M.A. Kazantsev1, M.A. Radchenko2

1-2 Joint-Stock Company «Special-Purpose Enterprise «Radiosvyaz» (Krasnoyarsk, Russia)

1 mkaz@mail.ru, 2 maks.engineer@mail.ru

The digital transformation of industry and the increasing regulatory pressure regarding personnel safety and Critical Information Infrastructure (CII) protection necessitate a shift from passive video surveillance to active, intelligent systems. The core problem lies in selecting an optimal video analytics architecture that balances reliability, compliance with regulatory requirements (including import substitution and CII certification), and sufficient flexibility for adaptation to specific production processes. This study aims to conduct a comprehensive analysis of modern video analytics systems used in industrial settings, focusing on three leading Russian developers: TRASSIR (LLC «NPP Trassir»), Macroscop (LLC «Macroscop Trade»), and VIPAKS (LLC «Vipaks+»). Furthermore, the paper provides a comparative assessment of these commercial solutions against those developed by independent (free) developers using open-source technologies. The research methodology included a detailed analysis of official technical documentation, a comparative evaluation of functional modules, a review of open-source frameworks (OpenCV, TensorFlow, PyTorch, YOLO), and an examination of the Russian r egulatory landscape governing CII, particularly the requirements of the Federal Service for Technical and Export Control (FSTEC) and the Ministry of Digital Development, Communications and Mass Media of the Russian Federation (Ministry of Digital Development). Pilot projects were conducted at JSC «NPP «Radiosvyaz» to validate the practical performance of TRASSIR and VIPAKS solutions. The results demonstrate that each commercial platform occupies a distinct niche: TRASSIR excels in large -scale, integrated projects requiring import substitution compliance; Macroscop is ideal for flexible integration into existing infrastructure with a foc us on personnel safety (e.g., Personal Protective Equipment (PPE) detection); and VIPAKS is optimal for high-risk facilities due to its specialized anti-terrorist modules and high -speed fire detection (claimed 98% accuracy, 10 –17 s response time). A critical finding is the syste matic inapplicability of open -source solutions on CII objects, primarily due to the lack of mandator y FSTEC certification, absence of developer security clearances, and insufficient legal liability. The practical significance of this work lies in providing methodological guidelines for selecting a video analytics system based on production type, risk level, and regulatory constraints. The findings are directly applicable to the design and modernization of security systems at industrial, energy, and logistics facilities in Russia and other jurisdictions with similar regulatory frameworks. The study concludes that while open-source offers unparalleled flexibility for bespoke tasks, certified commercial platforms remain the only viable option for CII, with VIPAKS emerging as the preferred choice for high-risk industrial environments based on pilot project validation.

Kazantsev M.A., Radchenko M.A. Application of video analytics syste ms in industrial environments. Dynamics of complex systems. 2026. V. 20. № 1. P. 62−69. DOI: 10.18127/j19997493-202601-06 (in Russian).

1. Trassir, OOO «NPP». Videoanalitika [E`lektronny`jresurs]. URL: https://trassir.ru/products/programmnoe_obespechenie_trassir/ trassir_videoanalitika/ (dataobrashheniya: 18.08.2025).
2. Makroskop trejd, OOO. Kontrol` speczodezhdy` [E`lektronny`jresurs]. URL: https://macroscoptrade.com/st#rec143091316/ (dataobrashheniya: 18.08.2025).
3. Makroskop, OOO. Produkty` iuslugi [E`lektronny`jresurs]. URL: https://macroscop.com/ (dataobrashheniya: 18.08.2025).
4. Vipaks+, OOO. Sistema promy`shlennogo videonablyudeniya [E`lektronny`jresurs]. URL: https://vipaks.com/solutions/promyshlennyepredpriyatiya/ (dataobrashheniya: 18.08.2025).
5. Vipaks+, OOO. Detektor ognya [E`lektronny`jresurs]. URL: https://vipaks.com/catalog/videoanalitika -0000000024/litsenziya-namodul-videoanalitiki-detektor-ognya/#props/ (dataobrashheniya: 18.08.2025).
6. Bradski G. The OpenCV Library. Dr. Dobb’s Journal of Software Tools. 2000. V. 25. № 11. P. 120–126.
7. Redmon J., Farhadi A. YOLOv3: An Incremental Improvement. arXiv preprint arXiv:1804.02767. 2018.
8. Abadi M. et al. TensorFlow: Large-Scale Machine Learning on Heterogeneous Systems. Proceedings of the 12th USENIX Conference on Operating Systems Design and Implementation (OSDI’16). Savannah, GA, USA. 2016. P. 265–283.
9. Paszke A. et al. PyTorch: An Imperative Style, High-Performance Deep Learning Library. Advances in Neural Information Processing Systems 32 (NeurIPS 2019). 2019. P. 8024–8035. 1
10. Prikaz FSTE`K Rossii ot 11.02.2013 № 21 «Ob utverzhdenii Trebovanij ozashhite informacii, ne sostavlyayushhej gosudarstvennu yu tajnu, soderzhashhejsya vgosudarstvenny`xinformacionny`xsistemax». M.: FSTE`K Rossii. 2013. 1
11. Postanovlenie Pravitel`stva RF ot 17.02.2023 № 234 «Ob utverzhdenii Pravil opredeleniya znachimy`x ob``ektovkriticheski vazh noj informacionnoj infrastruktury` Rossijskoj Federacii». M.: Pravitel`stvo RF. 2023.