AI Guards the City’s Light: Neural Networks Automate Street‑Lighting Control in Moscow Region

Automated Oversight Replaces Routine Inspections

Street‑lighting failures become especially visible during the fall and winter months, when daylight hours shrink. Traditional oversight—citizen complaints or periodic patrols—tends to be slow and incomplete. The technological system deployed in the Moscow Region is changing this dynamic.

Specialized neural networks analyze video feeds from thousands of cameras every day, detecting dark street sections and courtyards in real time. As soon as the system identifies a failed lighting point, it automatically generates a work order and sends it to the responsible utility or service provider. According to regional regulations, repairs must be completed within 24 to 48 hours.

AI‑driven monitoring improves safety and quality of life across several dimensions. Reliable lighting reduces crime risks, lowers accident rates after dark, and minimizes the amount of time any area remains unlit. AI also responds faster than any manual reporting system: there is no need for a resident to search for contacts or submit a request—the algorithm identifies the failure and initiates the repair workflow automatically.

The neural network operates around the clock and in all weather conditions, monitoring thousands of points simultaneously, ensuring that even remote or low‑traffic areas are not overlooked.

“Deploying an intelligent system for monitoring outdoor lighting is a major step toward building a safer, more comfortable urban environment. Automated data collection ensures that information about malfunctioning assets goes to operators instantly, allowing repairs to begin much faster.”

Nadezhda Kurtyanik

Minister of Public Administration, Information Technologies and Communications of the Moscow Region

Scaling the Model Across Russia

The success of the Moscow Region’s deployment is already being reviewed as a model for broader adoption. After visiting the system, Deputy Prime Minister Dmitry Grigorenko emphasized that the practice deserves nationwide rollout.

The Safe City platform is already active in Saint Petersburg, Yekaterinburg, Kaliningrad, Kazan, Voronezh, Novorossiysk, and Sochi. These cities can integrate the Moscow Region’s lighting‑failure detection algorithms with minimal adaptation, accelerating modernization of critical urban infrastructure.

From Public Safety to Infrastructure Intelligence

Launched in 2015 as a public‑safety video‑surveillance program, the Safe Region system has evolved into a comprehensive technological platform. As of early 2025, more than 150,000 cameras were connected to it—and the number continues to grow.

A turning point came in 2023, when a portion of the camera network was upgraded with neural networks for urban‑environment monitoring. AI learned to detect not only suspicious behavior but also infrastructure failures: potholes, illegal dumping sites, and—especially relevant in winter—malfunctioning streetlights.

In 2024 the system detected 10,812 lighting defects. Cities such as Mytishchi, Lyubertsy, Krasnogorsk, Leninsky, and Podolsk were among those with the highest number of reported issues.

AI Makes Cities Brighter

Automated detection of thousands of failed streetlights is not an isolated project—it represents a deeper shift in urban management. AI and big‑data tools are becoming routine instruments for improving public services, strengthening safety, and enhancing residents’ comfort.

This shift is reshaping sector benchmarks: the performance of municipal service providers is increasingly evaluated not by the volume of complaints received but by the speed of automated detection and resolution of problems.