AMAI’s goal is to bring innovative Artiﬁcial Intelligence (AI) solutions in Additive Manufacturing for in-line metrology to reconstruct the geometry of the part while it is being printed and quickly detect geometrical distortions.
High defectiveness and scrap rates still limit the industrial adoption of Additive Manufacturing (AM) processes. Several sources of undesired variability may lead to various kinds of defects with detrimental effects on the functional performances of the manufactured part. Moreover, complex shapes enabled by the AM production paradigm are difficult and expensive to inspect after the process.
AMAI aims at bringing to the market a novel solution for in-situ and in-line detection of geometrical errors and distortions in Powder Bed Fusion (PBF), to quickly detect anomalies while the part is being produced. The method relies on powder bed cameras that are already available in most PBF systems, but introduces a brand new combination of advanced image processing and data analytics to accurately reconstruct the geometry of each produced layer and detect possible geometrical deviations from the nominal shape. Thanks to this innovative capability, geometrical errors and distortions may be quickly detected while the part is being produced, anticipating the identification of non-conformities and enabling the reduction of scraps and costs associated to lack of quality. The tool can be easily installed on existing PBF machines, for both laser- and electron beam-based technologies.
The project was completed in 2021. It was coordinated by the Department of Mechanical Engineering, Politecnico di Milano, and it was carried out in partnership with GE Avio Aero, Beamit SpA, CNR Stiima and the Chalmers University of Technology. A beta version of the AMAI in-situ monitoring toolkit is now available for testing on any laser based-PBF system. The toolkit for electron beam-based PBF is currently available in alpha version. The AMAI team is working on the continuous improvement of the method, aiming at a full industrialization of the developed solutions.