Key Takeaways

  • Corescan and CSIRO’s HyLogger are the two systems that dominate this space — there isn’t a fragmented field of competitors to evaluate.
  • CSIRO’s MyLogger is the concrete “neural network on spectra” step that makes this an AI workflow rather than just a scanning workflow.
  • These systems generate roughly 800,000 spectral samples per meter of core — the resolution is what enables high-confidence clay models, not just the sensor’s presence.
  • This is a solid “Yes” — mature, deployed technology, not a speculative research direction.

TL;DR

High-resolution clay models come from running core through a hyperspectral scanner (Corescan HCI or CSIRO HyLogger) at sub-millimeter resolution, then using a trained interpretation model (like MyLogger’s neural network) to convert the spectra into quantitative clay mineralogy at every sampled point.

How Do I Predict High-Resolution Clay Models With AI?

The poster’s distinction between a “Clay/Geomet Model” and “Hyperspectral input Models” is really about pipeline stage: this node is specifically the raw data-generation step that everything downstream depends on. Two systems dominate: Corescan’s Hyperspectral Core Imager (HCI-3/HCI-4, now sold through Epiroc) and CSIRO’s HyLogger, both of which run VNIR-SWIR (and in HyLogger’s case, thermal infrared) reflectance spectroscopy across an entire core at roughly 0.5mm spatial resolution — around 800,000 spectral samples per meter of core, vastly denser than any manual sampling regime could achieve.

That density is the entire point: it’s what makes a genuinely high-resolution clay model possible, since you’re not interpolating between sparse manual sample points but classifying a near-continuous spectral record. The step that turns raw spectra into usable mineralogy, historically requiring a specialist interpreter, is now partially automated — CSIRO’s MyLogger applies a trained neural network to HyLogger’s thermal-infrared data to produce a first-pass mineral/lithology log automatically, closing what used to be the real bottleneck in this workflow.

If you’re setting this up: the hardware and interpretation software are sold by different parties (Corescan/Epiroc vs. CSIRO’s research program), so plan for an integration project connecting your chosen scanner’s output format to your block-modelling package, rather than expecting a single vendor to hand you an end-to-end clay-model pipeline.

Try It With Geocluster

Working through the practical integration questions — which scanner fits your core logistics, how MyLogger-style interpretation compares to Corescan’s own analytics — is exactly the kind of grounded comparison Geocluster is built to help with.