Research Article
PrintProbe: A Deep Learning-Based Real-Time Error Detection System for 3D Printing Using Faster R-CNN
@INPROCEEDINGS{10.4108/eai.28-4-2025.2358076, author={P. Suryanarayana Reddy and C. Shyamala kumari and Singam Ram Charan}, title={PrintProbe: A Deep Learning-Based Real-Time Error Detection System for 3D Printing Using Faster R-CNN}, proceedings={Proceedings of the 4th International Conference on Information Technology, Civil Innovation, Science, and Management, ICITSM 2025, 28-29 April 2025, Tiruchengode, Tamil Nadu, India, Part II}, publisher={EAI}, proceedings_a={ICITSM PART II}, year={2025}, month={10}, keywords={3d printing deep learning faster r-cnn error detection computer vision additive manufacturing real-time monitoring}, doi={10.4108/eai.28-4-2025.2358076} }- P. Suryanarayana Reddy
C. Shyamala kumari
Singam Ram Charan
Year: 2025
PrintProbe: A Deep Learning-Based Real-Time Error Detection System for 3D Printing Using Faster R-CNN
ICITSM PART II
EAI
DOI: 10.4108/eai.28-4-2025.2358076
Abstract
3D printing has emerged as a transformative additive manufacturing technology, yet persistent challenges with print failures continue to impact production efficiency. Common errors such as filament failure, bed level misalignment, and layer shifting result in significant material wastage and increased production time. This paper presents an automated system for real-time error detection using Faster R-CNN, trained on a comprehensive dataset of 4,165 images. The system identifies four critical printing errors: bed level misalignment, layer shifting, no filament, and spaghetti extrusion. Through integration with a Raspberry Pi, the system provides automated print halting and user notification capabilities. Our experimental results demonstrate robust performance with a mean Average Precision (mAP@0.5) of 60-70% and an F1-score between 72-82%, establishing a reliable foundation for automated print monitoring and error prevention.
