Towards Automatic 3D Printing: A Framework for Closed-loop Process Monitoring and Control
Presenter: Katie Xu
Master's Student, University of Toronto - Continuing Studies
Assistant Director, C-MORE
3D printing has important advantages over traditional manufacturing processes. However, as it is a relatively new class of manufacturing technologies, problems of reliability and parameter optimization remain largely unresolved. Our work focuses on addressing some of these issues through in-situ monitoring and closed-loop control, using machine learning as a tool for the endeavour. The idea is to analyze the condition of the process by predicting key characteristics of the final product, and then to use this analysis for adjusting process parameters on the fly.
We imagine that this framework of predictive analysis leading to closed-loop control can be extended to a variety of applications outside of 3D printing. In a more general maintenance scenario, sensor readings can be used to assess the condition of equipment and to predict the condition at a future time. This information can then be used to determine appropriate maintenance activities, such as triggering preventive maintenance, scaling back on the intensity of use, and ordering replacement parts, as well as the timing of these events.
For our case study in 3D printing, we have implemented in-situ monitoring hardware for a fused deposition modelling (FDM) printer and have constructed a dataset for modelling the process. The dataset consists of in-situ observations (photographs) and select mechanical property measurements for 359 fabricated parts. With this data, we demonstrate the ability of machine learning methods to capture the complex dynamics of a 3D printing process. Specifically, we train a neural network-based model which is able to predict mechanical properties of the final product based on in-situ photographs as well as parameter information. Predictions made by these models can then be used to assess the quality of products as they are being fabricated, thereby making it possible to correct errors or to improve the expected outcome through online parameter adjustments.
About the Presenter:
Katie is a master’s student in the Department of Mechanical and Industrial Engineering at the University of Toronto, and is a member of the Centre for Maintenance Optimization and Reliability Engineering (C-MORE). Her research interests are in machine learning, including deep learning, computer vision, and reinforcement learning, with a focus on applications in additive manufacturing.
About the Co-presenter: Dr. Janet Lam holds a PhD in Industrial Engineering from the University of Toronto. She has been working in the field of maintenance optimization since 2008, with an emphasis on optimal scheduling of inspections for condition-based maintenance. More recently, her research interests have extended to machine learning approaches for maintenance and asset management. Through her work at C-MORE, she has applied academic research directly with industry partners, including those in mining, utilities, transportation, and the military. As the Assistant Director of C-MORE, Janet is involved with cultivating strong relationships with industry partners and developing maintenance engineering resources that are both useful and current.