Embedded Artificial Intelligence for Quality Assurance in Additive Manufacturing and Advanced 3D Material Characterization

AIQuAM3D is a French-German research project that will open up application fields of artificial intelligence (AI) and improve the use of AI in cutting-edge technologies. The SMEs Xploraytion (DE) and Novitom (FR), together with the research organizations SOLEIL (FR) and Fraunhofer IPK (DE), constitute the project consortium and will work together on transferring AI research into the application fields of additive manufacturing (AM) and synchrotron micro-computed tomography (SR-µCT). With regard to AM technology, an embedded AI model for real-time bulk defects prediction in Laser-Powder Bed Fusion (L-PBF) will be developed, enhancing the quality assurance of AM process, leading to the wide application and industrialization of the AM technology. The approach goes from data fusion of in-situ process signals (level 0 to 3), through labelling and training of in-situ monitoring models with quality-optimized µCT data sets, up to defect prediction model development, validation and implementation as embedded solution into a L-PBF system. In regard to 3D material characterization, AIQuaAM3D will develop AI tools to integrate with a tomography data analysis pipeline. The integration of the AI analysis with the SR-µCT acquisition will enable real time data analysis, even for fast acquistions, providing feedback on the quality of results being collected. This closer integration means that the system can warn the user in the event of poor-quality data and/or image artefacts. The total project duration is 3 years, the subject of the project is aligned with the priorities of the funding program set by the partner countries.