Modern railway systems consist of a huge number and bride spectrum of components that generate a large amount of data. This data will be collected from heterogeneous sources and will be stored in Data Lakes. The collected data is used in many important tasks for railway systems operation, modeling, symbolic description, failure detection, and many more. Standard Data Lakes are rigid and could not be quickly adapted to constantly evolving railway systems. Moreover, the Data Lakes are often huge and complex, and even for domain experts, it is difficult to understand the meaning of the data and find interconnections in it. The main goal of this work is to add a semantic level to the Railway System's Data Lakes. This will be done using Knowledge Graphs, which are constructed from raw data, its metadata, and a railway systems ontology. An ontology is a formal description of a specific domain t railway systems in our case. Knowledge Graphs will enable the more flexible integration of new data sources, and will provide a better insight into the collected data.