Artificial intelligence improves waste sorting
Théophile Agresti, Théo Vitupier and Valentin Ibars, founders of WasteFlow (Photo: EPFL/Alain Herzog)
June 25, 2026 – Researchers at EPFL have developed an artificial intelligence system that improves the sorting of waste materials. The technology analyses images of mixed waste streams and recognises materials such as plastics, metals and paper, even when objects are dirty, damaged or difficult to distinguish.
Accurate sorting is essential because the quality of recycled materials depends largely on how well different waste streams are separated. By using artificial intelligence, sorting facilities can process waste more quickly while reducing the number of incorrectly classified items.
The researchers believe the system could make recycling more efficient and increase the amount of valuable materials that can be recovered for reuse. As a result, less waste may end up in incinerators or landfills, while demand for newly extracted raw materials can be reduced. According to the researchers The work demonstrates how digital technologies can strengthen the circular economy.
