Skip to main content

AI Demos by industry

Mining & critical minerals AI Demos

Explore mineral prospectivity, geological interpretation, and spatial evidence for resource assessment. Review measured data and model rankings while retaining their scientific limits. Browse interactive demonstrations relevant to this area, with methods, screenshots, and resource links.

Browse by industry

Choose an industry to explore relevant demonstrations. Research & scientific AI brings together methods used across industries.

5 of 5 entries

Browse entries by name

Geo Modeling

The geological modeling example brings surface terrain and subsurface structures together in a synthetic three-dimensional field. Wells, horizons, faults, and temperature intervals can be examined in the same scene to understand their spatial relationships. This integrated perspective helps geoscience teams explain subsurface geometry and discuss how well information relates to a broader geological interpretation.

  • Combines wells, geological horizons, faults, terrain, and temperature intervals in an interactive 3D scene.
  • Relates well trajectories and depth-dependent temperatures to their surrounding geological context.
  • Provides layer controls, terrain opacity, vertical relief, and well inspection.
  • Provides access to our open-source and proprietary ML for geospatial analyses.
  • Uses deterministic synthetic geology for illustration; the scene is not a field-validated subsurface interpretation.

Mine-Scale Lithium Prospectivity

This demo presents saved machine-learning prospectivity evidence for the Bald Hill lithium area in Western Australia. Ranking views are considered alongside measured assays, spatial support, cross sections, and interpreted pegmatite geometry. The combined evidence helps exploration teams examine where model rankings are supported by observations and how they behave across spatial holdouts.

  • Explores machine-learning lithium prospectivity rankings for Bald Hill, Western Australia.
  • Compares P70, P85, and P95 ranking heads with measured assays and an independent support mask.
  • Links 3D exploration to cross sections and explicitly interpreted pegmatite geometry.
  • Reviews spatial-holdout robustness; these saved rankings are not a grade, reserve, or orebody model.

Well Log Gap Filling

A geological well log records changes along depth, but gaps can interrupt the patterns needed for interpretation. This synthetic example uses neighboring measurements to estimate missing values while retaining the ordered depth structure of the record. Comparing the reconstructed log with held-back measurements helps users assess the behavior of the gap-filling method before considering a similar workflow for their own data.

  • Fills intentionally missing measurements along an ordered synthetic geological well log.
  • Uses our proprietary bounded RobustML neighbor method, with quick and stability-check configurations.
  • Compares observed and filled values using held-back RMSE and MAE.
  • Distinguishes the labeled precomputed example from a registered live workspace analysis.

Spatial Geology Gap Filling

Spatial gaps can obscure relationships between geological measurements collected across a field area. This compact synthetic example demonstrates how nearby observations can inform estimates at missing locations through a bounded, distance-weighted neighbor model. Viewing the reconstructed field alongside its evaluation metrics helps users judge both the spatial result and its agreement with held-back evidence.

  • Reconstructs missing values in a compact synthetic geological field sample.
  • Uses bounded, distance-weighted RobustML neighbors to estimate values at missing locations.
  • Shows spatial observations and predictions alongside held-back RMSE and MAE.
  • Retains a labeled precomputed fallback; live workspace execution is a separate action.

Feature Extraction (spatial)

Large geoscience datasets can contain recurring combinations of attributes that vary across a region. This demo uses NMFk to identify latent spatial factors and groups from selected measurements, with configurable preprocessing and factorization settings. Maps, biplots, and reconstruction views help users connect the extracted mathematical structure with the locations and attributes being studied.

  • Extracts latent spatial factors and clusters from selected dataset attributes with NMFk.
  • Configures the factor-count range, restart count, and normalization before factorization.
  • Provides matrices, latent maps, biplots, and reconstruction evidence to interpret recovered structure.
  • Supports optional regional interpolation after results exist; a factorization does not by itself establish a geological cause.