Intelligent Object.Identifier (IOI): AI-based object recognition for sophisticated sorting tasks
The Intelligent Object.Identifier (IOI) introduced at STEINERT many years ago uses artificial intelligence to reliably identify and sort even objects that are difficult to recognize. This AI-supported object recognition is particularly suitable for difficult sorting tasks where conventional methods reach their limits.
Our sensor-based sorting systems have been using artificial intelligence for years to recognize recyclable materials in the waste stream, recover them or separate them if they could damage the target fraction. The technology, which is based on deep learning, enables new ways of sorting.
What exactly is deep learning?
Deep learning is a sub-category of machine learning and AI. Deep learning uses artificial neural networks to solve particularly difficult tasks - tasks that traditional optical sorting techniques fail at. For example, STEINERT’S IOI can differentiate between food-grade and non-food-grade plastic packaging. Another example is the sorting of silicone cartridges from a polyethylene stream. While the outer wall of the cartridges is made of recyclable PE, silicone residue on the inside can contaminate the recycled product. The Intelligent Object.Identifier uses characteristic optical features to reliably identify the cartridges, enabling them to be separated in a targeted manner.
How the machine learns what needs to be sorted
Deep learning achieves its impressive accuracy through intensive training: our experts feed the neural network with millions of marked images. This teaches the system to recognize subtle differences, even if the packaging is crumpled, damaged or dirty.
Working with AI-supported sorting technologies makes it possible to economically exploit previously unusable material streams and make better decisions based on data. In combination with our tried-and-tested sensor-based sorting technology, this creates an innovation boost for high-quality recycling.
By using this innovative technology, sorting tasks can be solved that were previously not economically feasible. AI-based detection stabilizes the sorting process and significantly increases sorting performance at the same time. In this way, recyclable material streams can be processed more efficiently and higher recycling rates can be achieved.
AI-supported sorting solutions for the:
- Sorting PE silicone cartridges from material streams
- Sorting food packaging from non-food packaging
- Separation of aluminum cans from other materials made of aluminum
Your benefits:
- Identification and separation of difficult objects
- Stabilization and increase in sorting performance
- Contact us if we can support you.