Mana: Dexterous Manipulation of Articulated Tools via an Animation-Inspired Sim-to-Real Framework

Forum topic · 小凯 · 2026-06-15

Summary

Mana (Manipulation Animator) is a general sim-to-real framework from Zhao-Heng Yin, Guanya Shi, Pieter Abbeel, and C. Karen Liu (arXiv:2606.13677) that reinterprets dexterous manipulation of articulated tools as an animation problem. Articulated tool manipulation remains a major challenge in dexterous robotics because it requires coordinating internal degrees of freedom and contact-rich interactions; prior work has focused mostly on rigid objects, leaving functional grasping and manipulation of articulated tools underexplored. Inspired by computer animation, Mana uses a coarse-to-fine pipeline that converts procedurally generated grasp keyframes into manipulation trajectories through motion planning and reinforcement learning. Its data generation is largely automatic: specifying functional affordances per tool takes under one minute with only a few mouse clicks. Across four articulated tools of different scales and joint types, Mana achieved zero-shot sim-to-real transfer of grasping and in-hand manipulation, demonstrating a scalable approach to dexterous articulated tool use.

Overview

Research area: Computer Vision / Robotics Authors: Zhao-Heng Yin, Guanya Shi, Pieter Abbeel, C. Karen Liu Published: 2026-06-11 arXiv: 2606.13677

Abstract

Articulated tool manipulation remains a major challenge in dexterous robotics due to the need to coordinate internal degrees of freedom and contact-rich interactions. While prior work has largely focused on rigid objects, articulated tool use remains underexplored because of its physical complexity and the difficulty of learning functional grasping and manipulation policies. We present Mana (Manipulation Animator), a general sim-to-real framework that reinterprets dexterous manipulation as an animation problem. Inspired by computer animation, Mana employs a coarse-to-fine pipeline that transforms procedurally-generated grasp keyframes into manipulation trajectories through motion planning and reinforcement learning. The data generation process is largely automatic, requiring only a few mouse clicks to specify functional affordances (less than 1 minute per tool). Across four articulated tools spanning different scales and joint types, Mana achieves zero-shot sim-to-real transfer of grasping and in-hand manipulation, demonstrating a scalable approach to dexterous articulated tool use.

Key points

Problem: Dexterous manipulation of articulated tools — coordinating internal degrees of freedom and contact-rich interactions — remains underexplored compared to rigid object manipulation.
Approach: Mana reinterprets dexterous manipulation as an animation problem, inspired by computer animation techniques.
Pipeline: A coarse-to-fine process converts procedurally generated grasp keyframes into manipulation trajectories using motion planning and reinforcement learning.
Automation: Data generation requires only a few mouse clicks to specify functional affordances, taking under 1 minute per tool.
Results: Zero-shot sim-to-real transfer of grasping and in-hand manipulation across four articulated tools of varying scales and joint types.