AI-FDC: Automated Top Down System for Fab Wide Process Equipment Health Monitoring

We have developed a novel technique to handle Fault Detection and Classification data for Equipment Health Monitoring. While most techniques are very human intensive, this AI-FDC technique allows for less human interaction by taking advantage of recent advancements in Machine Learning. Raw traces ar...

Full description

Saved in:
Bibliographic Details
Published in2020 International Symposium on Semiconductor Manufacturing (ISSM) pp. 1 - 3
Main Authors Burch, Richard, Keleher, Michael, Kunitoshi, Kazuki
Format Conference Proceeding
LanguageEnglish
Published IEEE 15.12.2020
Subjects
Automatic Window Definition
Bootstrapping
Electric breakdown
Fail Mode Prediction
Fault detection
Fault Detection and Classification
Indicators
Machine learning
Manufacturing
Monitoring
Process control
Software
Online AccessGet full text

Cover

More Information
Summary:We have developed a novel technique to handle Fault Detection and Classification data for Equipment Health Monitoring. While most techniques are very human intensive, this AI-FDC technique allows for less human interaction by taking advantage of recent advancements in Machine Learning. Raw traces are automatically broken down into windows with consistent characteristics, relevant statistics are automatically calculated based on window characteristics, and anomalous traces are detected by the system without labels. This system will accelerate root cause diagnosis of Equipment Breakdowns and prevent subsequent breakdowns.
DOI:10.1109/ISSM51728.2020.9377500