Product quality assurance in Powder Bed Fusion through the application of ML/AI tools: a critical review

Abstract

Due to complex physical phenomena, which govern the process of Powder Bed Fusion (PBF), the feasibility and repeatability of the PBF production is often put in question, having as consequence unsatisfactory implementation of PBF at the industrial level. For this reason, many different research approaches have been carried out in terms of improving the governance of the process and its quality indicators. The overall boost of the Machine Learning and Artificial Intelligence (ML/AI) tools in the last decade has brought a significant advantage in this aspect, since these tools enable evaluation of massive amounts of data and extraction of complex relations which are not easily recognizable by the human eye or simple statistical tools. This paper provides an analysis of critical aspects of product quality in PBF, then offers a thorough critical review of the most relevant research efforts in the field of product quality assurance in Metal Additive Manufacturing through the application of ML/AI methods, to finally detect the major gaps, opportunities and future lines.