Title: Recent Advances in Artificial Intelligence Applications to Composites Fabrication

Authors: Scott Blake

DOI: 10.33599/nasampe/c.19.0658

Abstract: Since early 2018, artificial intelligence (AI) has been demonstrated as a viable mechanism for application development in composites fabrication. In the first known AI field test for composites application, a convolutional neural network (CNN) successfully generated an analysis algorithm for an automatic inspection system to detect foreign objects and debris (FOD) of various kinds. A second field test used third-party AI software to successfully generate an analysis algorithm that verifies the presence or absence of peel ply for bonding processes. In both cases, the resulting algorithm achieved near 100 percent accuracy in classifying images (which had been captured via machine vision by the automatic inspection system) as FOD/peel ply present or FOD/peel ply absent. Ongoing studies of AI application to composites fabrication include additional algorithm development (e.g. laps and gaps in automated tape laying and fiber placement applications) as well as AI-based deep learning in which data/images collected from actual manufacturing operations serve as raw data for AI-enabled design and process improvements. Especially given the relatively high number of process variables in composites fabrication (e.g. fiber and resin type, fiber form, fiber orientation, ply schedules, resin infusion/prepregging method, layup mechanism, cure mechanism), the potential for AI to accelerate composites fabrication development is significant. Both completed field testing results and planned AI-enabled composites fabrication development will be discussed.

References: 1. Hahn, Thomas. “Artificial Intelligence: Optimizing Industrial Operations.” The European Files 54 (2018): 30. 2. “Artificial Intelligence (AI) vs. Machine Learning vs Deep Learning.” 25 Jan 2019: https://skymind.ai/wiki/ai-vs-machine-learning-vs-deep-learning. 3. Cemenska, Joshua, T. Rudberg, M. Henscheid. “Automated In-Process Inspection for AFP Machines.” SAE Int. J. Aerosp. 8(22) (2015). DOI: https://doi.org/10.4271/2015-01-2608 4. Black, Sara. “Improving Composites Processing with Automatic Inspection.” CompositesWorld 4(2) (2018): 38-41. 5. Blake, Scott. “Manufacturing Readiness: The Case for Automatic Inspection in Composites Fabrication.” SAMPE Journal 54(2) (2018): 20-26. 6. Blake, Scott, A. Serna. “Automated Peel Ply Foreign Object Damage (FOD) Prevention in Aerospace Bonding Operations.” SAMPE Journal 51(6) (2015). 7. Thilmany, Jean. May 2018. “Artificial Intelligence Transforms Manufacturing,” ASME.org, 25 Jan 2019: https://www.asme.org/engineering-topics/articles/manufacturing-design/artificial-intelligence-transforms-manufacturing. 8. Ben-Bassat, Avner. Aug.-Sept. 2016. “Real-time tracking of parts & material genealogy improves throughput and quality,” JEC Composites Magazine 107 (2016): 34-35. 25 Jan 2019:https://www.plataine.com/wp-content/uploads/2016/10/JEC-Magazie107_Manufacturing_Platain-PRINT.pdf

Conference: CAMX 2019

Publication Date: 2019/09/23

SKU: TP19-0658

Pages: 8

Price: $16.00