DEVELOPING IN-SITU PROCESS MONITORING CAPABILITIES FOR MATERIAL EXTRUSION ADDITIVE MANUFACTURING

A machine agnostic framework for in-situ data collection during Material Extrusion (MEX) Additive Manufacturing (AM) builds with user-defined anomaly tagging is presented. To enable the use of Machine Learning (ML) algorithms for detection and identification of MEXAM build anomalies, a large set of training and test data is required. The tagging framework is integrated into a data collection system that includes infrared imaging, visible light imaging, accelerometer data, homography-based telemetry data, temperature, and environmental data. This data is registered both in time and 3D space, allowing the build anomaly data to be traced to a specific location on the as-built part. The presented framework allows users to create a database by identifying anomalies during a MEXAM build and automatically marks data around the anomaly time step across all collected sensor modalities. This tagged data can then be used as ground truth for ML training and validation.