivial. We contribute three improvements to the Inspire RH56DFX to transform it from a black-box device to a research tool: (1) hardware characterization (force calibration, latency, and overshoot), (2) a sim2real validated MuJoCo model for analytical width-to-grasp planning, and (3) a hybrid, closed-loop speed-force grasp controller. We validate these components on peg-in-hole insertion, achieving 65% success and outperforming a wrist-force-only baseline of 10% and on 300 grasps across 15 physically diverse objects, achieving 87% success and outperforming plan-free grasps and learned grasps. Our approach is modular, designed for compatibility with external object detectors and vision-language models for width & force estimation and high-level planning, and provides an interpretable, immediately deployable, and open-source (upon publishing) interface for dexterous manipulation with the Inspire RH56DFX hand.

Overview Video

Grasping Experiments

YCB and YCB-like Objects

Big Screwdriver

Bottle

Can

Charger

Metal Cup

Mustard

Orange

Pen

Small Screwdriver

Sugar Box

Delicate Objects

Egg

Nut

Paper Cup

Raspberry

Strawberry

Sim-to-Real Validation

Sim

Real

Sim (with UR5)

Real (with UR5)

Grasp Quality Visualization

Peg-in-Hole Experiments

FAQs

Code release status

The latest development branch is available on GitHub at the h12 branch of rh56_controller. This branch currently has the most flexible installation requirements; the main branch is more stable but may be harder to install.

Development on the Inspire RH56DFX hands is currently paused. Since late March 2026, we have been unable to read from the RX or TX pins on either hand, and we have not yet received sufficient vendor support to resolve the issue. Because of this hardware and support limitation, the official code release has been delayed.

At present, ROS2 support is available only for basic hand controls. The antipodal grasping and other higher-level control functionality remain implemented in native Python/Tkinter rather than ROS2.

BibTeX

@misc{tan2026characterizationanalyticalplanninghybrid,
            title={Characterization, Analytical Planning, and Hybrid Force Control for the Inspire RH56DFX Hand},
            author={Xuan Tan and William Xie and Nikolaus Correll},
            year={2026},
            eprint={2603.08988},
            archivePrefix={arXiv},
            primaryClass={cs.RO},
            url={https://arxiv.org/abs/2603.08988},
        }
        

Acknowledgements

The website template is adapted from Language to Rewards and ProgPrompt.