Manufacturing Trace Fingerprints for Secure Anomaly Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for determining associations between sensor data and anomalies in substrate processing equipment are inconvenient, often impossible, or lead to erroneous associations, resulting in underperforming products and waste of time, energy, and material, due to security concerns and the need for repeated processes.
Innovation Solution
A method that generates fingerprints based on features extracted from trace data, using fingerprint dimensions that are data agnostic and value-independent, allowing for the performance of corrective actions in manufacturing processes without revealing sensitive information, enabling the sharing of knowledge across facilities while preserving trade secrets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional methods are used to determine associations between sensor data and anomalies, then security concerns are addressed by keeping data private, but the ability to share knowledge and improve manufacturing processes is limited
Solution Approach 1:
The patent extracts only the essential pattern characteristics from sensor data to create fingerprints, removing sensitive raw data while retaining the information needed for anomaly detection. This allows knowledge sharing without exposing proprietary manufacturing details or sensitive sensor readings.
Solution Approach 2:
The patent creates simplified copies (fingerprints) of the original sensor data that capture the essential patterns and associations needed for anomaly detection. These fingerprints can be shared and compared across different facilities without requiring access to the original sensitive data, enabling knowledge transfer while maintaining security.
2Measurement precision
If detailed sensor data is analyzed to identify anomalies, then detection accuracy is improved, but processing time and computational resources increase
Solution Approach 1:
The patent extracts only the critical pattern features from sensor data to create compact fingerprints, eliminating the need to process entire datasets. This extraction approach maintains anomaly detection accuracy by focusing on the most relevant characteristics while dramatically reducing processing time and computational requirements.
Solution Approach 2:
The patent segments the sensor data analysis into two stages: offline fingerprint generation from historical data, and online anomaly detection by comparing current data against stored fingerprints. This segmentation allows complex pattern recognition to be performed once during fingerprint creation, enabling rapid real-time detection without reprocessing all original data.
3Productivity
If manufacturing processes are optimized for specific facilities, then production efficiency is improved, but the ability to transfer knowledge to other facilities is hindered by security concerns
Solution Approach 1:
The patent creates facility-agnostic fingerprints that capture the essential patterns and associations from manufacturing processes. These fingerprints can be copied and applied to other facilities with similar processes, enabling knowledge transfer and optimization without requiring sharing of sensitive facility-specific data or proprietary manufacturing details.
Solution Approach 2:
The patent develops a universal fingerprinting approach that can be applied across multiple facilities and manufacturing processes. The fingerprint methodology serves multiple functions: local optimization at each facility, cross-facility knowledge sharing, anomaly detection, and process improvement, making the system adaptable and versatile across different manufacturing environments.
Data Source
AI summary
A method includes receiving one or more fingerprint dimensions to be used to generate a fingerprint. The method further includes receiving trace data associated with a manufacturing process. The method further includes applying the one or more fingerprint dimensions to the trace data to generate at least one feature. The method further includes generating the fingerprint based on the at least one feature. The method further includes causing, based on the fingerprint, performance of a corrective action associated with one or more manufacturing processes.