Friday 14 March 2025
Researchers have made a significant breakthrough in the field of non-destructive testing, developing a new method that uses machine learning algorithms to identify defects in ultrasonic scans of composite materials. The approach, which involves training models on images of defects and normal material, has been shown to be highly effective in detecting even small flaws.
The technique is particularly useful for inspecting complex structures such as aircraft components, where traditional methods can be time-consuming and labor-intensive. By automating the process, manufacturers can quickly identify potential issues and take corrective action before they become major problems.
The new method uses a type of image segmentation called instance segmentation, which involves identifying individual objects within an image rather than just separating different classes of material. This allows the model to focus on specific defects or regions of interest, making it more accurate and efficient.
In addition to its speed and accuracy, the approach has several other advantages. For one, it requires minimal training data, meaning that manufacturers can start using it quickly without having to gather large amounts of data first. It also doesn’t require any complex pre-processing of the ultrasonic scan images, which can be a time-consuming and difficult task.
The researchers tested their method on a dataset of 72 ultrasonic scans of composite panels, which are representative of real-world aerospace structures. They found that the model was able to detect defects with high accuracy, even when they were small or poorly defined.
One potential application of this technology is in the inspection of aircraft components for defects caused by manufacturing flaws or damage during service. Currently, inspectors have to manually examine each component using a range of techniques, including visual inspection and ultrasonic testing. However, this process can be slow and labor-intensive, and may not always detect small or subtle defects.
With the new method, manufacturers could automate much of the inspection process, allowing them to quickly identify potential issues and take corrective action before they become major problems. This could help to improve the safety and reliability of aircraft components, as well as reduce the time and cost associated with inspecting and repairing defective parts.
The researchers are now working to further develop their method, including testing it on larger and more complex datasets. They also hope to explore its potential applications in other fields, such as medical imaging or materials science.
Cite this article: “Automated Defect Detection in Composite Materials Using Machine Learning”, The Science Archive, 2025.
Machine Learning, Ultrasonic Scans, Composite Materials, Non-Destructive Testing, Defects Detection, Instance Segmentation, Image Analysis, Aerospace Industry, Aircraft Components, Manufacturing Inspection.
