AI X-rays sort Brazil's potash before it ever reaches the plant — richer ore, less waste
Near Manaus in Brazil's Amazonas state, engineers from Brazil Potash and ERCOSPLAN have been teaching X-ray machines to judge rock. Their AI-powered sorting system scans each chunk of potash ore — potash is a key ingredient in fertiliser — and in initial tests lifted a sample from about 13% to about 27% potassium chloride, screening out roughly 20% of the low-grade material before it ever reached the processing plant. The sample was small, and the company is clear that bigger tests must follow once production starts — but sorting early could mean more production, or a longer mine life, without digging a third shaft.
The upbeat factorsmarter sorting at the gate means less rock moved, less energy spent, and more of what feeds crops.
SDG 9 · Industry, innovation and infrastructure — Mechanism: tech innovation
Explore:
Discovery