Robotic LEGO Assembly and Disassembly from Human Demonstration

TL;DR

This paper presents a framework for robotic LEGO assembly and disassembly learned from human demonstrations, utilizing a digital twin for verification and an end-effector tool for manipulation, successfully deployed on an industrial robot.

Contribution

It introduces a novel system combining learning from demonstration, digital twin verification, and specialized end-effector design for LEGO manipulation on industrial robots.

Findings

Effective learning of assembly/disassembly sequences from human demos

Successful deployment on FANUC industrial robot

Verification of task correctness via digital twin

Abstract

This paper studies automatic prototyping using LEGO. To satisfy individual needs and self-sustainability, this paper presents a framework that learns the assembly and disassembly sequences from human demonstrations. In addition, a digital twin is developed to verify the correctness of robot learning before deploying to the real world. Moreover, an end-effector tool (EOT) is designed, which allows large industrial robots to easily manipulate LEGO bricks. The proposed system is deployed to a FANUC LR-mate 200id/7L robot. Experiments demonstrate that the proposed system can effectively learn the assembly and disassembly tasks from human demonstrations. And the learned tasks are realized by the FANUC robot.