A novel machine learning system for early defect detection in 3D printing

This paper discusses a comprehensive study to develop a machine learning model for detecting unwanted vibrations during the 3D printing process. Undesired vibrations can significantly degrade print quality, leading to defects such as void formation, poor surface quality and improper layer bonding. Identifying and mitigating these vibrations is essential to ensuring the reliability and precision of 3D printed products, which is particularly crucial in sectors such as healthcare, automotive, and aerospace. The study introduced a novel system with an inertial measurement unit (IMU) mounted on the printer head, which records acceleration and angular velocity in three axes. The data is transmitted to a microcontroller and then to an acquisition device that controls a controlled vibration generator. The collected information formed a dataset for training and testing various machine learning models. Of all the models evaluated, the Dense Neural Network (DNN) showed the highest performance in accurately distinguishing normal print vibrations from unwanted vibrations. The study underscores the critical importance of early defect detection, which saves time and reduces costs, being essential for the widespread adoption of incremental manufacturing technology. Early identification of defects enables immediate intervention and correction of errors before they become serious defects affecting the quality of the final product. This is particularly important in the context of increasing automation and optimization of manufacturing processes.

---

Opracowanie rekordu ze środków MNiSW, umowa nr POPUL/SP/0154/2024/02 w ramach programu "Społeczna odpowiedzialność nauki II" - moduł: Popularyzacja nauki (2025).