~~~Transport and transport-technological systems of the country, its regions and cities, organization of production in transport~~~
In connection with the traceable global trend of increasing average speeds in railway transport, more and more often they are trying to exclude the intersection of railways and roads. However, in the Russian Federation, only 40 % of railway crossings are located near settlements or on busy roads, the construction of two-level interchanges at the remaining crossings is not economically profitable, and sometimes difficult due to climatic conditions. Therefore, a different approach is needed to prevent incidents and accidents at crossings. At the same time, the need for an individual construction of a security strategy for each object of the transport infrastructure should be taken into account. At present, in the existing information systems of the Ministry of Transport, the Ministry of Internal Affairs, as well as the owners of the railway infrastructure, a fairly large amount of statistical information has been accumulated on incidents and traffic accidents at railway crossings. But most often these data are used for certificates containing absolute values in each of the departments. At the same time, the relationships between these disparate sources of information are not comprehensively analyzed. Modern technologies for creating intelligent information systems based on existing large databases and tools such as machine learning, intelligent search, neural networks make it possible to build an expert system based on a specific set of rules that can analyze the situation at a particular crossing, assess the likelihood of emergency situations and suggest ways to prevent them. In addition, it is important to analyze not only the infrastructure, but also specific units of vehicles and the accident-free performance of the drivers driving them. This article discusses the prerequisites and technology for building a knowledge base of an intelligent system used to analyze the level of safety at railway crossings, which will allow us to combine currently information about the factors accumulated in various systems that affect traffic safety at crossings and based on the accumulation of knowledge give a forecast of the development of the situation at each specific object.