Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Fault Cause Identification across Manufacturing Lines through Ontology-Guided and Process-Aware FMEA Graph Learning with LLMs

Sho Okazaki, Kohei Kaminishi, Takuma Fujiu, Yusheng Wang, Jun Ota

TL;DR

The paper addresses fault-cause identification across heterogeneous manufacturing lines by transforming FMEA worksheets into a unified knowledge graph using ontology-guided LLM extraction. It then applies a process-aware Relational Graph Convolutional Network with an order-aware scoring function to enforce both semantic consistency and sequential process constraints, enabling transferable fault-cause prediction via link prediction. In a case study on automotive pressure sensor assembly, the approach achieves a top F1@20 of 0.523, outperforming a retrieval-based RAG baseline (0.267) and a standard RGCN (0.400), with ablations confirming the value of both domain conceptualization and process awareness. The framework demonstrates improved cross-line diagnostic reliability and offers a practical path for integrating domain-adaptive LLMs into smart manufacturing for more robust fault diagnosis.

Abstract

Fault cause identification in automated manufacturing lines is challenging due to the system's complexity, frequent reconfigurations, and the limited reusability of existing Failure Mode and Effects Analysis (FMEA) knowledge. Although FMEA worksheets contain valuable expert insights, their reuse across heterogeneous lines is hindered by natural language variability, inconsistent terminology, and process differences. To address these limitations, this study proposes a process-aware framework that enhances FMEA reusability by combining manufacturing-domain conceptualization with graph neural network (GNN) reasoning. First, FMEA worksheets from multiple manufacturing lines are transformed into a unified knowledge graph through ontology-guided large language model (LLM) extraction, capturing domain concepts such as actions, states, components, and parameters. Second, a Relational Graph Convolutional Network (RGCN) with the process-aware scoring function learns embeddings that respect both semantic relationships and sequential process flows. Finally, link prediction is employed to infer and rank candidate fault causes consistent with the target line's process flow. A case study on automotive pressure sensor assembly lines demonstrates that the proposed method outperforms a state-of-the-art retrieval-augmented generation (RAG) baseline (F1@20 = 0.267) and an RGCN approach (0.400), achieving the best performance (0.523) in fault cause identification. Ablation studies confirm the contributions of both LLM-driven domain conceptualization and process-aware learning. These results indicate that the proposed framework significantly improves the transferability of FMEA knowledge across heterogeneous lines, thereby supporting operators in diagnosing failures more reliably and paving the way for future domain-adaptive LLM applications in smart manufacturing.

Fault Cause Identification across Manufacturing Lines through Ontology-Guided and Process-Aware FMEA Graph Learning with LLMs

TL;DR

The paper addresses fault-cause identification across heterogeneous manufacturing lines by transforming FMEA worksheets into a unified knowledge graph using ontology-guided LLM extraction. It then applies a process-aware Relational Graph Convolutional Network with an order-aware scoring function to enforce both semantic consistency and sequential process constraints, enabling transferable fault-cause prediction via link prediction. In a case study on automotive pressure sensor assembly, the approach achieves a top F1@20 of 0.523, outperforming a retrieval-based RAG baseline (0.267) and a standard RGCN (0.400), with ablations confirming the value of both domain conceptualization and process awareness. The framework demonstrates improved cross-line diagnostic reliability and offers a practical path for integrating domain-adaptive LLMs into smart manufacturing for more robust fault diagnosis.

Abstract

Fault cause identification in automated manufacturing lines is challenging due to the system's complexity, frequent reconfigurations, and the limited reusability of existing Failure Mode and Effects Analysis (FMEA) knowledge. Although FMEA worksheets contain valuable expert insights, their reuse across heterogeneous lines is hindered by natural language variability, inconsistent terminology, and process differences. To address these limitations, this study proposes a process-aware framework that enhances FMEA reusability by combining manufacturing-domain conceptualization with graph neural network (GNN) reasoning. First, FMEA worksheets from multiple manufacturing lines are transformed into a unified knowledge graph through ontology-guided large language model (LLM) extraction, capturing domain concepts such as actions, states, components, and parameters. Second, a Relational Graph Convolutional Network (RGCN) with the process-aware scoring function learns embeddings that respect both semantic relationships and sequential process flows. Finally, link prediction is employed to infer and rank candidate fault causes consistent with the target line's process flow. A case study on automotive pressure sensor assembly lines demonstrates that the proposed method outperforms a state-of-the-art retrieval-augmented generation (RAG) baseline (F1@20 = 0.267) and an RGCN approach (0.400), achieving the best performance (0.523) in fault cause identification. Ablation studies confirm the contributions of both LLM-driven domain conceptualization and process-aware learning. These results indicate that the proposed framework significantly improves the transferability of FMEA knowledge across heterogeneous lines, thereby supporting operators in diagnosing failures more reliably and paving the way for future domain-adaptive LLM applications in smart manufacturing.
Paper Structure (18 sections, 7 equations, 9 figures, 3 tables, 1 algorithm)

This paper contains 18 sections, 7 equations, 9 figures, 3 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Literature Review
  3. Ontology-Based Methods
  4. Knowledge Graph and Machine Learning-Based Methods
  5. LLM- and RAG-Based Methods
  6. Research Gaps and Contributions
  7. Methodology
  8. Knowledge Graph Construction
  9. Ontology Definition
  10. Ontology-Guided LLM-Driven Entity Extraction
  11. Process-Aware Graph Learning
  12. Fault Cause Prediction
  13. Case Study
  14. Data Description
  15. Experimental Setup
  16. ...and 3 more sections

Figures (9)

  • Figure 1: Overall framework of the proposed method.
  • Figure 2: Ontology description of manufacturing and FMEA concepts.
  • Figure 3: Overview of the constructed knowledge graph. Instances from FMEA ontology and manufacturing ontology are linked through defined relations.
  • Figure 4: Architecture overview of process-aware graph learning.
  • Figure 5: Inference Process. Descriptions of new failures and their process contexts are input to the trained RGCN model to predict and rank candidate causes.
  • ...and 4 more figures