Multi-sensor data fusion

Many independent lines of evidence, weighed into one explainable score — and an honest read on how far to trust it.

Sign up free

No single dataset finds a deposit. A satellite alteration anomaly, a gravity high, a geochemical spike — each one on its own is ambiguous, and chasing any of them in isolation is how exploration budgets get burned. The signal that matters is where independent lines of evidence agree. Data fusion is how MineDSS brings satellite, geophysics, geochemistry and mapped geology onto common ground and weighs them together, target by target.

One national dataset from many sources.

We standardise disparate public geoscience — across Australia, the USA and Canada — onto common ground, so your area is scored against the full picture, not a thin local slice.

Many sources

Geochemistry
soil, stream & rock samples
Geology & lithology
mapped rock units & age
Geophysics
gravity, magnetics, radiometrics
Satellite & Earth observation
ASTER, Sentinel, SRTM
Mineral occurrences
known deposits & mines

One harmonised national dataset

Standardised & searchable — no proprietary lock-in.

AustraliaAustraliaUnited StatesUnited StatesCanadaCanada

Explainable output

Ranked drill targets
The evidence behind each
An honest confidence flag

The layers we read.

Every dataset becomes an independent line of evidence, brought onto common ground and weighed together:

GEOLOGY

Geology & rock age

mapped rock units and their ages

GEOPHYSICS

Gravity & magnetics

buried structures and intrusions

GEOPHYSICS

Radiometrics

potassium, thorium, uranium

TERRAIN

Terrain

elevation, slope, landform

SATELLITE

Satellite radar

surface texture and structure

SATELLITE

Alteration

mineral signatures from satellite

GEOCHEM

Pathfinder chemistry

the elements that point to your mineral

How a national model combines the evidence

Each national model weighs every line of evidence and assembles them into a plain reasoning chain for every target — you can see which signals pushed a target up the list and which pulled it down. Crucially, it also tracks how much real evidence stands behind each call: a target backed by several agreeing datasets is flagged differently from one resting on a single signal or on ground unlike anything the model has seen before. Where the evidence is thin, our national models says so rather than bluffing.

Independent evidence, not one dataset

Agreement across datasets that fail in different ways is far stronger than any single anomaly — and far less prone to false positives.

A reasoning chain you can defend

Every ranked target carries the evidence behind it in plain terms, so it stands up to a board, a JV partner or a regulator — not a black-box score.

An honest confidence flag

Each target is marked for how much evidence supports it, so you know which leads to back and where a few field samples would sharpen the model.

Data fusion is at the heart of every MineDSS national model — and it runs on the same national dataset that a custom model is tuned on.

Common questions

Why fuse datasets instead of using the best one?

Because every dataset has blind spots. Satellite reads only the surface, geophysics resolves structure and property but not chemistry, geochemistry is patchy in coverage. Fusing them means a target has to satisfy several independent lines of evidence at once — which is exactly what makes it trustworthy.

What does 'a line of evidence' mean?

One independent reading of the ground — the mapped geology, a geophysical structure, a radiometric response, the terrain, satellite alteration or pathfinder chemistry. MineDSS weighs these independent readings together, the way a geologist cross-checks a map sheet against the geophysics and the assays, so the final ranking reflects the whole mineral system rather than any single measurement.

Is our national models a black box?

No — that is the point of it. Every target comes with the reasoning behind it and a flag for how much evidence stands behind that reasoning. You can check the work, defend the target, and see plainly where the model is confident and where it is guessing.

See it on your ground.

Draw an area, pick a commodity, and explore the ranked targets and the evidence behind each.

Sign up free