Artificial Intelligence : Iren and Recycleye: A new partnership to install waste compositional AI analysis systems

Recycleye has announced a partnership with the Iren Group to install 6 AI systems for material compositional analysis at the Collegno MRF managed by Amiat, a company in the Group that handles Turin’s waste system. The computer vision systems, developed by Recycleye, will be installed on 6 lines in the paper sorting facility to track the composition of outgoing material flows and be consulted in near-real time by the operating team.

Iren, a leading Italian multi-utility, operates in the water, energy, gas and waste sectors across several regions, serving over 7 million inhabitants. Iren Ambiente provides collection services to 3.8 million inhabitants and processes 3.8 tonnes of materials per year. In Piedmont, the region in which the project will take place, 1.4 million inhabitants receive this
service from Iren, with 647k tonnes collected annually.

Recycleye is a London-based company that develops AI-powered technology for sorting and analysing material fractions processed at sorting centres, such as computer vision systems, robotic pickers and optical sorters.

The Amiat sorting facility treats fractions of paper, cardboard and pulp from curbside and door-to-door collection. Iren is currently renovating the plant, updating its systems to increase production capacity and the purity of treated fractions. The operations team was therefore seeking a solution to measure the success of the newly revamped plant, capable of analysing the composition of the outgoing material flows. To that end, Iren invested in 6 Recycleye computer vision systems to analyse the composition of the materials arriving at the sorting cabin across 6 belts; 3 systems will analyse streams of various paper types, 2 will analyse multi-material liquid carton streams and one will analyse a cardboard stream.

Recycleye developed this innovative technology and each unit comprises a camera, LED light modules and a compute unit. The computer vision system scans each piece of waste travelling on the belt and, using proprietary machine learning algorithms, identifies the type of material and object. Artificial intelligence detects using a range of visual characteristics, similar to the human eye. This means that, unlike other technologies such as NIR, which distinguish different fractions through spectroscopy, the AI system makes distinctions by shape, colour, size, etc. it can also detect multi-material objects, such as liquid cartons. Therefore, once installed in early 2025, the Recycleye systems will provide Amiat with near real-time granular data on the percentages of each material detected in their respective streams, made available via a user-friendly dashboard. This investment will support Iren management to monitor the purity of its outputs, data that can guide decisions on plant and machinery operations, indicating any anomalies or highlighting potential improvements.

Turi Lo Sardo, Technical Sales Manager (Italy) at Recycleye, added: "We are proud to collaborate with Iren on this innovative project. Our partnership represents an important starting point; today artificial intelligence is an increasingly indispensable tool for Italian waste management companies to achieve higher quality output and detailed data analysis. I am confident that this can be the first of many fruitful collaborations between our companies."