How AI Sorting Algorithms Work in Recycling Deep Learning TOMRA

Artificial intelligence and sorting algorithms are revolutionizing material recovery for recycling. Learn how TOMRA’s deep learning technology works.
Sensor-based sorting systems by TOMRA have relied on artificial intelligence for decades. In our relentless pursuit of innovation, our team of engineers and recycling experts have developed an AI-powered solution that is set to transform the recycling industry, significantly enhancing material circularity and sustainability.
Our deep learning technology identifies hard-to-classify objects that cannot be achieved with traditional optical waste sorting equipment. Material recovery facilities and recyclers maximize sorting performance and operational efficiency by automating these complex sorting tasks.
What is deep learning, and how do sorting algorithms work?
Deep learning technology is a subset of artificial intelligence and machine learning based on artificial neural networks. It uses multiple layers of different functions that identify and classify objects. The more complex the sorting task, the more layers of data, and hence, the deeper the learning.
At TOMRA, we apply this technology with our groundbreaking GAINnext™. The sorting system offers a wide and ever-growing application ecosystem developed explicitly for the recovery and recycling of:
Food-grade plastic packaging (PET, PP, HDPE)
Higher quality rPET and white opaque bottles
Solid wood and MDF from waste wood
High-purity paper for deinking processes
Aluminum used beverage cans (UBC)
More applications in development