Deep semi-supervised learning of dynamics for anomaly detection in laser powder bed fusion

Abstract

Highly complex data streams from in-situ additive manufacturing (AM) monitoring systems are becoming increasingly prevalent, yet finding physically actionable patterns remains a key challenge. Recent AM literature utilising machine learning methods tend to make predictions about flaws or porosity without considering the dynamical nature of the process. This leads to increases in false detections as useful information about the signal is lost. This study takes a different approach and investigates learning a physical model of the laser powder bed fusion process dynamics. In addition, deep representation learning enables this to be achieved directly from high speed videos. This representation is combined with a predictive state space model which is learned in a semi-supervised manner, requiring only the optimal laser parameter to be characterised. The model, referred to as FlawNet, was exploited to measure offsets between predicted and observed states resulting in a highly robust metric, known as the dynamic signature. This feature also correlated strongly with a global material quality metric, namely porosity. The model achieved state-of-the-art results with a receiver operating characteristic (ROC) area under curve (AUC) of 0.999 when differentiating between optimal and unstable laser parameters. Furthermore, there was a demonstrated potential to detect changes in ultra-dense, 0.1% porosity, materials with an ROC AUC of 0.944, suggesting an ability to detect anomalous events prior to the onset of significant material degradation. The method has merit for the purposes of detecting out of process distributions, while maintaining data efficiency. Subsequently, the generality of the methodology would suggest the solution is applicable to different laser processing systems and can potentially be adapted to a number of different sensing modalities.

Induced alterations driven compromised structural properties in additively manufactured products

“…The wide range of current AM control and monitoring systems propose such flexible techniques. Because pattern recognition modeling includes input or output correlations along with structured models, it enables efficient description of the physical nature of the problem (Larsen and Hooper, 2022).…”

In-process monitoring and prediction of droplet quality in droplet-on-demand liquid metal jetting additive manufacturing using machine learning

“…From a broader vista, machine learning has been used extensively for process monitoring and quality assurance in additive manufacturing (Meng et al, 2020;Qin et al, 2022;Wang et al, 2020). Recently, Larsen and Hooper (2022) developed a deep semi-supervised Variational Autoencoder machine learning model that used high-speed imaging data for anomaly detection in laser powder bed fusion (LPBF) additive manufacturing. The proposed approach was able to differentiate between optimal and undesirable processing conditions with an accuracy in terms of the area under the curve (receiver operating curve) ~0.99.…”

An Electrical Resistance Diagnostic for Conductivity Monitoring in Laser Powder Bed Fusion

“…Several X-ray synchrotron diagnostics for high-speed imaging of the LPBF process have been developed in the past five years [ 17 , 18 , 19 , 20 ]. These techniques aim to understand the sub-surface material behavior and dynamics during laser-induced melting such as quantifying pore formation mechanisms [ 21 ], spatter ejection mechanisms [ 22 ], melt pool and vapor depression morphologies [ 23 , 24 ], liquid flow in the melt-pool [ 25 ], and other physical phenomena during the LPBF process [ 26 ].…”