October 2021, Vol. 248, No. 10
Features
3-D Metal Loss Profiles from MFL Data
By Johannes Palmer and Andrey Danilov, ROSEN Technology and Research Center GmbH, Lingen, Germany Magnetic flux leakage (MFL) data have high quality and coverage even under rough inline inspection (ILI) measurement conditions. The ILI vendor’s vast experience allows for heuristic MFL data evaluation, accepted by the pipeline industry based on hundreds of performance tests. Neural networks and artificial intelligence further improve the result probabilities. However, assumptions are never a definite calculation. The conventional process allows for high performance but affects result repeatability and the value of run comparison. Moreover, it does not prevent outliers as the worst remai

Comments