“Gorodovoy” Keeps St. Petersburg Clean and Maintained
The State Administrative and Technical Inspectorate of St. Petersburg continues to train the Gorodovoy AI system. The latest update expands its ability to detect violations related to urban maintenance and public space conditions.
Initially, Gorodovoy could identify only a limited set of issues, including graffiti, air conditioners installed on building facades, and potholes. Over time, its capabilities have expanded. Today, the system can detect more than 20 types of violations across the urban environment, ranging from unauthorized kiosks to snow buildup and icicles on rooftops. The system operates in both “summer” and “winter” modes, with certain functions adjusted depending on the season.
This year, Gorodovoy is set to gain new capabilities. It will be able to detect overflowing trash bins near shopping centers, scaffolding installed without permits, and road surface subsidence during warranty periods.
The system relies on more than 20,000 smart surveillance cameras from the Bezopasnyy gorod (Safe City) network. In addition, mobile video units mounted on eight vehicles patrol the city and capture violations. All data is transmitted to video servers, where Gorodovoy processes it. When the system detects an issue, it automatically generates and assigns a task to the responsible municipal services. Last year, the AI identified about 2,500 violations, resulting in fines exceeding 170 million rubles (approximately $1.9 million).
Citywide Coverage
This year, Gorodovoy will expand to cover all of St. Petersburg. An additional 80,000 cameras will be connected to the system. Such AI-driven tools can significantly improve the quality of urban infrastructure. In practice, municipal services will be able to respond more quickly to violations and optimize cleaning and maintenance operations.
A comparable example can be seen in the Moscow Region. For more than two years, an AI system has monitored cleanliness and order across the region. Using 83,000 cameras from the Bezopasnyy region (Safe Region) network, it detects 15 types of violations and automatically assigns tasks to municipal services. The system has already identified more than 251,000 violations. Over the past three years, the number of AI-driven projects in the Moscow Region has increased more than fourfold, from 9 to 41.
Toward Cleaner, More Responsive Cities
With each new function, Gorodovoy is building a unified digital layer for monitoring urban maintenance. The expansion of AI systems in St. Petersburg’s city services marks a shift in how urban management operates, from reactive issue detection to continuous, proactive monitoring.
Computer vision technologies like this can be applied across a wide range of municipal tasks, from monitoring waste collection to detecting illegal advertising and parking violations. The approach can be scaled to other regions in Russia. Looking ahead, the inspectorate plans to deploy Gorodovoy on unmanned boats. These vessels will help detect violations that are difficult to capture from ground-level or fixed camera angles.
