EUVIMEDObjevte EUVIMED
Crossref

X-Ray Defect Recognition – Reliable and Efficient with AI Support

Christopher Zepp

e-Journal of Nondestructive Testing · 2026-07 · 31 (7)

DOI: 10.58286/33099 ↗

At a glance

Journal article in e-Journal of Nondestructive Testing, 2026-07. Written by Christopher Zepp.

AbstraktBibliografické údajeCiteRelated studies

Abstract (original language)

The advancement of artificial intelligence (AI) is transforming industrial non-destructive testing (NDT), particularly in radiographic testing (RT). This contribution presents AI-driven solutions that enhance inspection reliability and efficiency by automating X-ray image interpretation. Traditional RT evaluation depends strongly on human expertise, leading to variable results and extended assessment times. AI-based approaches address these limitations through automated interpretation models that improve agreement rates, ensure a consistent probability of detection (POD), and significantly reduce evaluation time. The presentation outlines the ongoing digital transformation in NDT — from manual interpretation to AI-assisted and increasingly automated inspection workflows. Real-world applications such as weld defect detection, surface-mount device (SMD) part counting, and image quality indicator (IQI) recognition demonstrate measurable improvements in throughput, decision consistency, and error reduction, with the highest yield currently observed in the casting industry. Industrial implementations already show substantial gains in detection reliability and process efficiency, confirming AI’s value in practical RT environments. The paper concludes with an outlook on next-generation developments combining AI with robotics and digital twin technologies to further enhance automation, standardization, and overall NDT process performance.

Cite

APA

Zepp, C. (2026). X-Ray Defect Recognition – Reliable and Efficient with AI Support. e-Journal of Nondestructive Testing, 31(7). https://doi.org/10.58286/33099

Vancouver

Zepp C. X-Ray Defect Recognition – Reliable and Efficient with AI Support. e-Journal of Nondestructive Testing. 2026;31(7). doi:10.58286/33099

BibTeX
@article{Zepp2026,
  author = {Zepp, Christopher},
  title = {X-Ray Defect Recognition – Reliable and Efficient with AI Support},
  journal = {e-Journal of Nondestructive Testing},
  year = {2026},
  volume = {31},
  number = {7},
  doi = {10.58286/33099}
}
Start for free →