Proposal of a Low-Cost Intelligent Fault Diagnosis System for 3D Printers

Additive manufacturing is a technology of Industry 4.0 widely used for various purposes. In the context of low-cost 3D printers, printing failures often manifest as a recurrent concern. These failures are typically diagnosed after a considerable amount of time has elapsed since the printing procedure, or upon its completion. This work proposes a method for the intelligent detection of failures in 3D printing, based on a low-cost accelerometer for vibration analysis. For each printing operating condition, the parameters were investigated, such as the printing speed, 3D printer bed leveling, filament extrusion temperature and bed temperature. A dataset was created based on data collected employing a widely used low-cost 3D printer. Two different algorithms for analyzing and classifying the samples were investigated, the Multilayer Perceptron (MLP) and Support Vector Machine (SVM). Intelligent fault diagnosis system developed presented good results, reaching until 92.3% of accuracy.