Russia to Create a Digital Twin of the Amur River
A combination of imaging technologies and neural network models will be used to recreate the river in virtual space.

Russian scientists have begun developing a digital twin of the Amur River, according to the Pacific National University. In 2026, specialists plan to photograph the riverbed. Additional data will be collected through aerial photography, sonar surveys, and laser scanning. Together, these methods will allow researchers to reproduce the river in a virtual environment.
This year, experimental water-based imaging is set to be conducted using Sanabot (underwater survey device). The system maps underwater terrain and will complement aerial surveys, researchers said. Part of the data has already been gathered using the Geoscan (unmanned aerial vehicle). In a single flight, it covered up to 10 kilometers of shoreline.
The collected information is expected to improve navigation, strengthen shoreline protection, support environmental measures, and enhance fisheries management.
Proven Methodology
Researchers have worked to document the river across different seasons and hydrological periods. The first surveys were conducted in late 2024. The focus has been on the coastal zone, covering up to 200 meters of water and roughly 150 meters of adjacent shoreline, including settlements, forests, and recreational areas. In 2025, the team used lidar, a laser-based remote sensing technology, to generate a dense “point cloud” that serves as a prototype for the digital twin. A multispectral camera captured details invisible to the human eye, helping researchers assess vegetation and soil conditions.
For the next stage, scientists developed neural network models to recognize the Amur’s water surface and shoreline objects. They also designed algorithms to forecast shoreline shifts and built a geoinformation system for data analysis and visualization. Ultimately, the team aims to create a validated methodology and software product capable of generating digital twins for any terrain.








































