Q&A: Automation takes flight in WA mineral exploration

Mining automation is increasingly moving into the earliest stage of the mining lifecycle: exploration.

According to GlobalData's Autonomous Mining: Unlocking Innovation and Competitive Advantage ​​​​​​​report, advances in AI, high-resolution geophysical surveys and three-dimensional geological modelling are helping miners make faster, safer and more informed decisions, extending automation across the mining value chain.

One technology driving that shift is drone-mounted geophysics. Equipped with high-resolution magnetometers, drones can survey prospective ground quickly and cost-effectively, while AI-assisted interpretation and 3D modelling help geologists identify drill targets before committing significant exploration capital.

American Mineral Resources, a brand under AMLM, is applying this approach at its recently acquired Higginsville Gold Project in Western Australia (WA). Located within the Southern Norseman-Wiluna Greenstone Belt, just 9km from the historic Higginsville Mine, the four-tenement project sits in a proven gold district with extensive historical drilling, rock-chip sampling and existing infrastructure.

Higginsville's geology makes it well suited to magnetic surveying. Gold mineralisation in the region is closely associated with major shear zones and fault structures, which magnetometry can map beneath shallow cover, allowing explorers to refine targets before drilling.

CEO Ryan Cunningham explains how drone-mounted magnetometry, AI-assisted data interpretation and geological expertise are being combined to guide exploration at Higginsville, and why digital technologies are changing not only how mines operate, but how they're discovered.

Ryan Cunningham: The transformation has been extraordinary. When I started out, exploration was almost entirely boots on ground, people were relying on hand samples, paper maps, and expensive fixed-wing geophysical surveys that took months to commission and even longer to interpret. Today, a small team can deploy a drone magnetometry system over a target area in days, process the data in the cloud overnight, and have AI-assisted anomaly maps ready before the field crew has left site.

What hasn't changed is the geology. The rocks don't care about the technology, you still need experienced geologists who understand structural controls, mineralisation styles and the regional context. What's changed is how fast and how cheaply we can screen ground and prioritise targets before we put a drill bit in. That's a genuine step-change in capital efficiency, and for a company like American Mineral Resources, where we're managing a portfolio across multiple continents, that matters enormously.

What hasn't changed is the geology. The rocks don't care about the technology, you still need experienced geologists who understand structural controls, mineralisation styles and the regional context. What's changed is how fast and how cheaply we can screen ground and prioritise targets before we put a drill bit in. That's a genuine step-change in capital efficiency, and for a company like American Mineral Resources, where we're managing a portfolio across multiple continents, that matters enormously.

Ryan Cunningham: Magnetics is often the first geophysical tool you reach for in an Archaean greenstone setting. Gold mineralisation in these terranes is almost always structurally controlled, you're looking for shear zones, faults, lithological contacts and magnetic data is exceptionally good at mapping those features even under cover. It's cost-effective, non-invasive, and the datasets integrate cleanly with geology and geochemistry. For greenstone gold, it's close to indispensable.

Ryan Cunningham: Higginsville sits on the western limb of the Greater Eundynie Anticline, within 2km of the Zuleika Shear Zone – one of the major gold-controlling structures in the Eastern Goldfields of WA. The tenements host faulted, thrust-repeated mafic and ultramafic sequences that have been through at least two deformation events: compression and later fracturing.

What we're specifically targeting are the north-west-trending fault systems that run parallel to the Zuleika Shear, and the cross-cutting north-east faults. The nearby Challenge-Swordsman operation produced 250,000 ounces from this structural setting in the 1990s, and the Higginsville Mine itself has over 1.1 million ounces of historical production. Magnetics will help us map those fault corridors precisely and identify where they intersect – those intersections are classically where gold deposits concentrate in this belt.

Ryan Cunningham: At this stage we work with specialist contractors. The Australian Goldfields market has excellent geophysical service providers with deep experience in Archaean greenstone surveys, and it makes more sense to leverage that expertise than to replicate it in-house.

What we bring in-house is the geological interpretation - knowing what questions to ask of the data and how to integrate it with the structural and geochemical picture.

Ryan Cunningham: Absolutely, in most cases. The cost per line-kilometre for drone mag has come down dramatically, and the ability to fly safely at low altitude gives you better resolution for the same budget. The real question is whether the survey is designed well enough to answer the geological question you're asking. A cheap survey that doesn't resolve your target is money wasted; a well-designed drone mag survey that identifies a drill target is one of the best returns on exploration capital you can get. Beats helicopters on cost hands down.

Ryan Cunningham: The raw data goes through processing – levelling, filtering, and gridding – before you start interpretation. We look at total magnetic intensity and reduced-to-pole maps, then move into analytical signal and tilt derivative products that help resolve edges and contacts. That interpretation gets overlaid onto the geology and any existing geochemical data. Where anomalies align with favourable structural positions and geochemical signatures, you have a drill target.

Ryan Cunningham: The processing is largely automated – the algorithms are well-established. But interpretation is still very much a human exercise. Pattern recognition in magnetic data requires understanding the regional geology, structural history and the mineralisation model for that specific terrane. AI tools are starting to assist with anomaly detection, but the experienced geologist sitting with the data asking “why does this anomaly exist, and does it make geological sense?” is still irreplaceable.

Ryan Cunningham: Right now, the clearest value is in data integration and target generation – taking large, heterogeneous datasets and identifying spatial correlations that human analysts might miss or take weeks to work through. AI is also showing real promise in drill core logging, where computer vision can identify lithological and structural features with consistency that's hard to maintain over thousands of meters of core.

I'm more cautious about AI as a primary decision-making tool for drill targeting. The geological reasoning still needs to be there – AI can surface candidates, but experienced geologists need to ground-truth them against the actual geology before capital gets committed.

Ryan Cunningham: Increasingly, yes. 3D inversion converts your magnetic data into a volumetric model of subsurface rock properties – a much more intuitive picture than 2D maps alone. Reliability depends heavily on the quality of input data and the geological constraints you feed in. When constrained by real drill data and good geological understanding, it can be a powerful predictor. Like any model, it's a hypothesis to be tested, not a fact to be trusted blindly.

Ryan Cunningham: To be transparent, American Mineral Resources acquired the Higginsville tenements in April 2026, and our own fieldwork programme is planned to commence in the second half of 2026. The historical geophysical and geochemical work was conducted by previous tenement holders. Our immediate focus has been integrating that historical dataset, including the 152-hole air-core programme and existing geophysical coverage to define priority drill targets before we go back into the field.

Ryan Cunningham: The historical data confirms the tenements are positioned on a highly prospective structural corridor. The north-west-trending fault system parallel to the Zuleika Shear and the cross-cutting north-east faults are clearly expressed in the geophysical data. Rock chip sampling from the Rainbow Waters prospect returned nine samples above 1.2 grams per tonne (g/t) of gold with a peak of 8.26 g/t, and surface prospecting has recovered nuggets up to 27 ounces, clear evidence the structural and mineralisation system is active across the tenement package.

Ryan Cunningham: The historical dataset has directly informed our target prioritisation. We have 164 historical drill holes, multiple prospect areas with rock chip sampling, and geophysical data to integrate. Our drilling programme, planned for the second half of 2026, will test the structural intersections identified in that combined dataset, particularly the north-west/north-east fault intersection zones that are the classic loci for gold deposition in this belt.

Ryan Cunningham: Completion of target definition and drill programme design; commencement of drilling at priority targets; and initial assay results. Higginsville is a cornerstone of American Mineral Resources' global footprint – we now operate across 12 projects in seven jurisdictions on four continents, and the Eastern Goldfields is one of the world's premier gold addresses.