Task-driven data fusion for additive manufacturing: Framework, approaches, and case studies

• DOI: 10.1016/j.jii.2023.100484
• Publication Date: 2023-06-16T03:43:42Z
• Authors: Fu Hu; Ying Liu; Yixin Li; Shuai Ma; Jian Qin; Jun Song; Qixiang Feng; Xianfang Sun; Qian Tang
• Keywords: Sensor Fusion; Data Analysis; Data reliability

Additive manufacturing (AM) has been envisioned as a critical technology for the next industrial revolution. Due to advances in data sensing and collection technologies, large amount of data, generated from multiple sources AM production, becomes available relevant analytics improve process reliability, repeatability, part quality. However, processes occur over wide range spatial temporal scales where generally involves different types, dimensions structures, leading difficulties when integrating then analysing. Hence, what way how integrate heterogeneous or merge information lead significant challenges systems. This paper proposed task-driven fusion framework that enables integration modalities based on tasks support decision-making activities. In this framework, activities involved task are identified first place. Then, required is identified, collected, characterised. Finally, techniques employed applied characteristics analytics. The best fit requirements selected final approach. Case studies research directions AM, including energy consumption prediction, mechanical properties prediction additively manufactured lattice investigation remelting density, were carried out demonstrate feasibility effectiveness approaches.