Towards explainable augmented intelligence (AI) for crack characterization
Résumé
Crack characterisation is one of the central tasks of NDT&E (the Non-Destructive Testing and Evaluation) of industrial components and structures. These days data necessary for carrying out this task are often collected using ultrasonic phased arrays. Many ultrasonic phased array inspections are automated but data interpretation is not. This paper offers an approach to design-ing an explainable AI (augmented intelligence) to meet this challenge. It describes a C code called AutoNDE, which comprises a signal-processing module based on a modified total focusing meth-od that creates a sequence of two-dimensional images of an evaluated specimen; an im-age-processing module, which filters and enhances these images; and an explainable AI module - a decision tree, which selects images of possible cracks, groups those of them that appear to rep-resent the same crack and produces for each group a possible inspection report for perusal by a human inspector. AutoNDE has been trained on 18 datasets collected in three different laborato-ries by imaging steel specimens with large smooth planar notches, both embedded and sur-face-breaking. The paper presents results of this training and describes in detail an approach to dealing with the main source of error in ultrasonic data - undulations in the specimen surfaces.
Fichier principal
applsci-11.pdf (1.91 Mo)
Télécharger le fichier
Correction_Fradkin_et_al_Towards_Explainable_Augme.pdf (286.29 Ko)
Télécharger le fichier