Real-World Dexterous Object Manipulation based Deep Reinforcement Learning

TL;DR

This paper presents a hierarchical deep reinforcement learning framework for real-world dexterous object manipulation, enabling a robot to manipulate a cube along a specified trajectory efficiently and effectively in both simulation and real environments.

Contribution

The authors introduce a hierarchical DRL approach that combines high-level contact position selection with low-level trajectory control, improving sample efficiency and adaptability for real-world robotic manipulation.

Findings

Successful transfer from simulation to real robot without fine-tuning

Effective manipulation of a cube along a specified trajectory

Demonstrated robustness and adaptability in real-world experiments

Abstract

Deep reinforcement learning has shown its advantages in real-time decision-making based on the state of the agent. In this stage, we solved the task of using a real robot to manipulate the cube to a given trajectory. The task is broken down into different procedures and we propose a hierarchical structure, the high-level deep reinforcement learning model selects appropriate contact positions and the low-level control module performs the position control under the corresponding trajectory. Our framework reduces the disadvantage of low sample efficiency of deep reinforcement learning and lacking adaptability of traditional robot control methods. Our algorithm is trained in simulation and migrated to reality without fine-tuning. The experimental results show the effectiveness of our method both simulation and reality. Our code and video can be found at https://github.com/42jaylonw/RRC2021ThreeWolves and https://youtu.be/Jr176xsn9wg.