While developing NEWASTE, our AI-assisted plastic waste classification project, one number really stayed with me:
In some recycling plants, impurity levels in plastic waste streams were reaching almost 30%.
Before working on the project, I honestly thought recycling mostly depended on people separating waste correctly at home. But I realised a huge part of the challenge happens later, during classification inside the plants themselves.
That’s why our solution focused on using computer vision and Deep Learning to help identify and remove improper waste faster and more accurately before contamination spread further through the process.
What surprised me most is how small improvements in sorting efficiency can have a much bigger impact on the entire recycling chain than people usually imagine. ♻️