Ore and material classification
Sensors and deep learning classify material on conveyors for sorting and grade control.
The problem
Grade estimates are slow and sorting decisions rely on sampling.
Data we need
- Sensor and image data from conveyors
- Assay results
- Plant process data
How it works
- Install sensors on the conveyor
- Train classification models against assays
- Validate on production runs
- Integrate with sorting equipment
- Retrain as ore bodies change
What it changes
Targets are agreed with you before we build, and measured on your data before release.
- Better grade control
- Less waste processed
- Faster decisions than lab sampling
Deep learning
Neural networks for problems where hand-written rules and classic models fall short.
- CNNs
- Transformers
- Autoencoders