Treasure in our rubbish: artificial intelligence and waste recognition

Artificial intelligence offers promising solutions for optimising waste management. Thanks to its advanced recognition capabilities, it can accurately identify and sort different materials, paving the way for a sustainable circular economy.

Autodiag, the smart camera that recognises waste

To ensure that our waste is correctly recycled, we need to sort it properly and group it together, including newspapers, packaging etc. Everyone is responsible for this at home, using the recycling bin. This sorting is then complemented by the work of the sorting centres, which narrow down the material by flow type (plastics, paper and cardboard) to provide a material that is "ready to recycle", and as pure as possible.
 In order to improve this process of classifying waste, SUEZ has developed the Autodiag solution. Using an arch equipped with a camera placed directly on the sorting table, Autodiag measures the quality of waste material in real time. How does it work? Algorithms, fed with labelled images of different types of waste, calculate the purity and uniformity of the different qualities of materials produced. 
98,5 %
Autodiag is now operational for plastic waste, with a success rate of around 98.5% in detecting incorrectly sorted waste.

QualiWaste UVE, the computer vision to collect and recover waste more effectively

The aim of QualiWaste UVE is to better identify and understand the nature of the waste arriving at EfW plants, so that all waste that can be recycled can be recycled.

With QualiWaste UVE, cameras are placed in the pit of the EfW (Energy from Waste) plant to film the dumping of material by waste-providing' trucks. The images are then sent to an algorithm that detects:

  • The type of waste brought. Identifying waste in this way will ensure that it complies with what has been declared by the contributors and will increase its traceability.
  • Unwanted waste. An alert system helps warn SUEZ teams and removes this waste using a grapple. This process enables unscheduled shut-downs of the furnace to be avoided and reinforces the safety of the site and the teams!
  • Recoverable waste. This waste is reported on, with its presence being quantified to help contributors improve how the materials they deliver to waste-to-energy plants is sorted and direct them towards the right recycling channels in future.
~ 90 %
La solution permet aujourd’hui un taux d’identification des déchets indésirables et des matières valorisables de presque 90 %.  Et les algorithmes, nourris en continu, sont en progression constante !  
Contact us

Want to know more about our innovations? Get in touch with us!

*mandatory fields

This form is only for companies, local authorities or public organisations. The information collected in this form is processed by SUEZ in order to provide you with the information requested on innovation. You can find your rights and SUEZ's data protection policy in our privacy policy.