Quickest Change Point Detection in Shape Inspection of Additively Manufactured Parts Under a Multi-Resolution Framework

Abstract

In-situ layer-by-layer inspection is essential to achieving the full capability and advantages of additive manufacturing in producing complex geometries. The shape of each inspected layer can be described by a 2D point cloud obtained by slicing a thin layer of 3D point cloud acquired from 3D scanning. In practice, a scanned shape must be aligned with the corresponding base-truth CAD model before evaluating its geometric accuracy. Indeed, the observed geometric error is attributed to systematic, random, and alignment errors, where the systematic error is the one that triggers an alarm of system anomalies. In this work, a quickest change detection (QCD) algorithm is applied under a multi-resolution alignment and inspection framework 1) to differentiate errors from different error sources, and 2) to identify the layer where the earliest systematic deviation distribution changes during the printing process. Numerical experiments and a case study on a human heart are conducted to illustrate the performance of the proposed method in detecting layer-wise geometric error.

Predictive modeling for online in-plane shape deviation inspection and compensation of additive manufacturing

“…; ng, respectively. After aligning the two shape profiles by applying the multiresolution alignment framework (Jin et al, 2020a(Jin et al, , 2020b, the in-plane deviation of layer l can be obtained by:…”

“…The key idea of the multiresolution alignment and inspection is for hierarchically interpreting the shape profile with different levels of detail. In Jin et al (2020aJin et al ( , 2020b, the multiresolution analysis is achieved by wavelet decomposition (Mallat, 1989) to decompose the original shape profiles into different levels of coefficient profiles, which are further applied in a three-step alignment process as illustrated in Figure 5.…”

“…. ; mg. To accommodate multiresolution alignment and inspection process for the shape with concave features, it is suggested to conduct the process with a feature-based inspection scheme as proposed in Jin et al (2020aJin et al ( , 2020b. Then the change point detection, as well as the corresponding predictive modeling and compensation, will be conducted on each feature.…”

“…In this paper, the proposed Seq2Seq model with an attention mechanism is implemented to predict the in-plane deviation of the subsequent layers to be printed and the occluded portion of the layer under inspection based on the deviation of previous layers. Based on the predicted deviations, a change point detection method (Jin et al, 2020a(Jin et al, , 2020b) is applied to decide if a compensation strategy should be performed or not. The proposed online compensated strategy is generated by changing the G-code directly to allow the machine to perform the fabrications by following the modified toolpaths.…”

“…In this work, the predictive modeling and compensation will be performed on the (x-y plane) shape deviation, which is referred to as the metric for evaluating geometric accuracy. As proposed in one of the recent works (Jin et al, 2020a(Jin et al, , 2020b, the in-plane…”