(Bloomberg) -- Codelco is using artificial intelligence to squeeze out more copper from its aging Chilean mines as suppliers of the metal look to boost efficiencies amid a dearth of new deposits and growing demand. 

Battling falling grades, rising costs and growing environmental scrutiny, Codelco introduced a digital data center in 2020 that uses machine learning. The platform is adding about 8,000 metric tons at its century-old Chuquicamata mine, or $80 million in annual profit, the state-owned company said in a statement.  

The machine learning systems use a stream of data on extracted ore to optimize processing through, for example, the use of blending. Now Codelco, the world’s biggest copper supplier, will look to introduce AI systems at other operations.

Mining companies around the world are turning to new technologies to extract more from current operations as new deposits become pricier and trickier to find and develop. The struggle to ramp up production is coming when demand for the wiring metal is expected to rise along with the transition away from fossil fuels. 

While Chile has the largest copper reserves, ore quality has been steadily falling. That means mines need to move more rock to produce the same amount, pushing up costs. Water restrictions and labor protests have also stifled output in recent years. Codelco saw production fall 9.3% last quarter from the prior year.

©2022 Bloomberg L.P.