AR Defect Visualization for Automated Fiber Placement Inspection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current automated fiber placement (AFP) manufacturing processes face challenges in efficient in-process inspection due to manual inspection methods, which are time-consuming, prone to human error, and lack reliable techniques for detecting defects like gaps, overlaps, and foreign object debris, leading to significant interruptions and increased production costs.
Innovation Solution
A defect visualization system utilizing augmented reality (AR) technology, including an AR viewing device and a visualization module that receives defect data, determines the device's position relative to the part, and generates display objects to align with defects, allowing for real-time visualization of defects in virtual three-dimensional space.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual inspection is used to detect defects in AFP processes, then defect detection capability is provided, but inspection time consumes 20-70 percent of total production time
Solution Approach 1:
The patent replaces manual visual inspection with an automated optical inspection system that uses cameras and image processing algorithms to detect defects. The system captures images of the composite part surface and automatically analyzes them to identify defects such as gaps, overlaps, and foreign object debris, eliminating the need for manual inspection and significantly reducing inspection time while maintaining or improving detection capability.
Solution Approach 2:
The patent introduces an automated image processing system as an intermediary between the manufacturing process and defect detection. This intermediary system processes images automatically using computer vision algorithms, serving as a mediator that transforms visual information into actionable defect data without requiring human operators, thereby speeding up the inspection process.
2Reliability
If manual inspection is used, then defect detection is performed, but inspection accuracy depends heavily on operator skill and training
Solution Approach 1:
The patent replaces the human operator's visual inspection system with an automated optical inspection system that uses standardized image processing algorithms. This substitution eliminates variability in inspection quality caused by differences in operator skill and training, providing consistent and repeatable defect detection results across different operators and shifts.
Solution Approach 2:
The patent transforms the inspection process from a human-dependent subjective evaluation to an objective automated analysis by changing the inspection parameters from human visual perception to machine vision algorithms. The system uses defined thresholds and detection criteria that remain constant, ensuring consistent defect detection regardless of operator expertise.
3Difficulty of detecting and measuring
If visual inspection is used, then defect identification is attempted, but low contrast between substrate and incoming tows makes visual identification difficult
Solution Approach 1:
The patent employs optical imaging systems that enhance the contrast between the substrate and incoming tows through sophisticated image processing techniques. The system captures images and applies algorithms that highlight subtle differences in reflectivity, texture, and color between the substrate and tows, making defects such as gaps, overlaps, and misalignments clearly visible even when the contrast is naturally low.
Solution Approach 2:
The patent may employ lighting systems that create controlled reflections or shadows to enhance defect visibility. By using specific lighting angles and intensities, the system creates optical effects that amplify the contrast between the substrate and tows, making defects more detectable through the imaging system.
Data Source
AI summary
A defect visualization system includes an augmented reality viewing device including a display, and a visualization module configured to receive defect data representative of the defects on the part, determine a position of the augmented reality viewing device in relation to the part, generate one or more display objects, wherein each display object of the display objects corresponds to a respective defect of the defects, and present the display objects on the display so that the display objects appear in alignment with the defects on the part.