Application of artificial intelligence methods in nondestructive testing

Abstract

Non-destructive examination serves detecting material inconsistencies, assessment of material properties and measurements of object dimensions without changing their utility properties. Methods and techniques of non-destructive examination are applied in industry due to necessity of providing high quality semi-products, finished products, devices and structures. They are commonly applied to examine condition of technical devices utilised by people in order to detect defects which might lead to destruction of a structure. They are particularly significant in rocket, aviation, chemical, petrochemical, armaments, nuclear and conventional energy industries, ship building, railway industries, road, bridge and tunnel constructions as well as in other areas, and especially there, where threats to human life and environment can be found. Currently, non-destructive examinations develop in three directions: improvement of known and applied research methods, implementation of new methods and application of artificial intelligence to support a process of examination itself and assessment of quality of an examined element. The first and second direction of development shall not be discussed in detail in this study, except for a small remark.