Quality monitoring method implementation for polymeric extrusion through supervised machine learning

Abstract

In the context of modern industry, additive manufacturing technology has been developed to enable productivity and quality gains in fabrication processes. In order to keep manufacturing effectiveness, machine learning techniques have been applied to assist with processes quality control. Thus, this project aims to create a quality monitoring method for a large single-screw extruder, which will be used for 3D printing of polymeric parts up to 1m3 of volume. For that, temperature, humidity, and acceleration sensors will be integrated to a fast-prototyping board. The collected data will be used as the input for supervised machine learning, with the objective of finding patterns that indicate failures in the extrusion operation.(AU)