Real-Time Defect Detection and Intervention System for FDM Printers Using Accelerometer Data

Kemel Shomenov, Md. Hazrat Ali, Essam Shehab

2025

Closed access Metadata (CSL-JSON)
DOI 10.1109/icspc68261.2025.11326218
Data confidence 70 medium
How this score is calculated

It is built from checkable properties of the record itself: whether the article has a DOI, how many independent sources confirmed its data, whether the required fields are filled in, and how recent the latest confirmation is. Disagreement between sources and a missing identifier lower it. A value of 80 or above means the record was published without further checking; 60 to 79 means it is published at reduced confidence, with a manual check advisable. A record verified by an editor by hand is never pushed below 90 by a recalculation. This scores our data about the article, not the article itself.

Abstract

Fused Deposition Modeling (FDM) 3D printers are widely used for prototyping and small-scale production; however, they require continuous supervision to prevent defects, which can lead to materials and time wastage. Excessive vibrations during the printing process often cause such defects. Therefore, it is essential to develop a vibration monitoring mechanism. Although prior research studies have focused on employing vibration data and machine learning to detect problems such as nozzle clogging and filament jams, little research attention has been given to real-time, automatic corrective measures. This study focuses on developing an economical, autonomous defect detection system based on accelerometer data. The system monitors vibration patterns in real-time and applies the Fast Fourier Transform (FFT) to analyze frequency domain data. Upon detection of defects, the system, which, implemented using Python, automatically pauses or stops the printer. The defect identification accuracy is enhanced using Cosine Similarity, which compares real-time data with the normal state of the printer. Experimental results indicate that the system can effectively detect changes in vibration patterns, such as those caused by loose belts, and implement immediate corrective actions. This autonomous approach significantly reduces material waste and improves the overall reliability of the 3D printing process. This research study contributes to the field by integrating defect detection with real-time interventions, providing a practical solution for improving the efficiency of FDM printing.