Interpretation

INTERPRETATION OF MEASUREMENT DATA

Our simulation and automated measurement approaches generate large volumes of data that need real-time interpretation.

We develop novel machine learning (ML) and artificial intelligence (AI) approaches to modelling our data and interpreting measurement results to enable real-time compensation of our measurement processes.

MACHINE LEARNING

We use Machine Learning techniques to validate acoustic propagation in complex materials (with material scattering and thermal gradients). This is used to correct imaging algorithms as shown in the figure.

Machine Learning is also used to automatically detect defects in complex NDT data sets.

Machine Learning

SIMULATION

Simulation plays a key role in understanding how transducers interact with our sample materials, and to understand the influence of changes in geometry and material properties when inspecting them.

Fast simulations are enabled using the latest GPU accelerated computer hardware to bring supercomputer performance to the desktop.

Simulation

SIGNAL PROCESSING

Signal conditioning and signal processing are core skills enabling us to make measurements in challenging materials and noisy environments.

Signal Processing

INTELLIGENT DECISION MAKING

Interpretation of NDT measurement data is traditionally highly manually intensive. By bringing together data and knowledge driven approaches, we are able to facilitate reduction in interpretation of measurement data from diverse structures such as power stations, and nuclear reactors.

Intelligent Decision Making

SEARCH is based at The University of Strathclyde in Glasgow

Transduction

Deployment

Interpretation

Asset Inspection
In-process NDE
Manufacturing