Contribution
Data-Driven Anomaly Detection of Powder Bed Fusion with Acoustic Emission Data
* Presenting author
Abstract:
Powder bed fusion (PBF) is an additive manufacturing process used extensively in high-value industries such as aerospace, automotive, and medical engineering, due to its possibility to produce complex and lightweight components. Ensuring consistent quality and detecting defects in real time is critical to optimizing production quality and minimizing waste. This work proposes a data-driven approach for anomaly detection in PBF using acoustic emission (AE) data. AE sensors, with their high sampling rate, provide rich information in the high-frequency range, enabling the detection of abrupt changes in the process that may indicate anomalies. By applying convolutional neural networks (CNNs), adapted from image processing, to the spectrogram of the AE data we aim to efficiently detect defects, which shows the potential to save energy, material, time, and costs in manufacturing.