In Riom, near Clermont-Ferrand, the Syndicat du Bois de l’Aumône (SBA) has begun testing an artificial-intelligence solution designed to reduce recycling mistakes and lower waste-management costs. The technology, developed by Grenoble-based startup Ficha, uses cameras installed inside two of the SBA’s 22 collection trucks to detect sorting errors in yellow recycling bins.
Each time the AI identifies an incorrect item, most commonly opaque plastic bags, it captures an image and sends it to a central dashboard. The system cross-references the data with tagged bins to identify households responsible for the errors. Residents are then contacted through personalized letters to encourage proper recycling practices.
The €40,000 pilot project aims to address a persistent problem: nearly one in three recycling bins in the area contains sorting mistakes, costing the SBA around €1 million annually in incineration and lost recycling revenue. If results remain positive by the end of the year, the AI system could be rolled out to the entire fleet and potentially to collective drop-off points.
The initiative forms part of the SBA’s wider modernization strategy, covering 161 communes in the Puy-de-Dôme department, to improve waste-sorting quality, reduce emissions, and cut operational costs without adding workload for collection agents.
