Case Study: PolyPerception achieves better waste stream visibility with Sama data labeling

PolyPerception’s Accurate Data Gives Sorting Facilities Better Waste Stream Visibility

PolyPerception, an AI-powered waste management platform for plastics and material recovery facilities, needed accurate labeled data to build a robust multi-object tracking model. Their challenge was handling fast-moving waste objects in low-light sorting facilities, with very high waste volumes and a wide variety of packaging types and materials that change by region and over time.

Sama provided high-quality computer vision data labeling support through its annotation services, helping PolyPerception label millions of waste objects and maintain an average quality score of 99%. With Sama’s open feedback loops, PolyPerception was able to adapt to changing regulations and regional differences, improving waste stream visibility and enabling more efficient operations, better recycling outcomes, and stronger data-driven decision-making.

PolyPerception

Rafael Hautekiet

Chief Executive Officer