← Sticker Wall AGENT · DETAIL

Plotlogic OreSense: Mining AI mapping system, single-mine annual subscription model securing clients like BHP

Workflow: The team transports hyperspectral imaging and LiDAR equipment to the mine site, deploying it to pit walls, underground h

AGENT

Key Fields

FIELD STAMPS

IndustrySaaS / Enterprise Software

RegionGlobal（澳大利亚）

ScaleSME

ChannelOnline

🔧 Workflow

The team transports hyperspectral imaging and LiDAR equipment to the mine site, deploying it to pit walls, underground heading faces, or ore piles within hours to a day. The system automatically scans for 3-5 minutes, and within 15 minutes outputs 3D maps of ore and waste rock boundaries, grades, and mineral compositions. Mine dispatchers make short-cycle mining decisions based on this, with human engineers acting as the final arbiters. System availability is maintained above 95%, supporting four deployment forms: vehicle-mounted, fixed, bucket-integrated, or containerized. Scan results directly integrate into the mining company's existing scheduling and ore blending workflows, repeatedly producing new boundary maps by shift every day. The equipment can also continuously output the distribution of harmful substances such as clay in ore pile and stockyard scenarios.

🛠 Setup Requirements

Requires hyperspectral and LiDAR hardware integration capabilities, machine learning mineral classification algorithms, and on-site mining engineering experience. Hardware reaches IP67 protection and weighs under 40 kilograms. This is heavy industrial hard tech requiring a combined geology and sensing team, with an R&D plus field validation cycle measured in years, making it unsuitable for lightweight individual starts. Building it also requires handling the collection and labeling of a 400-2500 nanometer spectral library, fine-tuning classification models迁移 (transfer) for different minerals such as gold, iron, lithium, and copper, and completing edge deployment on the mine network to keep data on-site.

🧰 Toolchain

🔧 Hyperspectral imaging sensor (400-2500 nanometers)
🔧 LiDAR
🔧 Proprietary AI mineral classification engine
🔧 Containerized or vehicle-mounted deployment hardware
🔧 Mine network infrastructure for cloud training and edge inference

💰 Revenue

Annual subscription fee based on mine deployment, typically with a minimum of 12 months, public pricing not disclosed. Referring to its 12-month support contract signed with Pilbara Minerals in 2023 and the $18 million funding round in February 2026, the annual subscription for a single mine is estimated to be in the hundreds of thousands of Australian dollars. Fees are tiered based on mine scale, configuration (pit face, underground, stockyard), and value delivery. Hardware is provided with the subscription rather than as a one-time sale; the early philosophy was no results, no fee, and renewals create compounding revenue.

💸 Cost

Costs mainly consist of hardware manufacturing, sensor procurement, and on-site deployment and maintenance, with system availability maintained above 95%. On the software side, costs include machine learning training and cloud computing expenses, with exact figures undisclosed. Estimated roughly at hundreds of thousands of AUD per mine annual subscription, the gross profit of a single project depends on hardware reuse and remote operations efficiency. Marginal costs decrease significantly after containerized mass production and multi-mine reuse, which is also the primary investment direction for the $18 million financing.

⏱ Time Investment

Corporate operation is a full-time commitment, with teams covering algorithms, on-site engineering, and marketing roles. A single mine deployment takes only a few hours to a day, followed primarily by remote operations. The annual rhythm involves parallel deployment of multiple mines plus model iteration by mineral type. Overseas projects also require compliance and localization adaptation according to host country mining regulations.

🚀 Getting Started

Novices are not suited to directly copy the hardware route. A viable first step is to enter a mining tech company or geological survey unit in an AI algorithm role, accumulate experience in hyperspectral or 3D geological modeling, and then break in through data labeling, model fine-tuning, or mine digitalization consulting. To validate demand, one can first turn public mining aerial surveys and core photos into ore recognition demos, pitch three-month paid pilots to small mining companies, and consider heavy hardware formats only after the model works.

🔑 Keys to Success

✅ Delivering results in 15 minutes instead of days of laboratory assays, offering an extremely strong value proposition
✅ Annual subscription per mine rather than one-time hardware sales, creating compounding renewal revenue
✅ Clients include major mining companies like BHP and Pilbara Minerals, providing strong endorsement
✅ Commercial operation in mines across multiple countries including New Zealand, Indonesia, Brazil, and Botswana, with system availability exceeding 95%, proving cross-region replication capability

⚠️ 风险

⚠️ Heavy hardware and heavy sales model with long deal cycles per mine; cash flow relies on financing
⚠️ No personal side-hustle or lightweight SaaS version; individuals cannot directly replicate this model
⚠️ High customer concentration; losing a single major mine contract has a significant impact

📌 Real Cases

📌 Secured a 12-month contract with Pilbara Minerals in 2023 to support lithium mine sites in boosting ore output and reducing carbon emissions
📌 Announced $18 million in financing in February 2026 to expand hyperspectral imaging to more mines
📌 Completed field tests of hyperspectral imaging technology at Western Australian gold and iron mines in March 2021, verifying the industrial reliability of ore and waste rock boundary recognition

Sources