How a neural network in the Moscow region imposed discipline in courtyards and parking lots

Meanwhile, the neural network is already actively working in the urban economy of the Moscow Region - it is fighting for cleanliness in the courtyards, putting things in order and discipline/In total, more than 130 thousand cameras of the Safe Region system are currently installed in the Moscow Region. Recently, about 66 thousand cameras of the Moscow region have been connected to the neural network.

Artificial intelligence, vigilantly observing through the lenses of cameras behind the streets and courtyards of cities and towns near Moscow, is trained to distinguish and record 14 types (for starters) of different violations. So, the neural network is able to assess the quality of garbage collection, the state of container sites, issue fines to motorists for parking on lawns, monitors the cleanliness of children's playgrounds and order in the yards. Also, AI is able to report "where necessary" about illegal graffiti on the walls of houses.

The Minister of Public Administration, Information Technologies and Communications of the Moscow Region Nadezhda Kurtyanik emphasizes that the entire video surveillance and artificial intelligence system of the "Safe Region" in the Moscow Region runs on domestic software.

After combining the cameras of the "Safe Region" with AI, a quick process of improving the urban economy began. The most illustrative example: complaints and signals from residents about the accumulation of yard rubbish, graffiti-painted fences and houses littered with containers not taken out with garbage had to be checked by specially trained people. That is, to fix the violation and take the necessary measures - a fine, administration to the Criminal Code, a staff of employees, vehicles, human hours, etc. was required.

In our technological time, the problem is most often solved before writing complaints. The neural network sees garbage in the children's play area, photographs the mess and independently calculates the Criminal Code responsible for cleanliness on the site. Then he sends a photo application with a description of the problem to the management company.

The neural network does not stop there, it takes control of the issue. After the time allotted to eliminate the violation, the AI ​ ​ again photographs the problem area. If the garbage is removed, the application is closed. And if the Criminal Code is in no hurry, then the information goes to higher authorities and a fine is imposed on the sluggish management company. The effect is amazing: 350 employees of various departments of the government of the Moscow Region previously barely coped with this work, which is now performed by AI. And the issues were resolved much more slowly.

Another typical example that is close and understandable to all residents of high-rise buildings near Moscow: car parking near garbage containers, which makes it difficult to access garbage trucks. Very often, garbage in the courtyards accumulates and is not taken out on time due to the fact that the entrances to the tanks are blocked by personal "ferraris," "lamborghinis" and "moseratti" of local residents. Therefore, the next smart function that the regional authorities hastened to give the cameras of the "Safe Region" was control over the parking of container sites.

Note that the legislation provides for a fine if motorists block the entrance of special vehicles to garbage sites. But before, in practice, the law was rarely applied. The traffic police inspectors were supposed to write out the fine, but this happened occasionally, because - where to find inspectors for each yard?

The introduction of neural networks saved the residents of the Moscow region from this problem. Now the camera takes two pictures of the blocked area with a difference of ten minutes. If the parked auto-intruder has not left the containers during this time, the AI ​ ​ automatically sends both photos for a fine. Since the beginning of 2025, the number of such violations, using a neural network, has decreased by 40 percent.

The functionality of AI in the role of a "yard overseer" is constantly expanding - since April, artificial intelligence has learned to additionally record incorrect car parking, with a stop on the lawn.

Neural networks have also learned to deal with another unpleasant feature of life in multi-storey buildings. These are old dressers, sofas, mattresses and other bulky garbage thrown out by residents, which could previously "decorate" courtyards near Moscow for weeks. And sometimes it also serves as a "relaxation room" for stray animals and more.

Old furniture, construction waste and broken household appliances should actually be thrown into large-sized special containers. But very often irresponsible residents throw this rubbish in the middle of the night onto a regular container site. What's next? Garbage trucks take only the contents of the tanks, and what is nearby is not their diocese. Bulky garbage is disposed of at other sites, by other companies, it is necessary to send a separate car after it, and this is a considerable cost.

As a result, a set of old furniture can take root at the trash can and even become a landmark of the yard. Then it will enter tourist routes and fall into the UNESCO material heritage list. And then you won't get rid of old things at all.

Neural networks, fortunately, indicated a way out of this cultural impasse. Having discovered non-standard garbage near the containers, artificial intelligence sends a photo to the Criminal Code with an order to remove furniture, or building materials. "We give the management company 48 hours to clean up the trash. If the Criminal Code does not cope at a given time, it is fined 50 thousand rubles, "said Vladislav Murashov, Deputy Chairman of the Moscow Region Government, Minister for Territorial Maintenance and State Housing Supervision of the Moscow Region. In the first three months of 2025 alone, the management companies of the Moscow Region were generally fined 19.4 million, precisely for bulky garbage near container sites.

Now the Criminal Code themselves are interested in catching violators among residents. Employees of management companies view CCTV cameras, identify the specific culprit and write statements to the police asking them to bring the citizen to justice. The scope of responsibility is determined by the regional Code of Administrative Offenses - a fine of one to five thousand rubles per person. The decision on the specific amount of the fine is made by the administrative commission, where the police transfer the protocol of violation. Many guilty of cluttering yards have already parted with blood money.

Although, of course, here it is necessary to pay attention to the elimination of objective causes of offenses - the lack of large containers. According to the standards, one such container should be installed on several yards. But often their number in microdistricts is determined simply by the presence (or absence) of a large site where cleaning equipment can maneuver.

As soon as the work of neural networks with car parking and garbage collection is finally debugged, the authorities of the Moscow region intend to take on the next stage of the project - the placement of cameras of the "Safe Region" with AI in children's play areas. In this case, the neural network will be designed to record violations on the part of beer drinkers and hang out on carousels at late and night.