Bridging the Gap Between AI and Reality • Rhodes, Greece
Time: Tuesday, 4.11
Room: Room C
Authors: Denis Dowling, Cathal Hoare, Tiziana Margaria
Abstract: Metal Additive Manufacturing (AM) offers significant advantages over traditional methods but continues to face challenges in process consistency, repeatability, and reliability. Small and medium-sized enterprises (SMEs), often lacking digitalization expertise, require accessible solutions. This paper proposes an incremental strategy for adopting Digital Shadows (DSs) via a Minimum Viable Digital Shadow (MVDS), which leverages hardcoded physical models to provide early value and generate curated datasets for future machine learning (ML) applications. The MVDS is implemented on a Service-Oriented Architecture and initially focuses on physics-based anomaly detection in Laser Powder Bed Fusion (LPBF) using in-process monitoring data and imagery. This approach constructs a graph-based digital thread that evolves over time to incorporate ML models. A case study illustrates how SMEs can transition from basic MVDS implementations to more advanced DSs, supporting continuous improvement and broader digital integration in metal AM.