The Chinese translation is being made; the original is shown for now.
In what it claims is one of the first cases of its kind in the mining industry, artificial intelligence models developed by Stratum AI are underpinning the publicly reported mineral resource estimate of an operating mine.
The models created by Stratum AI, a Toronto-based deep-tech company, were developed in collaboration with Perth-based mining company St Barbara Ltd. The miner has incorporated Stratum AI’s deep-learning models of weathering and material type (oxide, transitional and sulphide) into the mineral resource estimate for its Simberi gold operations in Papua New Guinea.
The resource estimate, which is based on these AI models, is reported in accordance with the JORC Code, which is incorporated into listing guidelines that companies must meet to remain listed on the Australian Securities Exchange.
“St Barbara has harnessed Stratum AI’s proprietary algorithms to classify certain ores that were previously considered to be unsuitable for Simberi’s existing CIL circuit,” Farzi Yusufali, Stratum AI’s CEO and co-Founder, says. “What began as an AI pilot has evolved into a technology that now underpins the mineral resource estimate in Simberi’s Feasibility Study. We see this as a significant milestone in the adoption of AI in mining.”
St Barbara released the study on December 10, 2025, the same day it announced a strategic transaction with Lingbao Gold Group that values 100% of the New Simberi gold project, the expansion of the existing Simberi gold mine, at approximately A$800 million ($562 million).
Stratum AI has been working with St Barbara for over four years at the Simberi gold mine to develop an AI-based approach to better classify mineralisation. Stratum’s AI models were initially deployed in pilot programs to analyse geological exploration data. They were then used to analyse operational data, allowing the miner to determine which ores were best suited to be fed through the existing carbon-in-leach (CIL) plant that treated oxide ores, or needed to be stockpiled for later treatment with sulphide ores.
Being able to differentiate between the types of ores allowed the miner to decrease costs, especially costs associated with generating extra waste, and saved wear and tear on the mill.
After the models were validated against grade-control reconciliation, St Barbara adopted them as its preferred model for material-type classification. The modelling helped build Simberi’s 2024 resource and reserve update and the mining operations’ 2025 Feasibility Study.
In May 2024, the mining company said it was able to identify an additional 143,000 oz of treatable material at Simberi, extending the mine’s current oxide processing life by 17 months. A month later, St Barbara said in a press release Stratum’s modelling had helped it increase its proved and probable ore reserves by 40%.
Stratum AI’s experience at Simberi illustrates how the company is fulfilling its mission to help producing mines apply machine learning to existing geological and operational data to make better decisions that can lower costs and maximise value, the company says. Mining generates vast amounts of data throughout its lifecycle, yet much of its value remains under-utilised during production and expansion.
The post Stratum AI models gaining recognition with Simberi resource JORC milestone appeared first on International Mining.