Jump to content

Connect SuperML | Leeroopedia MCP: Equip your AI agents with best practices, code verification, and debugging knowledge. Powered by Leeroo — building Organizational Superintelligence. Contact us at founders@leeroo.com.

Implementation:Haosulab ManiSkill FloatingInspireHand

From Leeroopedia
Revision as of 12:53, 16 February 2026 by Admin (talk | contribs) (Auto-imported from implementations/Haosulab_ManiSkill_FloatingInspireHand.md)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Knowledge Sources
Domains Robotics, Simulation, Dexterous_Manipulation
Last Updated 2026-02-15 08:00 GMT

Overview

Concrete tool for simulating the Floating Inspire Hand (right and left variants) dexterous robot hand in ManiSkill environments.

Description

The FloatingInspireHand defines two classes -- FloatingInspireHandRight (uid: floating_inspire_hand_right) and FloatingInspireHandLeft (uid: floating_inspire_hand_left). Each is a 5-finger dexterous hand mounted on a 6-DOF floating base (3 translational + 3 rotational root joints). The hand has 2 wrist joints (pitch, yaw), 6 actively controlled finger joints (thumb CMC yaw/pitch, index/middle/ring/pinky MCP), and 6 passive mimic joints (thumb MCP/IP, index/middle/ring/pinky PIP). Self-collisions are disabled for simulation efficiency. Finger link materials use high friction (static=2.0, dynamic=2.0). The robot has 20 total DOF (6 root + 2 wrist + 6 finger active + 6 passive).

uid: floating_inspire_hand_right / floating_inspire_hand_left

URDF path: {PACKAGE_ASSET_DIR}/robots/inspire_hand/RH56DFX-2LR/urdf/inspire_hand_right_floating.urdf (right) / inspire_hand_left_floating.urdf (left)

Supported control modes: pd_joint_pos, pd_joint_delta_pos

Usage

Use the FloatingInspireHand for dexterous in-hand manipulation tasks such as rotating objects, reorienting items, and precise finger-level grasping. The floating base allows free positioning in 3D space. Suitable for benchmarking dexterous manipulation policies where the hand needs to manipulate objects without a fixed arm attachment.

Code Reference

Source Location

Signature

@register_agent()
class FloatingInspireHandRight(BaseAgent):
    uid = "floating_inspire_hand_right"
    urdf_path = f"{PACKAGE_ASSET_DIR}/robots/inspire_hand/RH56DFX-2LR/urdf/inspire_hand_right_floating.urdf"
    disable_self_collisions = True
    root_joint_names = [
        "root_x_axis_joint", "root_y_axis_joint", "root_z_axis_joint",
        "root_x_rot_joint", "root_y_rot_joint", "root_z_rot_joint",
    ]
    keyframes = dict(
        palm_side=Keyframe(...),
        palm_up=Keyframe(...),
    )

@register_agent()
class FloatingInspireHandLeft(BaseAgent):
    uid = "floating_inspire_hand_left"

Import

from mani_skill.agents.robots.inspire_hand.floating_inspire_hand import FloatingInspireHandRight, FloatingInspireHandLeft

I/O Contract

Inputs

Name Type Required Description
(inherited from BaseAgent)

Outputs

Name Type Description
robot agent FloatingInspireHandRight / FloatingInspireHandLeft Configured floating dexterous hand with root, wrist, finger, and passive controllers

Usage Examples

Creating Environment with Robot

import gymnasium as gym
import mani_skill.envs

env = gym.make("RotateSingleObjectInHand-v1", robot_uids="floating_inspire_hand_right")
obs, info = env.reset()

Related Pages

Page Connections

Double-click a node to navigate. Hold to expand connections.
Principle
Implementation
Heuristic
Environment