FBS Model-based Maintenance Record Accumulation for Failure-Cause Inference in Manufacturing Systems

TL;DR

The paper addresses failure-cause inference in manufacturing by requiring explicit target-system knowledge and long causal failure chains. It introduces a Diagnostic Knowledge Ontology that encodes $FBS$ (Function-Behavior-Structure) alongside a maintenance-record accumulation workflow, linking design-phase knowledge with maintenance data. Experiments on a LEGO car assembly line show that incorporating the $FBS$ model into maintenance records improves inference accuracy, especially when related cases are scarce or vocabulary varies. This work lays a foundation for design-maintenance knowledge sharing and suggests future enhancements such as node-level embeddings and shop-floor interfaces for broader deployment.

Abstract

In manufacturing systems, identifying the causes of failures is crucial for maintaining and improving production efficiency. In knowledge-based failure-cause inference, it is important that the knowledge base (1) explicitly structures knowledge about the target system and about failures, and (2) contains sufficiently long causal chains of failures. In this study, we constructed Diagnostic Knowledge Ontology and proposed a Function-Behavior-Structure (FBS) model-based maintenance-record accumulation method based on it. Failure-cause inference using the maintenance records accumulated by the proposed method showed better agreement with the set of candidate causes enumerated by experts, especially in difficult cases where the number of related cases is small and the vocabulary used differs. In the future, it will be necessary to develop inference methods tailored to these maintenance records, build a user interface, and carry out validation on larger and more diverse systems. Additionally, this approach leverages the understanding and knowledge of the target in the design phase to support knowledge accumulation and problem solving during the maintenance phase, and it is expected to become a foundation for knowledge sharing across the entire engineering chain in the future.

Figures (7)

• Figure 1: Overview of the proposed framework
• Figure 2: Diagnosis knowledge ontology
• Figure 3: At the system commissioning stage, the FBS model of the target system is constructed. After the system begins operation, descriptions of failures are accumulated in a manner that links them to this model.
• Figure 4: Overview of inference in the experiment and comparison of chunking methods between the proposed method and the baseline. In the proposed method, chunks are limited to the hierarchical level where the assumed failure occurred and below, and they include both failure and system information. In contrast, the baseline method does not consider hierarchy and treats a whole causal chain of failures as a single chunk.
• Figure 5: The manufacturing system used in this experiment is a part of a LEGO-car assembly line. It consists of six processes: roof assembly, roof press-fitting, roof-height inspection, image inspection, performance inspection, and product collection.