Annotation: At present, a number of subprocesses of the requirements management process are almost completely not automated. This is mainly due to the fact that the vast majority of regulatory documents (laws, standards, technical specifications, etc.) are presented in plain text format and are not directly perceived by technical devices of computer technology. Modern achievements in the field of information technology, especially such as the creation of machine-comprehensible content perceived by computers without human participation, make it possible to directly move to the automation of requirements management processes in full, i.e. the direct inclusion of the requirements of various regulatory documents in the general contour of planning, development, production and operation of various product samples from individual samples to complex distributed systems. The article proposes a variant of the reference structure of the automated requirements management system, i.e. the general structure on the basis of which specific decisions can be made.

Annotation: The article analyzes the conceptual apparatus of the draft Federal Law and some of its provisions, taking into account the accumulated experience in the development and use of artificial intelligence systems, considers the problems of technical feasibility of the provisions of the draft law and the sufficiency of the relevant regulatory and methodological support. Some proposals for the development of technical regulation in the implementation of the provisions of the draft law are formulated.

Annotation: An algorithm for constructing a regression equation for processing the results of bird monitoring in the study area in the form of a multidimensional polynomial by the least squares method is proposed. An interactive program Birds_MLS is developed in the Python algorithmic language, which provides the calculation of the equation coefficients using the NumPy, Pandas, SciPy, Matplotlib, and PyQt6 libraries. The program contains a full cycle of calculations, starting with the introduction of initial data, building a regression model, and ending with checking its validity and visualizing the results. The validity of the proposed model is assessed using the R^2 determination index and the F-Fisher criterion. The adequacy criteria are determined by comparing the known dependence of the u(x,y) function on two coordinates with the results of calculations obtained using the v(x,y) regression equation. The testing experiments confirmed the adequacy of the proposed algorithm when studying the dependence of population size on multiple variables in a multivariate setting.

Annotation: The article examines the current state of legal and technical regulation of intelligent transport systems (ITS) in the Russian Federation. The research is structured around four interrelated blocks: the current state of regulation, the impact of standardization on ITS development, prospects for standardization development, and a practice‑oriented model for its improvement. It is shown that regulatory legal acts form the strategic framework for the digital transformation of the transport industry, road safety, scientific and technological development, experimental legal regimes, and the deployment of highly automated vehicles. Standardization documents, in turn, translate these objectives into applied requirements for architecture, subsystems, integration platforms, data, testing, and operation of ITS. It is substantiated that the key challenge at the current stage is not the absence of a regulatory framework per se, but rather the insufficient integration of standards with the life cycle of local projects, procurement practices, performance assessment, and technological advancement. A practice‑oriented model for improving ITS standardization is proposed, based on a matrix of standards applicability, a reference architecture, requirements verification, testing, operational monitoring, and feedback to update the standardization program.

Annotation: Robotic process automation (RPA), which enables the creation and use of "software robots" to perform routine, repetitive tasks within the interfaces of conventional computer programs in digital systems instead of humans, is currently one of the most sought-after technologies for automating workflows in business and manufacturing. Developing methods for evaluating the performance of such systems is a highly relevant task, as the speed and efficiency of a robotic process automation system in processing and executing program-defined tasks affects both the performance of the entire information system and the capital expenditures involved in planning these systems, as well as the operating costs during operation. The effectiveness of robotic process automation technology depends not only on the efficient use of written software but also on external factors, such as the impact of the IT network infrastructure, packet network equipment, and the length of communication lines used to solve functional problems. This article discusses methods for evaluating the flow of traffic through robotic process automation systems in a common multiplexed data stream through packet switching nodes. Estimates of the impact of the bit rate of the multiplexed stream on the RPA service are calculated. The potential of using the geometric probability method to estimate the timing characteristics of real-world bit traffic passing through packet switching nodes is demonstrated. The impact of delays occurring at network junctions with different NTP servers is also discussed. The causes of delays are analyzed, and a method for their estimation is proposed.

Annotation: Matrix transformations form the foundation of modern data mining and machine learning algorithms. The relevance of the study stems from the necessity to enhance the skills of novice programmers in software implementation of matrix transformations, as well as to strengthen their software testing skills for the early detection of defects and errors, in compliance with the guidelines established in GOST R 56920–2024 and GOST R 56939–2024. The aim of the research is the automation of the generation and assessment process of exercises on the software implementation and testing of matrix transformations. The scientific novelty of the method proposed in this paper consists in the synergistic combination of the following components: a term rewriting system with constraints tailored to the specific nature of matrix transformations; mutation analysis based on prepared programs with injected mutations used to evaluate the quality of students’ testing functions; interpretation of expression trees with both full and reduced precision of matrix operations to generate cross-platform test suites. The application of the proposed method in the educational process freed programming instructors from the burden of manually checking 12269 programs submitted by students over the course of a single semester. Incorporating mutation analysis as a distinct stage of the automated assessment pipeline has improved the robustness of exercises against solutions generated by large language models using the zero-shot approach, achieving an 80% resistance rate. Furthermore, it increased the absolute rejection rate of programs by 33.5% when compared to the automated assessment algorithm operating without mutation analysis. As a result of students’ work on enhancing the quality of their testing functions in initially rejected programs, the average number of assert statement executions in accepted programs increased by a factor of 2.24. The time required to generate unique exercises for over 1400 students using a software tool implemented in Python does not exceed several minutes, every exercise consisted of 5 tests for automated assessment and 3 modified programs for mutation analysis. The average time consumed for checking and mutation analysis of a single program within the sandbox environment, implemented as an isolated Docker container, is 8.4 seconds. The obtained results can be used in academic courses, in open online courses, and in programming competitions.

