Original scientific paper
https://doi.org/10.7307/ptt.v37i3.791
Visual and Statistical Methods for Analysing Irregularity Data in Rolling Stocks – An Application on Turkey’s Freight Wagons Fleet
Ömür AKBAYIR
orcid.org/0000-0002-8747-4238
; Eskisehir Technical University, Vocational School of Transportation
Serdar BENLİGİRAY
orcid.org/0000-0001-7593-7971
; Anadolu University, Faculty of Business Administration
Ahmet ONAY
; Eskisehir Technical University, Vocational School of Transportation
*
* Corresponding author.
Abstract
Maintaining freight wagons is an essential operational process and a significant cost factor for rail transport companies. Analysing detected irregularities in the freight wagons offers distinctive and valuable insights for planned maintenance. The primary purpose of this study is to provide various techniques to shed light on the characteristics of these irregularities and identify any interrelations between them. This study also reveals the general characteristics of Turkey’s freight wagon fleets and maintenance depots in relation to the detected irregularities. New-generation visualisation tools, such as heat maps and chord diagrams, were utilised in this study. To determine the relationships between pairs of irregularities based on wagon type, irregularities with a high co-occurrence rate were identified and tetrachoric correlation analyses were conducted. In the final stage, Stepwise Poisson Regression Models were tested to explain the irregularities for each wagon type. The analysis techniques exemplified in this study were proven to reveal many interrelations between irregularities. The methods proposed in this study have the potential to provide crucial information for maintenance planning, parts supply and wagon repair processes. However, their practical application requires careful interpretation and detailed consideration by expert railway managers and engineers.
Keywords
freight wagon; irregularity; maintenance; data visualisation; tetrachoric correlation; Poisson regression
Hrčak ID:
331759
URI
Publication date:
5.6.2025.
Visits: 229 *