I conducted this research project in the University of Pennsylvania’s Architected Materials Laboratory under Prof. Jordan R. Raney. View my UPenn Fall Research Expo poster here.
My research project investigates the application of convolutional neural networks to additive manufacturing. Combining machine learning and computer vision for intelligent, multi-material 3D printing allows printers to monitor uncommon events and defects, evaluate structure and material properties, and improve design printing strategy.
In Summer 2023, I received the competitive Penn Undergraduate Research and Mentorship (PURM) Program Award, providing me with funding and support to continue this project. To collect and process live feed from the printer, I am using an iDS Imaging camera combined with Python libraries OpenCV and Pyueye. For the convolutional neural network, I am using Keras and Tensorflow for multiclassification, specifically to recognize errors caused by low/high extrusion pressure and low/high distance from the nozzle.
About the Architected Materials Laboratory
New manufacturing methods such as 3D printing have enabled an unprecedented degree of control over the properties of materials. This has the potential to revolutionize materials properties and to thereby greatly expand the design space available to engineers. How can we leverage these tools to produce materials and structures with unprecedented engineered capabilities?
At the Architected Materials Laboratory we explore this question, with our work organized into three major Thrusts: (1) Geometric control of nonlinear behavior; (2) 3D-printable composites and responsive materials; and (3) Mechanical logic and autonomous materials.
Research Pitch
Live Demo