Environmental AI to Track Illegal Landfills Across Russia with Satellite Imagery
Environmental advocates are increasingly turning to artificial intelligence to locate illegal landfills, using technologies like the Geometry geoanalytical platform developed by Sber AI. This information emerged during the Russian Environmental Forum (REF) when Alexander Vedyakhin, First Deputy Chairman of the Board of Directors at Sberbank, spoke to the press service of the financial institution. (Source: Sber press service)
The pilot program is a collaboration among Sber Business Soft and the Russian Environmental Operator, with an agreement announced at the REF held at Sber University. The aim is to harness AI to scan satellite imagery and automatically signal potential landfill sites, enabling rapid action to curb illegal dumping and reduce environmental harm across the country. (Source: Sber press service)
At the core of the project is a neural network trained to interpret satellite photos and highlight zones that warrant closer inspection for possible waste dumps. By identifying these areas quickly, authorities and volunteers can curb the growth of unauthorized landfills and lessen ecological damage to Russia’s landscapes and communities. (Source: Sber press service)
For the implementation, the Russian Environmental Operator supplied Sber with a comprehensive dataset containing known legal and illegal dumping sites. This historical information was used to flag and train the AI model. In the next phase, the system will automatically analyze new satellite images to detect signs of natural waste accumulation. (Source: Sber press service)
From this point onward, human volunteers will review AI-generated findings, validating or refuting the existence of suspected dumps on the sites identified by the model. The process blends machine efficiency with on-the-ground verification. (Source: Sber press service)
Alexander Vedyakhin noted that more than 15,000 illegal landfills have been documented in Russia to date, though many more remain uncounted. New dumps continue to appear each year, and the exact numbers are difficult to determine. (Source: Sber press service)
“Our collaboration with the Russian Environmental Operator aims to build a comprehensive map of illegal landfills across the country. AI technologies will help reveal all instances of natural waste accumulation. Once the model demonstrates reliable and accurate recognition, the pilot can evolve into ongoing monitoring, allowing authorities to track changes rapidly and halt environmental violations,” Vedyakhin explained. (Source: Sber press service)
Denis Butsaev, General Director of the Russian Environmental Operator, emphasized that the joint project with Sber will enable a clearer assessment of environmental conditions across Russian regions and pave the way for federal-level proposals to adopt these results more broadly. (Source: Sber press service)
In the North American context, similar AI-driven monitoring programs are increasingly discussed as tools for environmental protection, urban planning, and compliance in Canada and the United States. Cross-border experts point to satellite data, machine learning models, and volunteer verification as a powerful mix for tracking illegal waste, informing policy, and improving response times. The ongoing collaboration between institutional operators and technology providers in Russia serves as a case study in how AI can scale environmental surveillance with community involvement and transparency. (Source: Sber press service)