How Emago turns a cuvette of sample into mineralogy, particle size and liberation.
After modest preparation, samples go straight into a cuvette. No polished sections, no vacuum, no harmless radiation.
This is an actual Emago frame of copper ore. Each scan captures many of them.
Deep learning traces each particle in the frame and measures its size.
Each particle is split into its mineral grains, the texture behind liberation.
A classifier trained against SEM-EDS and XRD names the mineral in every grain.
Not the tens a traditional count manages.
Each cuvette gets its own results, exported to Excel and PDF or sent to your LIMS.
Every particle is stored with its size, shape and grains. Mineralogy, assay, size, liberation and density are counted from those particles, not predicted or calibrated.
Emago puts mineralogy, particle size and liberation on the lab bench, within an hour instead of weeks. Ask us for a demo and an obligation-free quote.
The micrograph, particle outlines and grain identities are a real Emago scan of a copper ore.
Artificial Intelligence algorithms have advanced in other areas but applications in the minerals industry were still scarce — so it was uncertain whether AI was mature enough to identify complex minerals in process slurries in real-time. There is a market gap for real-time mineralogy so the right AI algorithms combined with the latest imaging hardware would open up mineralogy based process control. We executed extensive testing to select the appropriate microscopy configuration, sampling procedure, and AI algorithm.
We verified our methodology of detecting minerals using flotation samples from a poly-metallic sulphide deposit in NSW, Australia. We compared the grades calculated from the identified minerals against assay tests of the same samples from a laboratory. Our methodology was able to predict Cu, Pb, Zn, and S grades with high accuracy - confirming that our methodology detected the mineralogy correctly. The results were peer-reviewed and published in a journal.
With the concept proven, we focused our efforts to packaging the methodology into robust and portable forms for commercial use cases. With feedback from sites and domain experts, we planned for separate devices for use in the laboratory in batch mode and in the concentrator in continuous mode.
Delivered our first product, Emago, marking the beginning of our journey of turning the vision of real-time mineralogy into reality.
Taking the same technology from the laboratory into the concentrator with Elchemia, for mineralogy in seconds that informs process control.
