On the Moscow Central Circle (MCC), will install an AI system to detect ‘anti-social behaviour’ by passengers to prevent offences. The development cost of the platform will amount to 15.7 million rubles.
The tender for developing the system was announced by the Scientific Research and Design Institute for Informatization, Automation and Communications on Rail Transport (NIIAS), 75% of whose shares are owned by Russian Railways.
According to the tender documentation, the system is planned to be deployed at stations and in the carriages of the ‘Lastochka’ trains. It will analyse video footage in real time to identify signs of disorder and send notifications to the Unified Dispatch Centre of Russian Railways. The project also envisages the use of acoustic analytics tools to detect signs of violations of the rules for using railway transport. Among such signs are:
- screaming;
- signs of aggression (analyzing tone of voice and manner of speech);
- explosions and other loud sounds, atypical for the noise level inside the carriage.
NIIAS said the installation is driven by requirements for staying on railway transport facilities ‘in conditions of fully automated unmanned train movement on the MCC’. Previously, the company said it planned to launch autonomous ‘Lastochkas’ on the ring line, but this would not happen before 2022. Russian Railways needs to obtain the corresponding certificate.
There are no announced timelines for implementing the anti-social behaviour detection system on the MCC. NIIAS will announce the results of the tender on 28 October 2021.
In October, it emerged that Moscow’s traffic police are employing AI to track suspicious vehicles and recover partially obscured license plates.
In September, authorities announced plans to identify ‘dangerous schoolchildren’ by analyzing their essays with AI.
In August in St. Petersburg, they presented an AI-powered software suite for social media analytics that identifies users of ‘death groups’ or teenagers with suicidal tendencies.
Subscribe to ForkLog News on Telegram: ForkLog AI — all the news from the world of AI!