Sustainable Energy and Artificial Intelligence

Real-Time Deep Learning-Based Surface Defect Detection for Crankshaft Journals: Towards Zero-Defect Automotive Manufacturing

Articles in Press

Document Type : Original Article

Authors

Ata Ahmadabadi Asle Alamdari 1
Alireza Yazdanijoo 2
Seyed Abdullah Mousavi 3
Ashkan Moosavian 4
Amirhasan Kakaee 5

1 Faculty of Automotive Engineering, Master's degree, Iran University of Science and Technology, Tehran, Iran

2 Faculty of Automotive Engineering, Master's Degree, Iran University of Science and Technology, Tehran

3 Department of Mechanical Engineering, Technical and Vocational University (TVU), Tehran, Iran

4 Department of Agricultural Engineering, Technical and Vocational University, Tehran, Iran

5 School of Automotive Engineering, Iran University of Science and Technology, Narmak, Tehran, Iran

Abstract
Surface defects on crankshaft journals, such as micro-scratches and pitting, can significantly degrade engine performance and initiate catastrophic failure. This study presents a real-time, vision-based inspection system for detecting surface defects on stationary crankshaft journals under severe data scarcity constraints. A high-resolution dataset of real journal surfaces was acquired using a controlled optical setup to minimize specular reflections. Pixel-level defect localization was performed using a U-Net semantic segmentation model with a MobileNet backbone, trained with Categorical Focal Loss and advanced data augmentation to address class imbalance and limited samples. The MobileNet-based model achieved a classification F1-score of 88.9%, demonstrating effective learning without overfitting. For segmentation, the system attained a mean IoU of 71.5% without augmentation, improving to 81.7% with augmentation. The optimized model was deployed on an NVIDIA Jetson Nano at a bearing assembly station, validating its suitability for real-time, in-line industrial inspection. The proposed framework provides an efficient and scalable solution for automated quality control, advancing zero-defect manufacturing of critical engine components.

Keywords

Fault Detection
Internal Combustion Engine
Main Journals
Machine Vision
Convolutional Neural Network
Subjects
Sustainable Energy and Artificial Intelligence

Articles in Press, Corrected Proof
Available Online from 15 September 2026

  • Receive Date 10 June 2026
  • Revise Date 19 July 2026
  • Accept Date 23 August 2026
  • First Publish Date 15 September 2026

Home

Guide for Authors

Submit Manuscript

Reviewers

Contact Us