Innovation and searching for new solutions is a key element of continuous improvement of railroad undertakings' safety management systems. Therefore, supporting the existing systems with IT solutions gives wide and new possibilities for presentation and analysis of data collected in different areas of railroad companies' operation. Such a situation occurs, among others, in the case of information on events and incidents occurring on the railroad network. Graphical presentation of data, which so far were stored mainly in tabular form, creates entirely new potential for their further analysis and use. The map of railroad events should be gradually implemented into various processes of safety management systems, especially in aspects related to training and current information of employees about safety-relevant locations or exchange of information with external entities.

The problem of railroad incidents and events gives wide opportunities related to data analysis, both for individual railroad undertakings and for all railroad market entities operating on Polish railroad network. Detailed research should be conducted on basic data about events, i.e. location, time, weather conditions, direct causes of events, as well as on detailed issues related to the influence of human factor on occurrence of dangerous situations or local conditions at the place of event. Conducting scientific research in this area should be reflected in the future in the possibility of predicting the place of occurrence of railroad events, their potential effects, factors having the greatest influence on them, and consequently changing the approach to the training process of employees working at railroad stations or building public awareness of the interaction of various transport systems.

Due to the large number of collected, detailed data on railroad incidents occurring on PKP PLK S.A. railroad network, which in 2013-2021 were registered more than 13 500, due to their repetitive nature, as well as the closed system of railroad transport limited to a finite number of railroad lines with different traffic volumes, level crossings, stops and stations, etc., it is possible to develop a predictive model, including with the use of machine learning, the implementation of which can really affect the improvement of railroad transport safety, allow preventing the occurrence of railroad incidents and limit their consequences.