AutoMEX: Streamlining material extrusion with AI agents powered by large language models and knowledge graphs
Additive manufacturing (AM), particularly material extrusion (MEX), has become a versatile and widely adopted technology with significant applications in the consumer goods and healthcare industries. Despite its affordability, adaptability, and user-friendliness advantages, MEX faces challenges in s...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-03-01
|
| Series: | Materials & Design |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S0264127525000644 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832540447453478912 |
|---|---|
| author | Haolin Fan Junlin Huang Jilong Xu Yifei Zhou Jerry Ying Hsi Fuh Wen Feng Lu Bingbing Li |
| author_facet | Haolin Fan Junlin Huang Jilong Xu Yifei Zhou Jerry Ying Hsi Fuh Wen Feng Lu Bingbing Li |
| author_sort | Haolin Fan |
| collection | DOAJ |
| description | Additive manufacturing (AM), particularly material extrusion (MEX), has become a versatile and widely adopted technology with significant applications in the consumer goods and healthcare industries. Despite its affordability, adaptability, and user-friendliness advantages, MEX faces challenges in scaling for mass production due to limited process automation and fragmented domain knowledge, leaving gaps in end-to-end workflow integration. We propose AutoMEX, an innovative framework that integrates large language models (LLMs) as artificial intelligence (AI) agents to automate the MEX process to address these limitations. AutoMEX utilizes a knowledge graph (KG) derived from the scientific literature to enable LLMs to provide expert recommendations on material selection, process parameters, and design considerations, thereby improving accessibility and efficiency. With minimal human intervention, the framework encompasses a complete workflow, including CAD model generation, printing parameter recommendation, slicing, and machine operation. Experimental validation demonstrated a query acceptance rate of 94.6% for the recommendation system and up to a 9.6% improvement in print strength when employing the recommended parameters. These results highlight enhanced quality, autonomy, and customization of AM outputs, making AutoMEX suitable for batch production and streamlined manufacturing processes. While the framework shows promising potential, challenges such as the computational demands of advanced LLMs and the need for continual updates to the KG remain areas for future work. Overall, AutoMEX offers a pathway toward broader adoption and scalability of MEX technology, advancing the field of AM through enhanced automation and efficiency. |
| format | Article |
| id | doaj-art-39c76ab5c229422493d772777da55b85 |
| institution | Kabale University |
| issn | 0264-1275 |
| language | English |
| publishDate | 2025-03-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Materials & Design |
| spelling | doaj-art-39c76ab5c229422493d772777da55b852025-02-05T04:31:05ZengElsevierMaterials & Design0264-12752025-03-01251113644AutoMEX: Streamlining material extrusion with AI agents powered by large language models and knowledge graphsHaolin Fan0Junlin Huang1Jilong Xu2Yifei Zhou3Jerry Ying Hsi Fuh4Wen Feng Lu5Bingbing Li6Department of Mechanical Engineering, National University of Singapore, 9 Engineering Drive 1, 117575, Singapore; National University of Singapore Chongqing Research Institute, Chongqing, 401123, ChinaDepartment of Mechanical Engineering, National University of Singapore, 9 Engineering Drive 1, 117575, SingaporeDepartment of Mechanical Engineering, National University of Singapore, 9 Engineering Drive 1, 117575, SingaporeDepartment of Mechanical Engineering, National University of Singapore, 9 Engineering Drive 1, 117575, SingaporeDepartment of Mechanical Engineering, National University of Singapore, 9 Engineering Drive 1, 117575, Singapore; National University of Singapore Chongqing Research Institute, Chongqing, 401123, ChinaDepartment of Mechanical Engineering, National University of Singapore, 9 Engineering Drive 1, 117575, Singapore; National University of Singapore Chongqing Research Institute, Chongqing, 401123, ChinaDepartment of Mechanical Engineering, National University of Singapore, 9 Engineering Drive 1, 117575, Singapore; Department of Manufacturing Systems