Key Takeaways
- Leapfrog Geo (Seequent), Micromine and Datamine are the established platforms for building the mine geological model.
- Leapfrog’s implicit modelling engine uses radial basis function (RBF) interpolation, an ML-adjacent mathematical technique, to build 3D geological surfaces directly from data. It is faster than traditional explicit wireframing.
- Implicit modelling, pioneered by Leapfrog, has since been adopted by Micromine, Mintec’s Minesight and Maptek’s Eureka. It is now the industry-standard approach rather than a differentiator.
- Marked “Emerging” because RBF-based implicit modelling isn’t the same as a trained predictive model. Genuinely learned (for example, deep-learning-based) geological modelling is still research-stage.
TL;DR
Build your mine geological model in Leapfrog Geo, Micromine or Datamine using implicit modelling rather than manual wireframing. Feed it drillholes and mapping, then let new data update the surfaces automatically while a geologist checks the result. It is fast and industry-standard, although it is ML-adjacent rather than a trained AI model.
What Is a Mine Geological Model?
A mine geological model is the 3D representation of lithology, structure and alteration beneath a deposit. It guides exploration targeting, resource estimation and mine design. Geologists build it from drillhole logs, assays, surface mapping and survey data.
Traditionally, the model came from explicit wireframing. A geologist manually drew contact surfaces between rock units, section by section. It is accurate when done well, but slow. Every new drillhole potentially means redrawing surfaces by hand. As a result, models often lagged months behind the latest drilling.
How Do I Build Mine Geological Models With AI?
Modern practice breaks into three steps: choose the method, pick the platform and decide how far to push machine learning.
Replacing wireframes with implicit modelling
Implicit modelling changed the workflow. Leapfrog Geo’s engine builds geological surfaces mathematically using radial basis function interpolation. In short, it fits a continuous surface to your data points rather than connecting them by hand. When you add a new drillhole or mapping observation, the model updates automatically. That is a genuinely different workflow from re-wireframing.
Choosing a mine geological model platform
This approach, first popularised by Leapfrog, is now supported by Micromine, Mintec’s Minesight and Maptek’s Eureka. Therefore, the choice today is less about who has implicit modelling. It is more about integration with your existing toolchain. For what it’s worth, Leapfrog imports both Datamine and Micromine file formats, so migration isn’t a hard lock-in barrier.
Adding trained machine learning on top
RBF interpolation has real kinship with machine learning, since it approximates a function from scattered data. However, it is not a trained predictive model in the way a recovery-prediction model is. Genuinely learned approaches, such as diffusion-model-based methods trained on geological patterns, are an active research direction. Our post on AI-assisted implicit geological modelling covers them in detail.
How the Mine Geological Model Fits the Wider Workflow
The model sits between data capture and mine planning. Upstream, a drone survey in mining supplies the terrain surface, and drilling supplies the subsurface data.
Downstream, geological domaining splits the model into zones for estimation. Those zones then populate the mining block model that planners schedule against.
Expect limits. Implicit models are only as good as the data and the geologist’s structural interpretation. Sparse drilling produces smooth surfaces that look convincing but may be wrong. Therefore, keep a geologist reviewing every automatic update.
FAQs
What software is used to build a mine geological model?
Leapfrog Geo, Micromine and Datamine are the established platforms. Maptek and Mintec (Minesight) also support implicit modelling. The best choice usually depends on the software your planning and estimation teams already use.
Is implicit modelling a form of AI?
Not strictly. Implicit modelling uses radial basis function interpolation, which is mathematically related to machine learning but is not trained on examples. It is best described as ML-adjacent. Trained geological AI remains mostly at the research stage.
How often should a mine geological model be updated?
Ideally, update it whenever significant new drilling or mapping data arrives. Implicit modelling makes this practical, because surfaces rebuild automatically from the new data. Many sites refresh the model before each resource update or major planning cycle.
Can AI replace the geologist in a mine geological model?
No. Software speeds up surface building, but geological judgement still decides contacts, fault offsets and continuity. The geologist’s interpretation remains the main source of model quality.
Try It With Geocluster
It pays to understand where the line sits between ML-adjacent interpolation and genuinely trained geological AI before you invest engineering time. You also need to know which one actually solves your modelling problem. The Geocluster Research Harness helps teams make that call systematically. Read how the Geocluster Research Harness works to learn more.