Foundation models assist in human–robot collaboration assembly

Abstract Human–robot collaboration (HRC) is a novel manufacturing paradigm designed to fully leverage the advantage of humans and robots, efficiently and flexibly accomplishing customized manufacturing tasks. However, existing HRC systems lack the transfer and generalization capability for environme...

Full description

Saved in:
Bibliographic Details
Main Authors: Yuchen Ji, Zequn Zhang, Dunbing Tang, Yi Zheng, Changchun Liu, Zhen Zhao, Xinghui Li
Format: Article
Language:English
Published: Nature Portfolio 2024-10-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-024-75715-4
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850204493613891584
author Yuchen Ji
Zequn Zhang
Dunbing Tang
Yi Zheng
Changchun Liu
Zhen Zhao
Xinghui Li
author_facet Yuchen Ji
Zequn Zhang
Dunbing Tang
Yi Zheng
Changchun Liu
Zhen Zhao
Xinghui Li
author_sort Yuchen Ji
collection DOAJ
description Abstract Human–robot collaboration (HRC) is a novel manufacturing paradigm designed to fully leverage the advantage of humans and robots, efficiently and flexibly accomplishing customized manufacturing tasks. However, existing HRC systems lack the transfer and generalization capability for environment perception and task reasoning. These limitations manifest in: (1) current methods rely on specialized models to perceive scenes; and need retraining the model when facing unseen objects. (2) current methods only address predefined tasks, and cannot support undefined task reasoning. To avoid these limitations, this paper proposes a novel HRC approach based on Foundation Models (FMs), including Large Language models (LLMs) and Vision Foundation Models (VFMs). Specifically, a LLMs-based task reasoning method is introduced, utilizing prompt learning to transfer LLMs into the domain of HRC tasks, supporting undefined task reasoning. A VFMs-based scene semantic perception method is proposed, integrating various VFMs to achieve scene perception without training. Finally, a FMs-based HRC system is developed, comprising perception, reasoning, and execution modules for more flexible and generalized HRC. The superior performances of FMs in perception and reasoning are demonstrated by extensive experiments. Furthermore, the feasibility and effectiveness of the FMs-based HRC system are validated through an part assembly case involving a satellite component model.
format Article
id doaj-art-0b346d482d044dd28fcb7a2a4cc84fe3
institution OA Journals
issn 2045-2322
language English
publishDate 2024-10-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-0b346d482d044dd28fcb7a2a4cc84fe32025-08-20T02:11:17ZengNature PortfolioScientific Reports2045-23222024-10-0114112110.1038/s41598-024-75715-4Foundation models assist in human–robot collaboration assemblyYuchen Ji0Zequn Zhang1Dunbing Tang2Yi Zheng3Changchun Liu4Zhen Zhao5Xinghui Li6College of Mechanical and Electrical Engineering, Nanjing University of Aeronautics and AstronauticsCollege of Mechanical and Electrical Engineering, Nanjing University of Aeronautics and AstronauticsCollege of Mechanical and Electrical Engineering, Nanjing University of Aeronautics and AstronauticsCollege of Biophotonics, South China Normal UniversityCollege of Mechanical and Electrical Engineering, Nanjing University of Aeronautics and AstronauticsCollege of Mechanical and Electrical Engineering, Nanjing University of Aeronautics and AstronauticsShenzhen International Graduate School, Tsinghua UniversityAbstract Human–robot collaboration (HRC) is a novel manufacturing paradigm designed to fully leverage the advantage of humans and robots, efficiently and flexibly accomplishing customized manufacturing tasks. However, existing HRC systems lack the transfer and generalization capability for environment perception and task reasoning. These limitations manifest in: (1) current methods rely on specialized models to perceive scenes; and need retraining the model when facing unseen objects. (2) current methods only address predefined tasks, and cannot support undefined task reasoning. To avoid these limitations, this paper proposes a novel HRC approach based on Foundation Models (FMs), including Large Language models (LLMs) and Vision Foundation Models (VFMs). Specifically, a LLMs-based task reasoning method is introduced, utilizing prompt learning to transfer LLMs into the domain of HRC tasks, supporting undefined task reasoning. A VFMs-based scene semantic perception method is proposed, integrating various VFMs to achieve scene perception without training. Finally, a FMs-based HRC system is developed, comprising perception, reasoning, and execution modules for more flexible and generalized HRC. The superior performances of FMs in perception and reasoning are demonstrated by extensive experiments. Furthermore, the feasibility and effectiveness of the FMs-based HRC system are validated through an part assembly case involving a satellite component model.https://doi.org/10.1038/s41598-024-75715-4Human–robot collaborationFoundation modelsLarge language modelsVision foundation modelsIntelligent manufacture
spellingShingle Yuchen Ji
Zequn Zhang
Dunbing Tang
Yi Zheng
Changchun Liu
Zhen Zhao
Xinghui Li
Foundation models assist in human–robot collaboration assembly
Scientific Reports
Human–robot collaboration
Foundation models
Large language models
Vision foundation models
Intelligent manufacture
title Foundation models assist in human–robot collaboration assembly
title_full Foundation models assist in human–robot collaboration assembly
title_fullStr Foundation models assist in human–robot collaboration assembly
title_full_unstemmed Foundation models assist in human–robot collaboration assembly
title_short Foundation models assist in human–robot collaboration assembly
title_sort foundation models assist in human robot collaboration assembly
topic Human–robot collaboration
Foundation models
Large language models
Vision foundation models
Intelligent manufacture
url https://doi.org/10.1038/s41598-024-75715-4
work_keys_str_mv AT yuchenji foundationmodelsassistinhumanrobotcollaborationassembly
AT zequnzhang foundationmodelsassistinhumanrobotcollaborationassembly
AT dunbingtang foundationmodelsassistinhumanrobotcollaborationassembly
AT yizheng foundationmodelsassistinhumanrobotcollaborationassembly
AT changchunliu foundationmodelsassistinhumanrobotcollaborationassembly
AT zhenzhao foundationmodelsassistinhumanrobotcollaborationassembly
AT xinghuili foundationmodelsassistinhumanrobotcollaborationassembly