The breakthrough addresses a longstanding challenge in nuclear safety inspections. Critical components in nuclear power plants undergo regular non-destructive testing using ultrasonic inspection methods to detect potential damage such as cracks that may develop during operation. These inspections generate vast amounts of measurement data that traditionally required specialist engineers hours to analyse manually.

The AI-powered automated assessment works alongside human inspectors, significantly improving both efficiency and reliability. The artificial intelligence system filters through the data and flags areas that may contain incipient damage, which inspectors then evaluate to make final determinations. Tasks that previously took hours can now be completed in minutes.

“The assessment that used to take hours is now done in minutes,” the company explained in the press release. The technology also makes inspections more repeatable and reduces the risk of human error, whilst ensuring inspectors retain ultimate decision-making authority.