| About this Abstract |
| Meeting |
MS&T23: Materials Science & Technology
|
| Symposium
|
Additive Manufacturing: Design, Materials, Manufacturing, Challenges and Applications
|
| Presentation Title |
Monitoring Additive Manufacturing in Real-Time via Image-Based Machine Learning Techniques |
| Author(s) |
Peter Warren, Md Shahjahan Hossain, Pranta Sarkar, Asher Perez, Daniel Homa, Gary R. Pickrell, Ranajay Ghosh, Ramesh Subramanian, Subith Vasu, Navin Manjooran |
| On-Site Speaker (Planned) |
Navin Manjooran |
| Abstract Scope |
Monitoring the printing process in real-time is crucial for ensuring high-quality prints and identifying defects early on. This paper proposes a novel approach utilizing machine learning techniques to monitor and detect defects during the printing process using vision-based monitoring. Simulated additive manufacturing data serves as a representative dataset to train the machine learning models. The proposed system employs computer vision algorithms to analyze the real-time video feed of the printing process. By capturing frames at regular intervals, the system extracts visual features such as layer consistency, filament deposition, and surface texture. These features are then fed into a trained machine learning model, such as a convolutional neural network (CNN), to classify the printed object as defective or non-defective. To generate the training data, various types of simulated defects are introduced in the visual data, including layer shifting, over-extrusion, under-extrusion, and inconsistent extrusion. These defects are labeled to create a comprehensive dataset for training the machine learning model. |