The strategic development of Russian Railways (JSCo «RZD») is inextricably linked with the introduction of robotic complexes and intelligent control systems. The article substantiates the need for a transition from traditional, often reactive, management methods to an end-to-end digital methodology covering both the operation of marshalling yards and the polygon as a whole. The proposed approach focuses on the use of «windows of opportunity», which make it possible to form long-distance through and two-group trains, bypassing unnecessary reprocessing. This can radically reduce operating costs, shorten wagon turnover, and release significant internal reserves of the network. The key tool for implementation is the integration of artificial intelligence systems and big data analytics for predictive management and achieving synergy between digital stations and the control polygon. The proposed approach will ensure an increase in the share of self-formed through trains operating in long-distance service without reprocessing at intermediate marshalling yards. This will preserve the effect of the introduction of intelligent train formation algorithms within the «Digital Railway Station» system. In turn, the optimization of wagon flow routes across the railway network polygons throughout the entire transportation route helps to reduce overhead costs and minimize losses in the ratio of operational and tariff freight turnover.

The article considers the urgent problem of ensuring timely delivery of container cargo under infrastructure constraints on the Russian Railways network. A systematic approach to the organization of combined rail-road transportation with dynamic re-routing and reloading of cargo en route to bypass congested infrastructure sections is proposed. Based on the formalization of the route network as a weighted graph and the use of optimal route search algorithms (Prim, Floyd-Warshall), a model for choosing a rational delivery method that minimizes total delivery time and costs is developed. Special attention is paid to the conditions for making decisions about reloading before entering the limiting section, which allows avoiding downtime and fulfilling «just-in-time» obligations. For practical implementation of the model, the main categories of data necessary for building an adequate information base are identified.

This article examines the formation and transformation stages of intelligent transport systems on public roads in the Russian Federation, considering changes in the regulatory environment, technological priorities, and equipment structure. The study aims to identify patterns in intelligent transport systems evolution and factors driving the shift from local solutions to integrated digital traffic management platforms. The research included quantitative and qualitative analysis of technical means at different development stages, analysis of the domestic-foreign equipment ratio dynamics, and assessment of import substitution policies, compatibility requirements, and certification impacts on system architecture. Particular attention is given to federal and industry regulations shaping the current intelligent transport systems configuration. The results clarify trends in hardware optimization, the growing role of software and platform solutions, and the development of VATS and V2X technologies. The practical significance lies in applying the identified patterns to regional intelligent transport systems planning, defining equipment requirements, supporting investment decisions, and developing road industry digital transformation programs.

The main directions of technical and technological development of Russian railways for the period of about 40-50 years are considered. During this period, railways should significantly improve the key performance indicators of their work. The selection of promising areas for technical and technological development of Russian railways is based primarily on the use of the achievements of the Fourth Industrial Revolution, which began in the 1920s, as well as the already known, highly effective, but not yet implemented solutions of previous revolutions. It is necessary to study in depth the new opportunities generated by the 4th industrial revolution for the development of technical means and technologies in relation to the railway transport industry, infrastructure, rolling stock, traffic control systems, and transportation processes. The article discusses the main areas in which the 4th industrial revolution is developing and provides suggestions for their use, taking into account the specific features of railway operations.

The presented article describes an adaptive algorithm for processing measurements of sensors of inertial navigation systems of unmanned aerial vehicles. Its verbal and analytical descriptions are presented, as well as a comparison with existing algorithms: exponentially weighted moving average (software low-pass filter) and α-β filter. The novelty of the work lies in the development of an adaptive filter, which, unlike the well-known adaptive moving averages (for example, the Kaufman adaptive moving average), where the smoothing coefficient is determined by volatility or other statistical characteristics of the signal, uses the averaged (by means of an exponentially weighted moving average) difference of the measured signal values to dynamically adjust the filtering parameter. Additionally, automatic delay compensation is implemented through an exponentially smoothed error between the filtered and the original signal. The relevance of the work lies in the fact that the proposed algorithm ensures high accuracy of signal smoothing at a high response rate to signal changes, which is especially important for controlling maneuverable unmanned vehicles.

This article examines the problem of assessing the reliability of traffic incident detection subsystems (TIDS) based on computer vision technology. It is noted that the current regulatory framework of the Russian Federation, including GOST R 71159-2023, does not contain quantitative criteria for assessing the effectiveness of incident detection, which complicates the validation of systems and the comparison of solutions from different manufacturers. Internal «reliability» parameters in software systems cannot be independently verified, and the accuracy of spatial and temporal referencing depends on numerous factors that reduce the reliability of event recording. The purpose of the article is to determine key metrics for assessing the functional characteristics of the TIDS and justify their threshold values based on an analysis of technological capabilities, operational requirements and international experience. As a result of this work, a system of complementary metrics was developed, including detection completeness, detection accuracy, classification errors, and false positive rate. Criteria were also proposed, recommended as a methodological basis for developing regulatory and technical requirements for TIDS.

The article presents a methodological approach to regular monitoring of peripheral equipment and software of intelligent transport systems (hereinafter referred to as ITS) deployed on public roads in the Russian Federation. The relevance of the study is driven by the absence of a unified and reproducible mechanism for collecting and processing information on ITS elements (peripheral equipment and software), which results in fragmented datasets and complicates managerial decision making related to the development of digital road infrastructure. An organizational and methodological approach to systematic monitoring is proposed, covering the full cycle of primary statistical data management: formation of a respondent base, development of standardized survey forms, organization of methodological support for monitoring participants, and procedures for processing primary data. Particular attention is paid to ensuring the completeness, reliability, and comparability of information provided by ITS owners and operators. Within the proposed approach, procedures for data normalization, record deduplication, and format and logical validation are developed to minimize the influence of the human factor and improve the quality of analytical results. The proposed approach ensures the reproducibility of monitoring processes and can be used as a tool for regular observation of ITS infrastructure, as well as a basis for public procurement planning, strategic planning, and the establishment of statistical observation in the field of ITS.