AI and satellites to detect illegal landfills with 90% accuracy
A team of researchers from the Department of Electronics, Information and Bioengineering (DEIB) at Politecnico di Milano has developed an Artificial Intelligence-based system using satellite imagery, capable of detecting illegal waste dumps with an accuracy exceeding 90%. This technology, already operational in several European countries, represents a concrete tool to accelerate environmental investigations and combat illicit waste disposal activities.
The work of the Politecnico researchers is part of PERIVALLON, a project funded by the Horizon Europe program, involving 24 partners from 12 European and non-EU countries, including Fondazione SAFE, ARPA Lombardia (Regional Agency for Environmental Protection), and the Ecological Unit of the Carabinieri.
Among the most significant results, the pilot study conducted across over 100 municipalities in Lombardy in collaboration with ARPA demonstrated that the use of AI drastically reduces investigation times, enabling faster and more effective interventions to protect the environment and public health.
A key tool of the project is the AerialWaste dataset, the world’s first public archive for the recognition of illegal dumps from aerial images.
Thanks to Artificial Intelligence and satellite image analysis, we can detect signs of environmental crime on a large scale, quickly and accurately. These tools mark a paradigm shift, enabling systematic regional monitoring to support law enforcement and environmental agencies.
Piero Fraternali and Giacomo Boracchi, Department of Electronics, Information and Bioengineering