Vol 76, No 1 (2026)

This article examines the problem of dispatchers informed and prompt assignment of emergency response teams to respond to accidents. Existing approaches to dispatcher decision support typically focus on assessing the dispatcher's state and cognitive load. These approaches fail to consider the repair team as a single control entity, and the quality of the emergency response team assignment decision is only indirectly described. This paper proposes a three-dimensional ergatic model of «Human-Equipment-Logistics», formalizing the repair team as an integrated system. A team operational readiness index is introduced, a computational experiment is conducted simulating three typical team operating modes, and threshold values for the index are selected. It is demonstrated that the resulting indicator can be used in decision support systems to formalize the procedure for permitting teams to work and improve the transparency of emergency response team assignment.

This paper analyzes statistical indicators for the development of the agro-industrial complex for 2022-2024, identifies divergent trends in production, and examines the structure, functionality, and growth dynamics of the «Svoe» ecosystem, including expansion of product range, geographic coverage, and trade volumes. Particular attention is paid to the platform's impact on the socioeconomic development of rural areas: increasing producer incomes, diversifying the economy, reducing population outflow, and promoting sustainable farming and agritourism. Promising areas for further ecosystem improvement are identified, including deeper integration with government systems, the implementation of AI, export opportunities, and a focus on remote regions.

The article examines the problems of electric power development that have emerged in recent years, both as a result of the energy transition and in connection with the increasingly complicated international situation. The Republic of Cuba is taken as an example, where the energy situation is becoming more difficult day by day. Proposals are formulated for possible solutions to Cuba's energy problems with the help of Russia. The authors also emphasize that an unbalanced energy transition from traditional energy to the decarbonization of energy declared in recent years, and an unjustified increase in the use of renewable energy sources, could threaten the achievability of the stated goals. The conclusion is made on the need for serious analysis and use the principles of a systems approach when assessing the proclaimed changes in the structure of electricity generation, as well as when determining the potential of nuclear energy.

The problem of optimal control of the final state of a system is one of the simplest and constitutes the core of any other optimization problem. This paper considers the terminal problem of pursuing an evading object while measuring its distance and velocity in the absence of information about its possible trajectory, i.e., in the absence of a description of its dynamics. The problem statement includes a mathematical description of the object, constraints imposed on control actions, and a quadratic quality functional. The necessary optimality conditions in the problem of synthesizing the corresponding controls are written as the canonical Euler-Lagrange system with the corresponding boundary conditions. Synthesizing the corresponding controls involves finding solutions to boundary value problems, which is typically accomplished using numerical methods. This paper suggests an alternative way to solve two-point boundary value problems based on the assumption of the validity of the inverse principle of optimality by R. Bellman.

Transformer models are able to achieve state of the art results across multiple NLP tasks; however, there are still significant challenges in deploying them due to high inference latency and hardware costs. We empirically characterize transformer encoder inference across GPU generations based on over one thousand controlled measurements of DistilBERT, BERT-base, and BigBird-RoBERTa on NVIDIA T4, A10G, and L40S. Our analysis demonstrates that architectural efficiencies are consistent across models and that throughput-performance structure is smooth, low dimensional, and consistent across hardware generations. Utilizing these regularities, we develop a two-point cross-GPU scaling model to predict full throughput-sequence length curves using only a compute-dominated and a memory-dominated measurement. Prediction error decreases with batch size: from 12-13% MAPE at batch size 16 to 6% at batch size 256. The framework allows forecasting throughput, selecting models, planning hardware needs, and minimizing benchmarking on new accelerators.

Cervical carotid artery segmentation in contrast-enhanced CTA supports stenosis assessment and vessel geometry analysis, but routine scans are challenging due to noise, variable enhancement, calcified plaque, and complex bifurcation/skull-base anatomy. This paper presents a semi-automatic approach that combines multiscale Hessian vesselness with slice-wise mask propagation and local intensity-based refinement. Segmentation is initialized from a small set of manual lumen contours, propagated axially using vesselness thresholding and connected-component selection with simple area/centroid constraints, and refined by HU-based region growing within the vesselness neighborhood. Tested on 10 CTA scans, the method tracks the CCA and ICA over several centimeters in most cases, with generally stable behavior at the bifurcation and only occasional switching to adjacent vessels. Across the 8 cases without ICA/ECA switching, the segmentation achieved Dice = 0.925 ± 0.015 and HD95 = 1.13 ± 0.21 mm.