Overview
After obtaining a diploma in mechanical engineering in France (E.N.I of Tarbes ), I went to Cardiff University to start a PhD in optimisation and simulation of metal additive layer manufacturing using machine learning techniques.
My work is part of a European project called Manuela. This project funded by the European Union's Horizon 2020 research and innovation programme will develop and realize a metal additive manufacturing pilot line service covering the full AM development cycle.
During my first PhD year, I worked on the creation of a deep learning solution to classify layers quality in Electron beam melting (EBM) using a deep learning approach. This solution was able to classify electron-optical (ELO) layer images into 3 categories (porous, bulging and ideal) with an accuracy of 95%. A conference paper named "Automatised quality assessment in additive layer manufacturing using layer-by-layer surface measurements and deep learning" was published at the 14th CIRP Conference on Intelligent Computation in Manufacturing Engineering (2020).
Since the start of my second Pd.D year, I worked on the prediction of part's deformation caused by the additive manufacturing process. Deformation prediction is made with the use of geometrical deep learning using CAD meshes as inputs.
Publication
2022
- Liu, C., Le Roux, L., Körner, C., Tabaste, O., Lacan, F. and Bigot, S. 2022. Digital twin-enabled collaborative data management for metal additive manufacturing systems. Journal of Manufacturing Systems 62, pp. 857-874. (10.1016/j.jmsy.2020.05.010)
2021
- Le Roux, L. et al. 2021. Automatised quality assessment in additive layer manufacturing using layer-by-layer surface measurements and deep learning. Presented at: 14th CIRP Conference on Intelligent Computation in Manufacturing Engineering 2020, Naples, Italy, 15-17 July 2020, Vol. 99. Elsevier pp. 342-347., (10.1016/j.procir.2021.03.050)
Articles
- Liu, C., Le Roux, L., Körner, C., Tabaste, O., Lacan, F. and Bigot, S. 2022. Digital twin-enabled collaborative data management for metal additive manufacturing systems. Journal of Manufacturing Systems 62, pp. 857-874. (10.1016/j.jmsy.2020.05.010)
Conferences
- Le Roux, L. et al. 2021. Automatised quality assessment in additive layer manufacturing using layer-by-layer surface measurements and deep learning. Presented at: 14th CIRP Conference on Intelligent Computation in Manufacturing Engineering 2020, Naples, Italy, 15-17 July 2020, Vol. 99. Elsevier pp. 342-347., (10.1016/j.procir.2021.03.050)
Research
My current research is focused on the use of geometrical deep learning to predict the deformation caused by the printing process. This approach goal is to predict deformation for a defined printing parameter quicker than traditional simulation techniques to allow a faster iteration in the product design loop.
My previous work was on the images analysis and classification using convolution neural networks (CNN) of electron-optical (ELO) images from an Electron beam melting (EBM) printer at Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) university. This work resulted in the creation of a CNN with an accuracy of 95% and the publication of a conference paper ("Automatised quality assessment in additive layer manufacturing using layer-by-layer surface measurements and deep learning")