skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Layer-wise anomaly detection and classification for powder bed additive manufacturing processes: A machine-agnostic algorithm for real-time pixel-wise semantic segmentation

Abstract

Increasing industry acceptance of powder bed metal Additive Manufacturing requires improved real-time detection and classification of anomalies. Many of these anomalies, such as recoater blade impacts, binder deposition issues, spatter generation, and some porosities, are surface-visible at each layer of the building process. In this work, the authors present a novel Convolutional Neural Network architecture for pixel-wise localization (semantic segmentation) of layer-wise powder bed imaging data. Key advantages of the algorithm include its ability to return segmentation results at the native resolution of the imaging sensor, seamlessly transfer learned knowledge between different Additive Manufacturing machines, and provide real-time performance. The algorithm is demonstrated on six different machines spanning three technologies: laser fusion, binder jetting, and electron beam fusion. Finally, the performance of the algorithm is shown to be superior to that of previous algorithms presented by the authors with respect to localization, accuracy, computation time, and generalizability.

Authors:
; ; ;
Publication Date:
Research Org.:
Oak Ridge National Lab. (ORNL), Oak Ridge, TN (United States)
Sponsoring Org.:
USDOE Office of Energy Efficiency and Renewable Energy (EERE); USDOE Office of Nuclear Energy (NE)
OSTI Identifier:
1638745
Alternate Identifier(s):
OSTI ID: 1649400
Grant/Contract Number:  
AC05-00OR22725
Resource Type:
Journal Article: Published Article
Journal Name:
Additive Manufacturing
Additional Journal Information:
Journal Name: Additive Manufacturing Journal Volume: 36 Journal Issue: C; Journal ID: ISSN 2214-8604
Publisher:
Elsevier
Country of Publication:
Netherlands
Language:
English
Subject:
42 ENGINEERING

Citation Formats

Scime, Luke, Siddel, Derek, Baird, Seth, and Paquit, Vincent. Layer-wise anomaly detection and classification for powder bed additive manufacturing processes: A machine-agnostic algorithm for real-time pixel-wise semantic segmentation. Netherlands: N. p., 2020. Web. doi:10.1016/j.addma.2020.101453.
Scime, Luke, Siddel, Derek, Baird, Seth, & Paquit, Vincent. Layer-wise anomaly detection and classification for powder bed additive manufacturing processes: A machine-agnostic algorithm for real-time pixel-wise semantic segmentation. Netherlands. https://doi.org/10.1016/j.addma.2020.101453
Scime, Luke, Siddel, Derek, Baird, Seth, and Paquit, Vincent. Tue . "Layer-wise anomaly detection and classification for powder bed additive manufacturing processes: A machine-agnostic algorithm for real-time pixel-wise semantic segmentation". Netherlands. https://doi.org/10.1016/j.addma.2020.101453.
@article{osti_1638745,
title = {Layer-wise anomaly detection and classification for powder bed additive manufacturing processes: A machine-agnostic algorithm for real-time pixel-wise semantic segmentation},
author = {Scime, Luke and Siddel, Derek and Baird, Seth and Paquit, Vincent},
abstractNote = {Increasing industry acceptance of powder bed metal Additive Manufacturing requires improved real-time detection and classification of anomalies. Many of these anomalies, such as recoater blade impacts, binder deposition issues, spatter generation, and some porosities, are surface-visible at each layer of the building process. In this work, the authors present a novel Convolutional Neural Network architecture for pixel-wise localization (semantic segmentation) of layer-wise powder bed imaging data. Key advantages of the algorithm include its ability to return segmentation results at the native resolution of the imaging sensor, seamlessly transfer learned knowledge between different Additive Manufacturing machines, and provide real-time performance. The algorithm is demonstrated on six different machines spanning three technologies: laser fusion, binder jetting, and electron beam fusion. Finally, the performance of the algorithm is shown to be superior to that of previous algorithms presented by the authors with respect to localization, accuracy, computation time, and generalizability.},
doi = {10.1016/j.addma.2020.101453},
url = {https://www.osti.gov/biblio/1638745}, journal = {Additive Manufacturing},
issn = {2214-8604},
number = C,
volume = 36,
place = {Netherlands},
year = {2020},
month = {12}
}

Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record at https://doi.org/10.1016/j.addma.2020.101453

Save / Share:

Works referenced in this record:

Economics of additive manufacturing for end-usable metal parts
journal, February 2012


Deep learning
journal, May 2015


ImageNet Large Scale Visual Recognition Challenge
journal, April 2015