Underactuated Dexterous Robotic Grasping With Reconfigurable Passive Joints

2025-01-01

IEEE Robotics and Automation Letters

AI Summary

To address the challenge of simultaneously achieving dexterity and lightweight design in underactuated robotic hands, this paper introduces the Reconfigurable Passive joint (RP-joint)—a driveless, lightweight, and compact multi-state locking mechanism that reconfigures via tendon tension and is integrated into a three-finger underactuated hand. We further propose a one-shot imitation learning-based grasp planning method that synergistically incorporates kinesthetic feedback and contact optimization to enable automatic RP-joint configuration and dexterous manipulation. Key contributions include: (i) the first multi-state reconfigurable design for a driveless passive joint, and (ii) the first end-to-end grasp planning framework for underactuated hands enabled by a single demonstration. Evaluated on the IKEA dataset (42 objects) and the YCB dataset (370 grasps), our approach achieves success rates of 80% and 87%, respectively—demonstrating substantial improvements in robustness and generalization capability.

Technology Category

• Intelligent Robots: Manipulation
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Planning, Routing, and Scheduling: Replanning and Plan Repair

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

We introduce a novel reconfigurable passive joint (RP-joint), which has been implemented and tested on an underactuated three-finger robotic gripper. RP-joint has no actuation, but instead it is lightweight and compact. It can be easily reconfigured by applying external forces and locked to perform complex dexterous manipulation tasks, but only after tension is applied to the connected tendon. Additionally, we present an approach that allows learning dexterous grasps from single examples with underactuated grippers and automatically configures the RP-joints for dexterous manipulation. This is enhanced by integrating kinaesthetic contact optimization, which improves grasp performance even further. The proposed RP-joint gripper and grasp planner have been tested on over 370 grasps executed on 42 IKEA objects and on the YCB object dataset, achieving grasping success rates of 80% and 87%, on IKEA and YCB, respectively.

Problem

• Deformable Joints
• Dexterous Grasping
• Robot Gripper Design

Innovation

• Deformable Joints
• Learning Algorithm
• Optimization Technique