Post
Teaching Steel Industry Equipment to See
A practical look at how AI, computer vision, and deterministic geometry can work together to turn messy real-world perception into reliable measurements a machine can actually use.
Automated Vision-Based Inspection of Drilled CFRP Composites Using Multi-Light Imaging
Deep learning for detecting defects in aerospace composite structures
- Stage
- Research Prototype
- Published
- Who made it?
Ali Maghami
- Tags
- Share
Publication
Authors: Maghami, A., Salehi, M., and Khoshdarregi, M.
Journal: CIRP Journal of Manufacturing Science and Technology, 2021
Overview
This work explores using computer vision and deep learning to automatically detect defects in drilled composite structures, making quality control faster and more consistent.
Problem
Detecting microscopic defects around drilled holes in aerospace composites is time-consuming and inconsistent when done manually. We investigated whether automated vision could improve inspection processes.
Approach
We developed a system using multiple light angles to reveal different types of defects, then trained deep learning models to recognize damage patterns and anomalies automatically.
Results
The system achieved high accuracy for defect detection and significantly reduced inspection time compared to manual methods, with validation on real manufacturing samples.
Application
This work demonstrates the potential for automating quality control in precision composite manufacturing, where visual inspection is critical but challenging to automate effectively.