Engineering and Management, California State University Northridge, 18111 Nordhoff St, Northridge, CA 91330, USA; Corresponding author at: Department of Manufacturing Systems Engineering and Management, California State University Northridge, 18111 Nordhoff St, Northridge, CA 91330, USA.Additive manufacturing (AM), particularly material extrusion (MEX), has become a versatile and widely adopted technology with significant applications in the consumer goods and healthcare industries. Despite its affordability, adaptability, and user-friendliness advantages, MEX faces challenges in scaling for mass production due to limited process automation and fragmented domain knowledge, leaving gaps in end-to-end workflow integration. We propose AutoMEX, an innovative framework that integrates large language models (LLMs) as artificial intelligence (AI) agents to automate the MEX process to address these limitations. AutoMEX utilizes a knowledge graph (KG) derived from the scientific literature to enable LLMs to provide expert recommendations on material selection, process parameters, and design considerations, thereby improving accessibility and efficiency. With minimal human intervention, the framework encompasses a complete workflow, including CAD model generation, printing parameter recommendation, slicing, and machine operation. Experimental validation demonstrated a query acceptance rate of 94.6% for the recommendation system and up to a 9.6% improvement in print strength when employing the recommended parameters. These results highlight enhanced quality, autonomy, and customization of AM outputs, making AutoMEX suitable for batch production and streamlined manufacturing processes. While the framework shows promising potential, challenges such as the computational demands of advanced LLMs and the need for continual updates to the KG remain areas for future work. Overall, AutoMEX offers a pathway toward broader adoption and scalability of MEX technology, advancing the field of AM through enhanced automation and efficiency.http://www.sciencedirect.com/science/article/pii/S0264127525000644Material extrusionLarge language modelsArtificial intelligence agentsKnowledge graphsAutonomous additive manufacturing |
| spellingShingle | Haolin Fan Junlin Huang Jilong Xu Yifei Zhou Jerry Ying Hsi Fuh Wen Feng Lu Bingbing Li AutoMEX: Streamlining material extrusion with AI agents powered by large language models and knowledge graphs Materials & Design Material extrusion Large language models Artificial intelligence agents Knowledge graphs Autonomous additive manufacturing |
| title | AutoMEX: Streamlining material extrusion with AI agents powered by large language models and knowledge graphs |
| title_full | AutoMEX: Streamlining material extrusion with AI agents powered by large language models and knowledge graphs |
| title_fullStr | AutoMEX: Streamlining material extrusion with AI agents powered by large language models and knowledge graphs |
| title_full_unstemmed | AutoMEX: Streamlining material extrusion with AI agents powered by large language models and knowledge graphs |
| title_short | AutoMEX: Streamlining material extrusion with AI agents powered by large language models and knowledge graphs |
| title_sort | automex streamlining material extrusion with ai agents powered by large language models and knowledge graphs |
| topic | Material extrusion Large language models Artificial intelligence agents Knowledge graphs Autonomous additive manufacturing |
| url | http://www.sciencedirect.com/science/article/pii/S0264127525000644 |
| work_keys_str_mv | AT haolinfan automexstreamliningmaterialextrusionwithaiagentspoweredbylargelanguagemodelsandknowledgegraphs AT junlinhuang automexstreamliningmaterialextrusionwithaiagentspoweredbylargelanguagemodelsandknowledgegraphs AT jilongxu automexstreamliningmaterialextrusionwithaiagentspoweredbylargelanguagemodelsandknowledgegraphs AT yifeizhou automexstreamliningmaterialextrusionwithaiagentspoweredbylargelanguagemodelsandknowledgegraphs AT jerryyinghsifuh automexstreamliningmaterialextrusionwithaiagentspoweredbylargelanguagemodelsandknowledgegraphs AT wenfenglu automexstreamliningmaterialextrusionwithaiagentspoweredbylargelanguagemodelsandknowledgegraphs AT bingbingli automexstreamliningmaterialextrusionwithaiagentspoweredbylargelanguagemodelsandknowledgegraphs |