Anomaly Detection System Using Reference Information
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current anomaly detection systems in microelectronics and IT systems face challenges in accurately identifying deviations from normal behavior due to contaminated baselines, which can lead to false negatives, and manual analysis methods are slow and costly for supply chain risk management, especially in the absence of golden reference items.
Innovation Solution
A method and system that automate the analysis of anomalies by using reference information to preprocess and analyze item data, employing techniques like Optical Character Recognition, Natural Language Processing, and computer vision to determine confidence in identification, trust, and compliance levels, and generate output data indicating whether items are expected or abnormal.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual analysis methods are used for supply chain risk management, then analysis accuracy can be maintained through human expertise, but the process becomes slow and costly
Solution Approach 1:
The patent replaces manual mechanical analysis with an automated computer-based system that uses machine learning models, optical character recognition, and image processing to detect anomalies in supply chain items, thereby maintaining accuracy while dramatically improving analysis speed and reducing costs
Solution Approach 2:
The system enables self-service anomaly detection by automatically comparing item characteristics against learned patterns and reference data without requiring human analysts to manually examine each item, allowing the system to serve itself in identifying deviations and generating alerts
2Productivity
If anomaly detection systems are trained on existing baseline data, then the system can quickly deploy, but the baseline may already contain hidden anomalies leading to false negatives
Solution Approach 1:
The patent applies preliminary cleaning and validation actions to the baseline training data before model training, using automated methods to identify and remove potentially contaminated samples, thereby preventing hidden anomalies from being learned as normal patterns while maintaining rapid deployment capabilities
Solution Approach 2:
The system introduces an intermediary validation layer between raw baseline data and the training process, using reference information and cross-verification techniques to filter out corrupted data points before they can compromise the anomaly detection model's reliability
3Measurement precision
If golden reference items are available for comparison, then anomaly detection accuracy improves, but such reference items are often unavailable in commercial supply chains
Solution Approach 1:
The patent creates synthetic reference copies and virtual golden items through data augmentation and generation techniques, allowing the system to establish comparison benchmarks without requiring physical reference items, thereby enabling accurate anomaly detection in commercial supply chains where golden references are unavailable
Solution Approach 2:
The system develops universal anomaly detection capabilities that can operate across different supply chain contexts with or without reference items, using multiple detection modes including reference-based comparison when available and reference-free pattern recognition when unavailable, making the system adaptable to both controlled and commercial environments
Data Source
AI summary
Method and system for automating analyzing anomalies of an item for which no golden reference item is available, by using reference information, wherein the golden reference item is a known non-abnormal instance of an analyzed assembly, includes loading from a data storage, a memory, or via a communication, or user entry, item information of an analyzed item to be analyzed to be either expected or abnormal; loading reference information about the analyzed item; preprocessing both of the item information and the reference information to facilitate analysis; analyzing the item information and the reference information to determine a result that indicates whether elements of the item are confirmed by the reference information to be expected or abnormal; generating an output data with the result; and storing the output data pertaining to the result in a memory.