Russia Is Getting a Digital Twin of Its Population
Researchers in Samara have built a map that estimates how many people live in a specific building today, including their gender and age, and how those numbers are likely to change over the next five years.

Government agencies and businesses regularly make decisions about where to build schools, day care centers, stores, and other facilities. But they don't always know exactly how many people live near a proposed construction site – or how those numbers might change in the coming years.
Researchers at Samara National Research University, working with the company SAMIS, have opened beta testing of a digital model of Russia's population. The system can analyze areas ranging from entire regions down to individual residential buildings, estimate the number of residents and their age and gender breakdown, and forecast how those figures may change as infrastructure develops.
Building a More Detailed Picture
The new platform collects all available information about residential buildings, including their year of construction, number of floors and apartments, condition, and price per square meter. A machine-learning algorithm trained on buildings for which complete resident data is available then fills in the gaps for properties where such information is missing.
What sets the system apart is its ability to show not only the present but also possible future scenarios. For example, if a new cluster of high-rise buildings goes up in a neighborhood, the model can estimate in advance how many families with children might move there. It can also flag the need for a new school or day care center, or warn that public transportation may come under greater strain, before those problems arise. Estimates of the area's older population can help identify where medical services are most needed. Planners can test different scenarios: What happens if a residential complex is built here and an industrial park there? How many people will move in, and what facilities will they need? This kind of scenario analysis supports more informed decisions.
The system grew out of the Robot-kartograf (Robot Cartographer) project, which matches addresses with geospatial data, including information from Russia's Unified State Register of Real Estate (EGRN). The developers say that building-level detail produces more accurate estimates than aggregated figures for entire city blocks. Testing is scheduled to continue through the end of the year. The platform covers 85 regions, more than 23.5 million buildings, and over 71 million households.

Smart Models and Digital Twins
The Samara project is not the first attempt to use this kind of modeling. In 2022, for example, specialists at Vostokgosplan published research in which digital agents simulate demographic processes – birth rates, mortality, and migration – for scenario-based experiments. The focus was on the behavior of a hypothetical population, rather than linking demographic data to specific buildings.
In 2023, researcher Andrey Smirnov of the Komi Science Centre, Ural Branch of the Russian Academy of Sciences, unveiled an interactive website with detailed demographic data for Arctic territories, down to the municipal and settlement levels. The tool included ranking, clustering, and forecasting functions. Unlike the Samara model, which aims to cover the entire country and map population data to individual buildings, the Arctic project focuses on spatial data visualization and demographic profiles of territories within a specific region.
Rostelecom developed the Digital Twin of Perm Krai. It is an example of an infrastructure digital twin rather than a demographic one, but it shows how the technology can be deployed at the regional level. During the pilot, nearly 800,000 assets were digitized, from residential buildings and roads to utility networks. The system supports practical tasks such as tracking capital repair projects, monitoring waste accumulation sites, and analyzing urban development constraints. In the future, this infrastructure model could complement the demographic one.

Looking Ahead Through Scenarios
For shopping mall owners, pharmacies, package pickup points, and municipal officials responsible for land-use planning, knowing that a city has a population of one million is not enough. They need to understand how those people are distributed across neighborhoods: where more young people live, where the older population is concentrated, and where families with children are most likely to be found. Developers need to know who is likely to buy homes in a new neighborhood and what infrastructure those residents will need. Population distribution helps transit operators decide where to establish routes and how frequently to run buses. For business owners, it can help identify the street where a new location is likely to become profitable fastest.
The digital population model offers a way to look ahead by testing different scenarios. It does not replace official statistics. Instead, it supplements them where data is missing and helps estimate how many people will live in a particular location five years from now. The project combines geographic information systems (GIS), machine learning, and demographic modeling. These technologies and capabilities could also be applied to other products and services, including smart city systems, real estate management, and mapping platforms.









































