Title: GENERATING GOOD DATA FOR AI-BASED AUTOMATIC INSPECTION AND REMEDIATION OF LARGE-SCALE COMPOSITE COMPONENTS

Authors: Scott Blake

DOI: 10.33599/nasampe/s.23.0149

Abstract: Artificial intelligence (AI) – specifically, deep learning (DL) – is moving rapidly into industrial applications, leveraging recently realized processing power to transform manufacturing processes. Such transformations are possible only when good data is readily available. Indeed, the availability of good data is central to achieving the benefits of Industry 4.0 through AI/DL and other enabling technologies. Collecting good data during the fabrication of large, complex composite structures has been difficult because the industry remains dependent on outdated manual inspection processes. Additionally, the size, complexity and variations comprising aerospace composite parts render inapplicable the current means of data acquisition employed in other manufacturing efforts. This paper describes an imaging system that captures small regions of large, complex fields, accompanied by photogrammetric transforms calibrating the pixels into the coordinate system of the nominal model. The calibrated images are easily tagged and used to train DL classifiers that support real-time automatic inspection. The paper explains both the DL training process and subsequent pinpointing of areas of concern identified in the calibrated images via laser projection on the actual work-in-progress (WIP). This novel approach to AI-based automatic inspection supports rapid verification and/or remediation of critical WIP attributes, yielding the lowest cycle time and cost.

References: 1. Andrew T. Modjeski, Interview by Scott Blake, September 12, 2022. 2. S. Blake, “The Democratization of Artificial Intelligence/Machine Learning to Enable Widespread Application of Automatic Inspection to Composites Fabrication,” CAMX Technical Paper TP-21-451, 2021. 3. T. Hahn, “Artificial Intelligence: Optimizing Industrial Operations.” The European Files 54 (2018): 30. 4. G. Nguyen, S. Dlugolinsky, M. Bobák et al., “Machine Learning and Deep Learning Frameworks and Libraries for Large-Scale Data Mining: a Survey,” Artif Intell Rev 52, 77–124 (2019). DOI: https://doi.org/10.1007/s10462-018-09679-z. 5. S. Fahle, C. Prinz and B. Kuhlenkötter, “Systematic Review on Machine Learning (ML) for Manufacturing Processes – Identifying Artificial Intelligence (AI) Methods for Field Application,” 53rd CIRP Conference on Manufacturing Systems Proceedings 93 (2020), 413–418. 6. T. Rudberg, J. Nielson, M. Henscheid, and J. Cemenska, “Improving AFP Cell Performance,” SAE Int. J. Aerosp. 7(2):2014, DOI:10.4271/2014-01-2272. 7. J. Cemenska, T. Rudberg, M. Henscheid et al., “AFP Automated Inspection System Performance and Expectations,” SAE Technical Paper 2017-01-2150, 2017, DOI:10.4271/2017-01-2150. 8. L. Scime, D. Siddel, S. Baird and V. Paquit, “Layer-Wise Anomaly Detection and Classification for Powder Bed Additive Manufacturing Processes: A Machine-Agnostic Algorithm for Real-Time Pixel-Wise Semantic Segmentation,” Additive Manufacturing 36 (2020), DOI:10.1016/j.addma.2020.101453. 9. Vincent Paquit, Interview by Scott Blake and Karen Mason, August 18, 2020. 10. J.F. Arinez, Q. Chang, R.X. Gao, C. Xu and J. Zhang (August 13, 2020). “Artificial Intelligence in Advanced Manufacturing: Current Status and Future Outlook.” ASME. J. Manuf. Sci. Eng. November 2020; 142(11): 110804. DOI: https://doi.org/10.1115/1.4047855. 11. U.S. Khan, J. Iqbal and M.A. Khan, “Automatic Inspection System Using Machine Vision,” 34th Applied Imagery and Pattern Recognition Workshop (APR’05), 2005, pp. 6-217, DOI: 10.1109/AIPR.2005.20. 12. “An Overview of the U.S. Commercial Aircraft Fleet,” by Forecast International, 10/1/2019, accessed 8/6/2020 on Defense & Security Monitor blog.

Conference: SAMPE 2023

Publication Date: 2023/04/17

SKU: TP23-0000000149

Pages: 11

Price: $22.00