Wafer Path Event Sequencing for Faster Root Cause Isolation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing semiconductor production systems face challenges in efficiently identifying the root cause of wafer quality excursions, as they rely on extensive data analysis and lack effective tools to narrow down possible causes.
Innovation Solution
An event sequence-driven approach is adopted, where the transitions between processing steps in a semiconductor fabrication facility are modeled to compute probabilities for each transition, identifying critical pairs that distinguish between good and bad wafers. This information is aggregated and cross-validated to confirm the accuracy of the model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If extensive data analysis is used to identify root causes of wafer quality excursions, then measurement precision is improved, but loss of time increases
Solution Approach 1:
The patent segments the extensive production data into discrete event sequences representing specific transitions between processing steps. By dividing the data into meaningful temporal segments (events), the system can analyze root causes more efficiently without examining every data point, thus reducing time loss while maintaining identification accuracy.
Solution Approach 2:
The patent performs preliminary action by pre-defining and cataloging event sequences and transitions before actual root cause analysis is needed. The system establishes a framework of expected transitions and anomalies in advance, allowing for rapid comparison against actual production data when quality excursions occur, thereby reducing the time required for root cause identification.
2Measurement precision
If extensive data from production runs is analyzed, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent extracts only the essential and relevant features from extensive production data - specifically event sequences and transitions between processing steps. By taking out only the critical information needed for anomaly detection rather than analyzing all raw data, the system maintains high measurement precision while reducing the complexity of the classification system.
Solution Approach 2:
The patent applies local quality by focusing analysis on specific critical transitions and event sequences rather than treating all data uniformly. The system identifies and emphasizes locally important patterns (specific transitions that lead to anomalies) rather than requiring complex global analysis of all production data, thereby simplifying the overall system while maintaining detection accuracy.
Data Source
AI summary
Wafer quality is determined by modeling equipment history as a sequence of events, then evaluating anomalous results for individual events. Identifying an event that generates bad wafers narrows the list of possible root causes.