Sensor modelling for BEM Technology: Gaussian Processes Approach
Overview
Broadband Electro-Magnetic (BEM) concepts have been deployed in non-destructive testing by Rock Solid Group (RSG) for almost two decades. Due to the presence of large parametric variations and difficulty in accurate physical modelling of the sensor to pipe interaction, data driven machine learning techniques have been proposed and are being investigated. The approach is validated through simulations and experiments.