MKMaxim Kosei

Maxim Kosei

*This is an algorithmically generated author record based on publicly available information.

Maxim Kosei has a total of 7 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Comparative analysis of UAV simulation platforms: challenges, opportunities, and future directions

    by Andrey Kupin, Yuriy Osadchuk, Maksym Kosei

    Unmanned aerial vehicles (UAVs) are deployed across agriculture, disaster response, infrastructure inspection, and military operations, but flight testing is expensive, hazardous, and slow, which has made simulation the default environment for developing and validating UAV algorithms. Dozens of simulation platforms exist, and a substantial secondary literature already compares them; what is missing is a consolidated view across those comparisons. This paper reports a meta-review of nine studies published between 2018 and 2025: six comparative surveys of simulators and agent-based frameworks, two simulator-development papers, and one review of UAV airframes and autonomy components. For each study we reconstruct its scope, the platforms it covers, and the evaluation criteria it applies, and consolidate the results in a single comparison. The surveys converge on a common set of evaluation axes – sensor support, physics fidelity, scalability, usability, and maintenance activity – but weight them differently and share no common benchmark, so their verdicts cannot be compared across publications. Coverage is highly uneven: Gazebo is assessed in five of the nine studies and most other platforms in only one. We argue that the principal gap in UAV simulation is therefore not platform capability but the absence of a shared benchmarking protocol, and set out what such a protocol would have to specify.

  • Architecture of intelligent system for webservices scaling

    by Andrey Kupin, Dmytro Zubov, Maxim Kosei, Vladyslav Holiver

    The key aspects of web services development, administration structures, and the use of machine learning technologies for server optimization are explored. The tendencies of web services development, scaling options, importance and basic concepts of microservice architecture are considered. The article highlights the general principles of artificial intelligence, machine learning, and deep learning and their impact on the functionality of web services. To enhance the operation of web services, an architecture of an intelligent system for automatic scaling is presented and machine learning algorithms with increased reliability are elaborated. The article optimizes the performance of such a system. Methods for detecting abnormal system behavior are proposed, which allows preventing failures or a decrease in overall performance.


  • Models and Technologies for Autoscaling Based on Machine Learning for Microservices Architecture

    by Kryvyi Rih National University, Serhiy Semerikov, Dmytro Zubov, University of Central Asia, Andrey Kupin, Maxim Kosei, Vladyslav Holiver