Currently, the distribution of defects in a printed part is obtained thanks to non-destructive inspection methods such as tomography or radiography, being necessary to reach minimum resolutions to be able to detect defects that could be critical for the correct functioning of the part. This process is cost and time inefficient and often leads to discarding parts after costly manufacturing, machining or other post-processing and inspection processes.

The SCALDATA project, approved within the Digital Technologies Projects line of the Call for Asociaciones Empresariales Innovadoras 2022 (AEI) promoted by the Ministry of Industry, Tourism and Trade of the Government of Spain, proposes the use of artificial intelligence to generate predictions for the typology, location and effect of defects in parts produced in powder metal bed additive manufacturing processes.

This type of advanced manufacturing process has many advantages from the point of view of both material savings and design optimisation. However, there are still many factors that prevent the adoption of this technology, especially the appearance of defects in the
manufactured parts, as defects are generally unpredictable, random and difficult to detect.

For all these reasons, the results of the SCALDATA project contribute to the implementation of this type of manufacturing technology, which allows the production of optimised parts while reducing the material and energy required for the production processes.