Automated Engine Blade Inspection via Multi-Modal Data Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Human inspectors often miss defects in images or videos from borescopes due to repetitive tasks and physical or mental fatigue, leading to potential customer dissatisfaction and increased costs in engine inspection and other applications.
Innovation Solution
An automated method using multi-modal analysis that combines image data with 3D CAD models and manufacturing records, employing data fusion and Bayesian analysis to identify and prioritize potential defects in engine blades, minimizing human error and improving detection reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If human inspectors manually examine images from borescopes, then flexibility and adaptability are maintained, but inspection reliability deteriorates due to fatigue and repetitive tasks
Solution Approach 1:
The system enables self-service inspection by automatically comparing blade images against CAD models and manufacturing records without requiring human inspectors. The automated defect detection system performs inspections independently, eliminating fatigue-related errors while maintaining consistent reliability across all inspections.
Solution Approach 2:
The patent replaces the mechanical human inspection process with an automated computational system. Image data from borescopes is automatically processed through multi-modal analysis, substituting human visual inspection with algorithmic comparison against reference data, thereby eliminating variability and fatigue associated with manual inspection.
2Measurement precision
If multiple modes of data are analyzed, then defect detection accuracy improves, but system complexity increases
Solution Approach 1:
The system merges multiple data modes (image data, CAD models, manufacturing records) into a unified analysis framework. By combining these diverse data sources through multi-modal analysis, the system achieves comprehensive defect detection that leverages the strengths of each data type while presenting a single integrated result to the user.
Solution Approach 2:
The defect detection system is designed with multi-functionality to handle various data types uniformly. The same processing pipeline can analyze image data, compare against CAD models, and cross-reference manufacturing records, providing a universal solution that adapts to different inspection scenarios without requiring separate specialized systems.
3Reliability
If comprehensive multi-modal analysis is performed on all blades, then defect detection quality improves, but inspection time increases
Solution Approach 1:
The system applies partial analysis by focusing computational resources on blades that show anomalies or deviations from expected patterns. Rather than performing exhaustive multi-modal analysis on every single blade, the system identifies suspicious cases and applies comprehensive analysis selectively, maintaining high defect detection quality while reducing overall inspection time.
Solution Approach 2:
The system performs preliminary filtering and pre-analysis to identify blades that require comprehensive multi-modal inspection. By conducting initial screening and prioritizing candidates based on risk factors or preliminary anomaly detection, the system prepares a reduced set of blades for detailed analysis, thereby improving throughput without compromising detection quality for critical cases.
Data Source
Figure 1
Figure 2
AI summary
A method and system (2) for performing automated defect detection is disclosed. The system may include at least one database (20), an image capture device (10) and a processor (14). The method may comprise providing at least one database (20) for storing information used in processing data to detect a defect in at least one member of a plurality of members in a device. The information may include a plurality of different modes of data. The method may further comprise providing a processing unit (14) for processing the information; receiving (102), by the database, updates to the information; identifying a potential defect in a first mode of data; applying (105), by the processing unit, analysis of a second mode of data, the analysis of the second mode of data triggered by the identifying, the second mode of data different than the first mode of data; and reporting defects based on the results of the applying.