Improving Inspector Training by Incorporating Artificial Defect Data in Guided Wave Testing
e-Journal of Nondestructive Testing · 2026-07 · 31 (7)
DOI: 10.58286/32902 ↗Kurz gefasst
Fachartikel in der Zeitschrift e-Journal of Nondestructive Testing, 2026-07. Verfasst von Diana olivia Martinez-trejo und 2 weiteren Autor:innen.
Abstract (Originalsprache)
Training inspectors in non-destructive testing (NDT) is challenging due to the lack of datasets that include defects. Traditional training methods rely on a small number of physical samples, which may have some defects, but do not show the wide variety of shapes, severities, and conditions that inspectors will face in the field. To address this issue, Guided Ultrasonics Ltd. (GUL) has created a new method that combines real guided wave signals with synthetic defect echoes made by the POGO program from Imperial College London. This approach generates realistic and diverse training datasets. As a result, it allows for fairer exams and better preparedness for inspectors. Using synthetic data has already improved training efficiency, fairness in examinations, and the development of analysis tools. This initiative tackles a vital challenge in the industry. It offers a scalable and effective solution to improve inspector skills and maintain high inspection standards.
Zitieren
APA
Martinez-trejo, D. O., Pavlakovic, B., & Vine, K. (2026). Improving Inspector Training by Incorporating Artificial Defect Data in Guided Wave Testing. e-Journal of Nondestructive Testing, 31(7). https://doi.org/10.58286/32902
Vancouver
Martinez-trejo DO, Pavlakovic B, Vine K. Improving Inspector Training by Incorporating Artificial Defect Data in Guided Wave Testing. e-Journal of Nondestructive Testing. 2026;31(7). doi:10.58286/32902
BibTeX
@article{Martineztrejo2026,
author = {Martinez-trejo, Diana olivia and Pavlakovic, Brian and Vine, Keith},
title = {Improving Inspector Training by Incorporating Artificial Defect Data in Guided Wave Testing},
journal = {e-Journal of Nondestructive Testing},
year = {2026},
volume = {31},
number = {7},
doi = {10.58286/32902}
}
EUVIMED