A Low-Cost Multi-Sensor Data Acquisition System for Fault Detection in Fused Deposition Modelling

Area of Science:

• Additive Manufacturing
• Quality Control
• Sensor Technology

Background:

• Fused Deposition Modelling (FDM) is a key Industry 4.0 technology offering benefits like rapid prototyping and cost-effectiveness.
• However, FDM 3D printing is prone to defects, posing challenges for real-time fault diagnosis.
• Effective monitoring requires robust data acquisition and analysis systems.

Purpose of the Study:

• To develop a low-cost, multi-sensor data acquisition system (DAQ) for detecting faults in FDM 3D printed products.
• To analyze sensor data under various fault conditions to improve defect detection.
• To implement a machine learning model for accurate classification of normal and faulty printing conditions.

Main Methods:

• A multi-sensor DAQ system was built using an Arduino microcontroller, integrating vibration, current, and sound sensors.
• Time and frequency domain analyses were performed on sensor data, with feature selection using the chi-square method.
• A Convolutional Neural Network (CNN) model was trained for fault classification, utilizing K-means for data clustering and normal condition thresholds derived from normal distribution curves.

Main Results:

• The developed DAQ system successfully captured real-time multi-sensor signals during FDM printing under various fault conditions.
• Feature vectors were created, and significant features were identified for training the CNN model.
• The CNN model achieved approximately 94% accuracy in classifying normal versus fault conditions, with performance evaluated using recall, precision, and F1 score.

Conclusions:

• The proposed low-cost multi-sensor DAQ system is effective for real-time fault detection in FDM 3D printing.
• The combination of sensor data analysis and CNN classification provides a reliable method for quality control.
• This approach contributes to enhanced monitoring and defect diagnosis in additive manufacturing processes.