Abstract:Fault analysis (FA) is the process of collecting and analyzing data to determine the cause of a failure. It plays an important role in ensuring the quality in manufacturing process. Traditional FA techniques are time-consuming and labor-intensive, relying heavily on human expertise and the availability of failure inspection equipment. In semiconductor industry, a large amount of FA reports are generated by experts to record the fault descriptions, fault analysis path and fault root causes. With the development of Artificial Intelligence, it is possible to automate the industrial FA process while extracting expert knowledge from the vast FA report data. The goal of this research is to develop a complete expert knowledge extraction pipeline for FA in semiconductor industry based on advanced Natural Language Processing and Machine Learning. Our research aims at automatically predicting the fault root cause based on the fault descriptions. First, the text data from the FA reports are transformed into numerical data using Sentence Transformer embedding. The numerical data are converted into latent spaces using Generalized-Controllable Variational AutoEncoder. Then, the latent spaces are classified by Gaussian Mixture Model. Finally, Association Rules are applied to establish the relationship between the labels in the latent space of the fault descriptions and that of the fault root cause. The proposed algorithm has been evaluated with real data of semiconductor industry collected over three years. The average correctness of the predicted label achieves 97.8%. The method can effectively reduce the time of failure identification and the cost during the inspection stage.

The Study Of Knowledge Representation And Uncertainty Reasoning For Computer Aided Failure Analysis Expert System

Intelligent Fault Diagnosis of CNC Machine Tools Based on Rough Set Theory

Quality problem oriented extension knowledge representation and diagnostic reasoning

Knowledge acquisition development in failure diagnosis analysis as an interactive approach

Study on the Classification and Disposal of Uncertain Knowledge in Intelligent Fault Diagnosis Systems

A Practical Approach To The Integrated Expression Of Shallow And Deep Knowledge In Fault Diagnosis Systems

Building a Knowledge Base of a Fault Diagnostic System for Chemical Processes

A Survey of Knowledge-Based Intelligent Fault Diagnosis Techniques

The Application of Mechanical Failure Diagnosis Expert System for the Improvement of Vibration Fault Monitor and Diagnosis Based on Rough Set Theory

Research on Multiple Faults Diagnosis Method Based on Case Reasoning

Algorithms for Knowledge Acquisition in Incomplete Information System

Fault Diagnosis Uncertainity Reasoning Based on Dual Factors

A Knowledge Graph-Based Failure Information Fusion Method for Enhancing Reliability in Sustainable Systems

Nature language processing (NLP) and association rules (AR)-based knowledge extraction for intelligent fault analysis: a case study in semiconductor industry

A Kind of Representation of Common Knowledge and its Application in Requirements Analysis

A method for root cause diagnosis with picture fuzzy sets based dynamic uncertain causality graph

Elicit the Requirements on Software Dependability: A Knowledge-Based Approach

Enhancing Fault Troubleshooting through Human-Machine Collaboration: A Multi-Stage Reasoning Approach

UNCERTAINTY REASONING IN EXPERT SYSTEM OF BRONZE IDENTIFY

Expert System for Managing Policy of Technological Security in Uncertainty Conditions: Architectural, Algorithmic, and Computing Aspects

Managing engineering analysis knowledge