Highlights

  • The business states it has started discussions with prospective mining partners about pilot deployments that could generate revenue.
  • No pilot agreement has been signed, and further detail will follow only once one is executed.
  • According to the company, the device combines LiDAR scanning, three-dimensional mapping, spatial positioning and onboard analysis.
  • The technology sits within Stealth Technologies Pty Ltd, a wholly owned investment of Strategic Elements.

Strategic Elements (ASX:SOR) has told the market it has launched a pilot program for EdgeiQ Portable, a handheld system for mapping and analysing underground mines, and that it has reached a technical milestone in running artificial intelligence on the device itself.

Quarterly Developments

The business states it has started discussions with prospective mining partners about pilot deployments that could generate revenue. No pilot agreement has been signed, and further detail will follow only once one is executed.

What Edgeiq Portable Does

According to the company, the device combines LiDAR scanning, three-dimensional mapping, spatial positioning and onboard analysis. It is designed for underground settings where GPS is unavailable and connectivity may be limited, and it processes data on the unit without GPS, internet access or external cloud computing.

Strategic Elements says the product was developed through several hardware generations and five completed underground field trials with a Tier 1 mining company. It says those trials form the technical basis for the commercial pilot program and that the existing capabilities could support more than one distinct mining product.

The technology sits within Stealth Technologies Pty Ltd, a wholly owned investment of Strategic Elements. SOR is a registered Pooled Development Fund, a structure that it says provides eligible shareholders with certain tax concessions.

The Physical Ai Milestone

The business reports that an initial AI model has run directly on EdgeiQ Portable hardware with no extra computing equipment and no internet or cloud connection. It was deployed with Curtin University's Western Australian School of Mines. The model was trained on 1,000 simulated mining rounds, using ground conditions and drilling parameters, and aims to predict overbreak and associated dilution. Initial test outputs took about 100 milliseconds per inference.

SOR also says it has implemented input validation and a repeatable process for moving AI models onto the hardware, which it describes as a base for more advanced models. A Tier 1 mining company has approved access to operational mining data, and data supply has begun.

The planned timeline is further model development using that data in the fourth quarter of 2026, followed by a potential underground field trial in the first quarter of 2027. The company states this depends on technical progress and site access, and that the trial is not contractually committed.

Market Context And Estimates

The announcement cites GlobalData's estimate of about 2,300 operating underground mines worldwide, and the MINEDEX database's listing of 58 operating underground operations in Western Australia.

For its existing fixed-site product, EdgeiQ OnSite, the company points to a first commercial contract with a major global gold producer for underground haulage measurement, announced on 29 January 2026. It gives an internal estimate of a theoretical global market above A$500 million for that application. A separate internal estimate puts the cost of overbreak at more than A$2 billion a year globally. The company stresses that both figures are indicative market estimates, not forecasts of its own revenue.

Risks And Conditions

The company is clear about the limits of what has been shown. The AI model was trained on simulated data, and its predictive accuracy against real mining outcomes has not been established. It gives no assurance that a pilot agreement will be executed or that the technology will reach commercial deployment.

Risks it lists include technical performance, data quality, access to trial sites, partner arrangements, intellectual property, competition and market adoption.

What To Watch

The confirmed milestones are limited: the Q4 2026 model development using real mining data and the potential Q1 2027 underground trial. Any signed pilot agreement would be the first evidence of a paid deployment for the Portable product, and the company has said it will release further detail at that point.