One Recording, Several Readings: Checking an AI Flaw Call in Non-Destructive Testing of Metals and Bridges Against the Instrument's Own Settings

Inspectors test welds, pipes, metal parts and bridges without breaking them, using ultrasound and X-rays, and AI programs now help to call out flaws. Today the AI gives one answer from one way of processing the recording. This note describes a simple check that can sit on top of any such program. The part is inspected once. From the raw recording the instrument holds, several pictures are made by changing only the instrument's own processing settings, within the range the maker and the inspection procedure permit. The AI reads every picture. If the answers stay together, the recording supports the flaw call firmly. If they spread out, the call depends on a setting and deserves a closer look. Other methods shake the data with invented changes, such as flips or artificial noise. This method uses the settings the instrument itself allows, so the uncertainty it measures is real, and it is read for each recording. One number, the spread value, tells which case it is. A computer-decided shunting action then moves the call that needs care to the front of a senior inspector's list, while all other calls carry on. The check needs no second visit, no second exposure and no change to the AI program or to the qualified procedure. Safer bridges, pipelines and aircraft parts, fewer needless stoppages and fewer repeat visits for inspectors are the benefits to society. Patent pending. Keywords: non-destructive testing; ultrasonic inspection; AI flaw detection; bridges; instrument settings; uncertainty; spread value; shunting action

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-10-27
DOI
https://doi.org/10.5281/zenodo.23692182
Primary Topic
Non-Destructive Testing Techniques
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

One Recording, Several Readings: Checking an AI Flaw Call in Non-Destructive Testing of Metals and Bridges Against the Instrument's Own Settings

Prakash Vaithyanathan
Zenodo (CERN European Organization for Nuclear Research)
Non-Destructive Testing Techniques
article

One Recording, Several Readings: Checking an AI Flaw Call in Non-Destructive Testing of Metals and Bridges Against the Instrument's Own Settings

Prakash Vaithyanathan
article en

Abstract

Inspectors test welds, pipes, metal parts and bridges without breaking them, using ultrasound and X-rays, and AI programs now help to call out flaws. Today the AI gives one answer from one way of processing the recording. This note describes a simple check that can sit on top of any such program. The part is inspected once. From the raw recording the instrument holds, several pictures are made by changing only the instrument's own processing settings, within the range the maker and the inspection procedure permit. The AI reads every picture. If the answers stay together, the recording supports the flaw call firmly. If they spread out, the call depends on a setting and deserves a closer look. Other methods shake the data with invented changes, such as flips or artificial noise. This method uses the settings the instrument itself allows, so the uncertainty it measures is real, and it is read for each recording. One number, the spread value, tells which case it is. A computer-decided shunting action then moves the call that needs care to the front of a senior inspector's list, while all other calls carry on. The check needs no second visit, no second exposure and no change to the AI program or to the qualified procedure. Safer bridges, pipelines and aircraft parts, fewer needless stoppages and fewer repeat visits for inspectors are the benefits to society. Patent pending. Keywords: non-destructive testing; ultrasonic inspection; AI flaw detection; bridges; instrument settings; uncertainty; spread value; shunting action

Zenodo (CERN European Organization for Nuclear Research)
Industry, innovation and infrastructure
Openalex Percentile: Top 23%
Non-Destructive Testing Techniques
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon