Abstract
Rail transport is a key component of modern traffic and transportation infrastructure. In both freight and passenger transport, the quality and reliability of rail products are of crucial importance, as they make a significant contribution to the safety, durability, and economic efficiency of railway operations. This thesis investigates the occurrence of surface defects in the rail rolling process at voestalpine Rail Technology GmbH. The focus is placed on scab-related defect patterns in switch rails. As part of the study, the rail rolling process was systematically analyzed and evaluated with regard to potential causes of failure and their effects by means of a process FMEA in accordance with VDA/AIAG 2019. The results show that material-related pre-damage in the semi-finished material, temperature-dependent influences, and mechanical deviations in the Break-Down Mill and Ultra-Flexible Rolling process steps are the main contributing factors to scab formation and the resulting surface defects. In addition, it became evident that many of these defects can only be detected to a limited extent during the ongoing process. This results in a higher priority for the implementation of corrective measures. Based on these findings, a possible concept for the integration of an AI-supported camera system was developed. This concept includes three defined image acquisition points along the process and enables continuous visual documentation from the bloom to the finished rail. The aim of the system is to identify a correlation between the cut surface of the blooms and the later formation of scabs, as well as to improve defect detection during the production process. The results of this thesis show that the use of image-based analysis systems offers promising potential for the early identification of defect causes. In the long term, such a system can contribute to improving process stability, product quality, and efficiency in the rail rolling process.
| Translated title of the contribution | Process analysis and concept development for AI-based surface defect detection in the rail rolling process at voestalpine Rail Technology GmbH |
|---|---|
| Original language | German |
| Awarding Institution |
|
| Supervisors/Advisors |
|
| Award date | 26 Jun 2026 |
| Publication status | Published - 2026 |
Bibliographical note
embargoed until 20-05-2031Keywords
- rail rolling
- Failure Mode and Effects Analysis (FMEA)
- surface defects
Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver