Analysis of wagon failure data with clustering methods

Based on wagon fault records in Turkey, the destination maintenance workshop for wagons will be predicted using clustering methods by utilizing information on wagon type and fault type. Similarly, using wagon type and workshop information, fault type prediction will also be conducted using clustering methods. The clustering methods to be employed in this context include Self-Organizing Maps (SOM) and K-Means clustering. The SOM method visualizes datasets by representing high-dimensional data on a two-dimensional map, allowing for the clustering of data with similar characteristics. K-Means clustering, on the other hand, assigns data points to a predefined number of K cluster centers based on their proximity to these centers, grouping data points closest to these centers into the same cluster. The combined use of these methods will enable more efficient and systematic management of wagon maintenance processes. Moreover, accurately predicting fault types and directing wagons to appropriate workshops will expedite maintenance processes and reduce costs. Consequently, the operational efficiency of railway enterprises will increase, and maintenance planning will be conducted more effectively. This integrated approach aims to enhance the overall maintenance strategy, ensuring timely and cost-effective maintenance operations.