AI X-Ray CT Anomaly Detection for Non-Destructive Inspection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for distinguishing between manufactured goods that appear identical on the surface, such as counterfeit products, are inadequate and often require costly and destructive inspections to identify differences.
Innovation Solution
Utilizing X-ray computed tomography (CT) scan data to train artificial intelligence (AI) models for anomaly detection, enabling the differentiation between nominal and anomalous goods by analyzing 2D and 3D data derived from CT scans, including renderings, slices, and meshes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional inspection methods (visual, 2D imaging) are used to distinguish manufactured goods, then the inspection process is simple and fast, but the detection precision is insufficient to identify counterfeit or defective goods that appear identical on the surface
Solution Approach 1:
The patent transitions from 2D surface imaging to 3D internal volumetric imaging using X-ray CT scanning. This dimensional change enables detection of internal anomalies, material composition differences, and structural defects that are invisible on the surface, directly resolving the contradiction between detection precision and device complexity by adding the third spatial dimension to the inspection process
Solution Approach 2:
The patent introduces AI/machine learning models as an intermediary between raw X-ray CT data and anomaly detection. The AI model processes complex 3D volumetric data, automatically identifying patterns and anomalies that would be difficult for human inspectors to detect, thereby maintaining operational simplicity while achieving high detection precision through intelligent data analysis
2Measurement precision
If destructive inspection methods are used to verify product authenticity and quality, then the detection precision is high, but the productivity is reduced due to loss of inspected goods
Solution Approach 1:
The patent replaces physical destructive testing methods with non-destructive X-ray CT imaging combined with AI analysis. This substitution allows complete quality verification without consuming or damaging the inspected product, enabling 100% inspection throughput while maintaining high detection accuracy for counterfeits and defects
Solution Approach 2:
The patent creates a digital 3D copy (volumetric image) of the product's internal structure through X-ray CT scanning. This digital twin can be analyzed repeatedly by AI models without affecting the physical product, allowing comprehensive quality verification while preserving the original item for continued use or sale
3Measurement precision
If exhaustive inspection methods are used to detect internal differences, then the detection precision is high, but the inspection time increases significantly
Solution Approach 1:
The patent performs preliminary action by training AI models on extensive datasets of normal and anomalous product features before actual inspection. This pre-training enables the model to rapidly recognize patterns during production inspection, achieving high detection precision without requiring time-consuming manual analysis of each product's 3D data
Solution Approach 2:
The patent enables rapid scanning through the product by utilizing the speed of X-ray imaging, which captures complete 3D volumetric data in seconds. The AI model then processes this data efficiently, skipping through the inspection process quickly while maintaining high detection accuracy for internal anomalies through pattern recognition
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables rapid and accurate identification of anomalies in manufactured goods without destructive testing, improving quality control and safety by distinguishing between authentic and counterfeit items.
Implementation Method 1
X-ray CT is a technique that can image the interior features and structures of such manufactured goods
Data Source
AI summary
Provided herein are methods, apparatuses, computer program products, and systems for anomaly detection using machine learning models. One method can include obtaining X-ray computed tomography (CT) scan data for a scan object of a predetermined object type; producing derived data from the X-ray CT scan data, wherein the derived data reveals at least one interior structure usable for anomaly detection in objects of the predetermined object type; inputting at least the derived data into a machine learning model, which has been trained using at least derived data produced from prior X-ray CT scan data for objects of the predetermined object type; receiving an output from the machine learning model that indicates that at least one anomaly has been detected for the scan object; and providing an output to a physical device based on the at least one anomaly having been detected for the scan object.