Annotation: Regular collection of publications from open mass media sources is complicated by the heterogeneity of source interfaces, the volatility of page markup and the need to attest the provenance and immutability of the accumulated corpus. The aim of the study is to design and experimentally verify an application architecture in which adding a source does not affect the program core, the completeness of periodic feed polling is monitored under a variable publication rate, and the state of the collected corpus remains verifiable. A combined architecture is proposed, in which coarse-grained functional modules are complemented by a coordination layer formed by a message broker with durable, at-least-once delivery and a distributed cache of short-lived state. The application combines a versioned declarative source template with reparsing of stored responses, adaptive feed polling based on overlapping publication windows, and verifiable corpus attestation using a Merkle tree and signed checkpoints. These mechanisms form a reproducible sequence linking source responses, extraction rules and normalised records. The experimental implementation was compared with Scrapy at matching process counts. With one to eight workers, the ratio of mean batch throughputs ranged from 1.70 to 2.00. A separate experiment used a shared request-rate limiter to characterise operation under a constrained input flow. Stored HTML pages from two news resources were subjected to a controlled local markup transformation and parsed using revised templates. No additional HTTP requests were sent during reparsing; the non-empty field values obtained with the original template were recovered without differences. Message redelivery occurred in the worker, RabbitMQ and Redis failure tests, but each of the three episodes ended with all unique results stored. With one million records in the tree, the mean inclusion-path size was 640 bytes and verification took 24.50 microseconds. The reported timings were obtained on one node for the stated software versions, using the measurement intervals specified for each experiment.

Annotation: The paper investigates the influence of clarifying instructions and semantic feedback on the quality of alignment between program models of standardization objects generated by large language models. The problem is that independently generated models may be syntactically correct and semantically close, yet difficult to align because of differences in the naming of entities, roles, states, relationships, and units of measurement. Unlike the preliminary validation on a small number of documents, the present study was conducted on a corpus of 32 RFC documents, six generative models, and five independent iterations for each combination of document, model, and scenario. In total, 2,880 models were processed; 14,400 pairs were formed from them, and 85,400 alignment operations were performed using two COMA-like methods: valentine and bdikit. The results show that general instructions on semantic naming and code structuring improve the relative alignment quality by 46.2%, whereas semantic feedback worsens the result, improving quality by only 40.3% relative to the control condition. This finding should be taken into account when developing tools for formal model generation, static analysis, model alignment, and the preparation of intermediate representations for the analysis of technical and regulatory documents.

Annotation: This paper inaugurates a series of research aimed at investigating configuration variants of pipeline special-purpose computing systems employing dedicated hardware blocks. The objective of the research is to extend the applicability of the system-level approaches previously developed by the Laboratory of Special-Purpose Computing Systems at RTU MIREA for constructing logical and topological representations of pipeline special-purpose computing systems utilizing dedicated hardware blocks embedded in field-programmable gate arrays (FPGAs). The relevance of developing specialized computing devices for tasks requiring minimal response time and high performance is discussed. The paper reviews the fundamental concepts and principles underlying the construction of logical and topological representations of special-purpose computing systems, as well as the issues of ensuring consistency between different representation variants while preserving the intended design characteristics. System-level approaches developed by the Laboratory of Special-Purpose Computing Systems at RTU MIREA for constructing representations of pipeline computing systems are presented. The applicability of these approaches to the design of pipeline special-purpose computing systems based on FPGA dedicated hardware blocks is examined. Particular attention is paid to Digital Signal Processing (DSP) blocks, which are widely used to implement multiply-accumulate operations, digital signal processing algorithms, and neural network computations. The architectural features of the DSP48E1 block and the specifics of its placement within AMD Artix-7, Kintex-7, and Virtex-7 FPGA families are analyzed. An automated framework has been developed to generate special-purpose computing system configurations, construct their topological representations, and obtain detailed implementation reports. Experimental results demonstrate the influence of pipeline depth on the achievable operating frequency of the special-purpose computing system, as well as significant differences in critical path delay across different representation variants. The paper concludes by outlining directions for further research.