APPLICATION OF ARTIFICIAL INTELLIGENCE TECHNOLOGIES IN AUTOMATIC TRAIN CONTROL FOR RAILWAY TRANSPORT

Abstract

An integrated artificial intelligence (AI) platform is proposed for optimizing train operations in railway transport. The concept combines a unified data platform, machine learning (ML) and reinforcement learning (RL) prediction modules, real‑time timetable rescheduling (MILP + RL), predictive maintenance, and cybersecurity standards. The goal is to reduce delays, improve capacity utilization, and enable proactive service management. Current literature shows that AI adoption in railways often remains at pilot stages due to fragmented data silos and limited integration across subsystems. This paper emphasizes the importance of data governance and standardization, while highlighting research gaps in predictive service integration, model robustness, and real‑time decision support. Based on Uzbekistan’s digitalization initiatives, a roadmap is proposed including evaluation metrics, risk assessment, and mitigation strategies.