Tue, 02 March, 2021
The conference is supported by the German Welding Society (DVS), International Institute of Welding (IIW) and the American Welding Society, and takes place from 9-10 March 2021.
Norbert's presentation is titled Electron Beam Probing and Deep Learning to Improve Weld Quality, and will be alongside his industrial supervisor from TWI Colin Ribton, as well as his academic supervisors from Lancaster University Yingtao Tian and Andrew Kennedy.
Regarding his presentation, Norbert said,
"Recently, I trained a neural network model for the analysis of beam probe signals. It allows applying an image classification model to recognise the focus state of the beam. It is anticipated that research outcomes could lead to higher weld integrity, better process monitoring and reduced failure rates in service."
The objective of Norbert's PhD research in general is to analyse multiple input quality indicators to derive a quality assurance system.
Overall research of the project focuses on creating a data-driven, multi-sensor fusion system using artificial intelligence tools.
The quality indicators examined include BeamAssure signal, backscatter electron monitoring and images collected by welding camera inside the EB machine.
The National Structural Integrity Research Centre PhD Programme is based at TWI Ltd in Cambridge, UK, and currently has over 70 active students conducting research in Structural Integrity and related fields.
The centre is manage and operated by TWI and is closely linked to The Welding Institute.