Method and system for generating time-efficient synthetic non-destructive testing data

This invention describes a software system that creates artificial non-destructive testing (NDT) data. It takes real NDT measurements and uses numerical simulations, such as Finite Element Analysis, to generate a large set of training data with various flaws. This training data then teaches a Deep Convolutional Generative Adversarial Network (DCGAN) to produce entirely new, synthetic NDT datasets, which can be used to train other AI models without needing more real-world experiments.