Abstract
This thesis examines the problem of evaluating the efficiency of AI based intelligent passenger flow management systems on metropolitan escalators. The aim of the study is to develop a comprehensive evaluation criteria system covering technical, operational, economic, and social-safety dimensions. As a result of the analysis, a four-group criteria system and the concept of an integral efficiency index are proposed.
References
1. The number of passengers transported on the Tashkent metro increased by 20 million in 2025 // Gazeta.uz. — 2026.01.13. — URL:
https://www.gazeta.uz/oz/2026/01/13/metro/
2. Akkaya I.B., Halici U. A YOLO-based method for real-time crowd detection // J. Real-Time Image Processing. — 2023. — Vol. 20. — DOI: 10.1007/s11554-023 01290-y.
3. Evaluating YOLO Models for Detecting Crowds in Sparse Regions // Springer LN. — 2025. — DOI: 10.1007/978-981-96-8796-1_5.
4. Distributed Deep Learning System for Real-Time Passenger Flow Measurement // Applied Intelligence (Springer). — 2025. — DOI: 10.1007/s10489-025-06954-9.
5. Osipov V. et al. Intelligent escalator passenger safety management // Scientific Reports (Nature). — 2022. — Vol. 12. — DOI: 10.1038/s41598-022-09498-x.