AI System to Identify Disorderly Passengers on Moscow’s MCC

The Moscow Central Circle (MCC) is set to implement an artificial intelligence (AI) system designed to remotely detect signs of antisocial behavior among passengers, according to a report by Vedomosti.

The software and hardware complex will be installed on “Lastochka” electric trains and at MCC stations. The system will automatically and in real time analyze video footage from surveillance cameras inside train cars and on railway platforms, as well as acoustic information, including speech. Using machine vision and AI, the system will recognize shouting, displays of aggression (such as raised voices and changes in speech patterns), explosions, and other unusual sounds not typical for the train environment.

The competition to develop this system was announced by the Research and Design Institute of Informatization, Automation, and Communication in Railway Transport (NIIAS), a subsidiary of Russian Railways (RZD). The results will be announced on October 28, and the project is valued at 15.7 million rubles, according to information on the government procurement website.

Currently, 242 pairs of trains operate on the MCC on weekdays and 211 pairs on weekends, including the high-speed “Lastochka” trains produced by Ural Locomotives. RZD plans to launch driverless “Lastochka” trains on the ring in 2024, with certification for these trains scheduled for 2022.

At present, all “Lastochka” trains on the MCC are equipped with video surveillance systems, with information transmitted to the train operator’s control panel. The operator is responsible for taking action in cases of antisocial passenger behavior. After the new AI system is implemented, data will be sent to a unified RZD dispatch center.