Multi-Sensor Coregistration for Pixel-Level Manufacturing Inspection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional manufacturing inspection processes treat data streams from various sensors separately, failing to leverage salient information for quality metrics, leading to inefficiencies in part assessment and process monitoring.
Innovation Solution
A technique for coregistering data streams from diverse sensors with image data, creating a rich tensor that combines spatial and temporal information, enabling advanced analytical approaches for improved quality control and defect detection in manufacturing processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If data streams from various sensors are treated separately in conventional inspection processes, then the inspection system is simpler to implement and operate, but the quality assessment capability and defect detection accuracy are insufficient
Solution Approach 1:
The patent merges multiple separate data streams from diverse sensors (imaging sensors, non-imaging sensors, process sensors) into a unified multi-dimensional tensor data structure. This combining approach enables comprehensive quality assessment by integrating spatial, spectral, and process information, directly resolving the contradiction between measurement precision and device complexity through systematic data fusion rather than separate analysis
Solution Approach 2:
The patent creates a universal tensor-based data structure that can accommodate multiple sensor types and data modalities within a single framework. This multi-functional approach allows the same inspection system to handle diverse sensor inputs (optical, thermal, mechanical, chemical) uniformly, improving quality assessment capability while providing a standardized processing pipeline that manages complexity
2Reliability
If comprehensive sensor data is collected and analyzed for every part, then quality control and defect detection improve, but inspection time and processing speed increase
Solution Approach 1:
The patent performs preliminary coregistration and organization of multi-sensor data into tensor structures during the manufacturing process itself, before final inspection decisions are made. By pre-processing and structuring the data in advance, the system enables faster real-time quality assessment and defect detection, reducing inspection time while maintaining comprehensive quality control through advance data preparation
Solution Approach 2:
The patent enables continuous monitoring and analysis of manufacturing processes by integrating sensor data collection with the manufacturing workflow. Rather than performing discrete, time-consuming inspections, the system continuously processes sensor streams in real-time, maintaining quality control without adding inspection delays through uninterrupted data flow and analysis
3Loss of information
If multiple diverse sensors are integrated to capture comprehensive manufacturing data, then the information richness for quality metrics improves, but the data processing and analysis complexity increases
Solution Approach 1:
The patent transforms multi-sensor data into a multi-dimensional tensor structure that adds organizational dimensions (spatial, spectral, temporal, process) to the data. This dimensional organization preserves all information from diverse sensors while providing a structured framework that simplifies processing through consistent mathematical operations on tensors, resolving the contradiction between information richness and processing complexity
Solution Approach 2:
The patent changes the parameter representation of sensor data by converting diverse sensor outputs into a unified tensor format with consistent data types and structures. This parameter transformation enables standardized processing algorithms to handle all sensor inputs uniformly, reducing analysis complexity while maintaining the full information content through appropriate tensor dimensions and properties
Data Source
AI summary
A method of manufacturing a physical object comprises, during a process of manufacturing the physical object by a machine, capturing image data of at least a portion of the physical object and other sensor data related to the machine or to the at least a portion of the physical object. The method further comprises, during the process of manufacturing the physical object, for each of the plurality of pixels of the image data, coregistering the image data with the other sensor data on a pixel-by-pixel basis, storing the coregistered image data and other sensor data in association with each other in a data structure, and using at least a portion of the coregistered image data and other sensor to detect an anomaly in the physical object or in the process of manufacturing the physical object.