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 AllegroHand

From Leeroopedia
Knowledge Sources
Domains Robotics, Simulation, Dexterous_Manipulation
Last Updated 2026-02-15 08:00 GMT

Overview

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

Description

The AllegroHandRight is a 16-DOF dexterous hand with four fingers: thumb, index, middle, and ring. Each finger has 4 joints (joint_0.0 through joint_15.0). The hand is fixed in space (no floating base). Tip links have high-friction materials (static=2.0, dynamic=1.0) for stable contact. The robot provides proprioceptive observations including palm pose and tip poses for all four fingertips. Joint stiffness is 4e2, damping is 1e1, and force limit is 5e1. The file also defines AllegroHandLeft which inherits from AllegroHandRight with a different URDF.

uid: allegro_hand_right / allegro_hand_left

URDF path: {PACKAGE_ASSET_DIR}/robots/allegro/allegro_hand_right_glb.urdf (right) / allegro_hand_left.urdf (left)

Supported control modes: pd_joint_delta_pos, pd_joint_pos, pd_joint_target_delta_pos

Usage

Use AllegroHand for dexterous in-hand manipulation tasks such as object rotation, reorientation, and finger gaiting. The 16-DOF design allows complex finger coordination. Suitable for benchmarking dexterous manipulation policies where a fixed hand must manipulate objects using only finger motions.

Code Reference

Source Location

Signature

@register_agent()
class AllegroHandRight(BaseAgent):
    uid = "allegro_hand_right"
    urdf_path = f"{PACKAGE_ASSET_DIR}/robots/allegro/allegro_hand_right_glb.urdf"
    joint_names = ["joint_0.0", ..., "joint_15.0"]  # 16 joints
    joint_stiffness = 4e2
    joint_damping = 1e1
    joint_force_limit = 5e1
    tip_link_names = ["link_15.0_tip", "link_3.0_tip", "link_7.0_tip", "link_11.0_tip"]
    keyframes = dict(palm_side=Keyframe(...), palm_up=Keyframe(...))

@register_agent()
class AllegroHandLeft(AllegroHandRight):
    uid = "allegro_hand_left"

Import

from mani_skill.agents.robots.allegro_hand.allegro import AllegroHandRight, AllegroHandLeft

I/O Contract

Inputs

Name Type Required Description
(inherited from BaseAgent)

Outputs

Name Type Description
robot agent AllegroHandRight Configured 16-DOF dexterous hand with PD joint controllers and proprioceptive tip/palm pose feedback

Usage Examples

Creating Environment with Robot

import gymnasium as gym
import mani_skill.envs

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

Related Pages

Page Connections

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