Any-ttach — Quick End-effector Swapping for Manipulation Dexterity
// ABSTRACT
Any-ttach is a tool-centric manipulation framework developed in the Duke Robotic Dexterity Lab that leverages mechanical modularity and treats end-effector swapping as a primary mechanism for dexterity. It combines a low-cost, automatic swapping mechanism for a 1-DoF parallel gripper, a handheld device for collecting human demonstrations, and a task planning framework that composes learned, parameterized, and planned skills. My contribution was on the hardware: designing and prototyping the modular mechanical tool interfaces, refining their geometry and actuation mechanisms, and preparing the interface for imitation-learning workflows with graduate researchers. The work is described in an arXiv preprint (2026) on which I am a co-author.
SKILLS
TOOLS
Overview
Any-ttach is a tool-centric manipulation framework developed in Dr. Xianyi Cheng's Robotic Dexterity Lab at Duke University (January 2025 – present). Rather than pursuing dexterity through increasingly complex multifingered hands, Any-ttach leverages mechanical modularity and treats end-effector swapping as a primary mechanism for dexterity. The system combines three components:
- A low-cost, automatic, and easily deployable swapping mechanism for a 1-DoF parallel gripper.
- A handheld device for collecting human demonstrations.
- A task planning framework that composes learned, parameterized, and planned skills for flexible use of diverse tools.
The same end-effector interface supports everyday tools, articulated tools such as scissors, and a low-cost anthropomorphic hand. In two long-horizon tasks, making a sandwich and preparing a cucumber, Any-ttach executes six tool-use subskills.
My Contributions (Hardware)
- Designed and prototyped modular mechanical tool interfaces, enabling rapid and repeatable tool interchangeability.
- Refined the prototype interface geometry and actuation mechanisms, validating tolerances and design robustness.
- Prepared and integrated the interface for imitation-learning workflows in collaboration with graduate researchers.
Publication
W. Ni*, J. Li*, H. Li, W. Pan, C. A. Alessio-Bunnell, X. Cheng. "Any-ttach: Quick End-effector Swapping Enables Manipulation Dexterity with Simplicity." arXiv preprint arXiv:2605.30569, 2026. (*Equal contribution)