An AI-enabled technology the consortium is developing uses algorithms that can map rock layers in lithological logs to reduce the time to needed to evaluate a drill core from weeks to minutes, and with increased accuracy, said LKAB.

"Through this collaboration, core logging takes a big step through machine learning and artificial intelligence," said says Jan-Anders Perdahl, specialist at LKAB's exploration department.

"A geologist can at an early stage place greater focus on the parts of the core that show chemical or other changes. Opportunities are opened up to gain increased knowledge about ore formation processes and alterations in a completely different way than before.  One can also get indications that you are close to a mineralisation and where it may be located and thereby streamline exploration," he said.

The partnership began with tests launched last March with Minalyzer's CS drill core scanner that aimed to improve the workflow for core logging.

The test led to a permanent installation in LKAB's Kiruna mine, and expansion to Malmberget where data from Minalyzer CS is used to help geological logging of the drill cores.

The partners now plan to take the project to "the next level" by leveraging Sentian's AI expertise to analyse boreholes in LKAB's deposits.

"We are at an interesting time when the hardware to extract consistently high-resolution data from drill cores is available and we can now take the next step and generate value from data together with our customers," said Annelie Lundström, chief executive of Minalyze.

"In this collaboration, we will develop algorithms that can map rock layers in so-called lithological logs with very high confidence. This can only be done by combining expertise from all three parties. The results from our collaboration will forever change how drill core logging takes place everywhere and will result in a more efficient, non-subjective and consistent process".