Case Study: MTC speeds up nuclear waste sorting annotation 10x with V7 Darwin

How MTC is Using V7 to Build AI for Sorting and Segregating Nuclear Waste

MTC, the Manufacturing Technology Centre, is an energy-sector research and technology organization working on autonomous waste sorting for nuclear decommissioning. To help build AI for identifying and separating mixed radioactive waste, MTC needed a faster, easier annotation workflow than its previous tool, especially for non-technical team members and for maintaining consistency across labelers. V7 Darwin was used for data labeling and segmentation.

Using V7, MTC built an efficient workflow for semantic segmentation with polygons, quality review, and pre-trained model-assisted labeling to annotate waste items such as rubble, metal, rubber, and batteries. V7 helped the team speed up labeling by 9–10x, improve annotation accuracy through robust QA and customizable workflows, and support pixel-perfect masks for model training; the team also highlighted V7’s intuitive interface and responsive support as key benefits.

MTC

Mark Robson

Technical Specialist