Making the Numbers Work for Us...

By Shelby Yee, CEO, GroundedAI

There’s no question that mining is in the midst of a data boom. In fact, so much data is being captured and generated underground that the industry now faces a new challenge – what do we do with it all?

Traditionally, the mining industry has relied on a single powerful data source: drill core data. Miners have drilled thousands of metres into the earth, generating detailed logs and defining valuable resources, but at significant expense. This data has formed the foundation for all components that go into planning and developing a mine, including mine schedules, ground-control management plans, and most critically, a validated resource estimate that is defined by the geological block model.

Yet, as real-world conditions begin to challenge the models that guide plans and production, the need for a broader data arsenal becomes clear. Today, mines have a rich and varied toolkit at their disposal for collecting data, including digital technologies that are accepted standards and some that are emerging as powerful new methods.

Geochemical data tracks variations in rock composition, delineating ore bodies and guiding extraction strategies, while geophysical sensors monitor dynamic underground events – like blasts and rock bursts – to maintain operational safety. Automated drill core analysis delivers high-fidelity geological logs that streamline resource evaluation and planning, giving hope for a more accurate and consistent representation of the orebody. High-resolution point clouds provide realistic 3D representations of active workings, offering essential spatial context.

The issue isn’t the lack of data – it’s the disconnect between the data captured and action.

Technology revolutions have followed a clear pattern: first, new tools increase the volume, precision, and types of data we can collect. Then, solutions emerge to make that data useful, integrating it into decision-making workflows and action. We’re at the second stage of that transition. Within the mining industry, data is being captured with the goal of understanding what’s happening underground, in real-time, and in great detail, as we continue to operate mines that are increasingly remote and at extreme depths.

The question now is: how will we make mining data truly actionable?

A flood of data, a drought of insights

For many in the industry, this is where data usage stalls.

So, what happens to this rich trove of data? Much of it – whether geological, geotechnical, or operational – ends up stored away in the digital equivalent of a dusty backroom, waiting for someone to connect it with an operational problem. Centralized software databases and solutions are being widely adopted across the industry and are forming critical infrastructure to safely store and connect data captured underground. Mining companies have also invested millions of dollars into new LiDAR and other data capture technologies. Millions more are spent hosting datasets on servers or cloud-based databases.

Despite significant investments in LiDAR and centralized databases, a critical gap remains in seeking actionable insights from these massive, unruly datasets. Few software solutions can fully harness the potential of raw, 3D-coloured point clouds, as their enormous file sizes and intensive computing demands often overwhelm standard systems at mine sites. While you might still have a few impressive 3D views of your underground workings, without efficient tools to process and integrate the data you’ve invested so much in collecting, what practical decisions can you truly make?

Imagine this: you’re a production geologist starting your day with a list of production faces to inspect. Your mission is to quickly map each face, log sample data, and provide clear instructions to the mining crew. A colourized, high-definition point cloud may seem like the perfect canvas for mapping and delineating geological structures – after all, it provides an accurate 3D representation of the rock face. On your desktop, the point cloud looks incredible – millions of data points capturing every contour in the rock face with millimeter precision. You can rotate it, zoom in, and maybe you take a screenshot for good measure. But now what? What decisions does this actually inform?

In practice, it introduces new challenges. Processing the data, mapping onto the point cloud surface, and integrating the mapped representation of the production face into the daily report becomes a time-consuming task. More importantly, mine operations rely on geologists to simplify and interpret rock faces, not just replicate them in extreme detail.

Beyond geological mapping, sample data requires accurate localization in 3D space, ore and waste calls need to be reported, and any instructions given to the mining crew needs to be documented. While there’s no question that using and integrating high-resolution point clouds enhances accuracy and auditability, the associated data handling requirements often add to a geologist’s workload rather than streamline it.

From data collection to decision-making

There has been a boom in technology being brought to the mining industry, from startups to large OEMs, and industry-adjacent software providers to internal innovation groups. The most successful have not only created or leveraged a rugged suite of sensors (designed specifically for the demanding conditions of the mining industry) but have identified specific use cases and outputs that better inform specific portions of the production cycle. The differentiating value of these solutions lies in removing the special skills and effort that is typically required to process point cloud data, and distill it into actionable information with minimal user intervention 

“This is not dissimilar to other industries,” says Duncan Turner, General Partner of SOSV, and the industrial hardware program, HAX. “LiDAR sensors are now commoditized. The novelty is now in the actual manipulation and utilization of the point clouds into insights.”

Take the mining tech start-up Point.Laz, for example. Their Lazaruss system provides automated 3D scanning of mine shafts using high-resolution LiDAR and 360° imaging. While the raw scan data is highly detailed and computationally demanding, Alexandre Grenier, Founder & CEO of Point.Laz recognized the most valuable output is a simplified PDF report illustrating shaft changes over time, such as cracks, corrosion, and convergence. The information is easy to interpret and sharable across departments, from engineering to mine operations. Instead of relying on individuals to load massive datasets, which can take hours or days to view, they have found a way to distill high-resolution point cloud data into clear and actionable insights that improve mine safety and operational efficiency.

Similarly, mining tech startup LoopX leverages state-of-the-art LiDAR sensors to enhance mining operations. Chao Yu, Founder & CEO of LoopX, recognized that exhaustively mapping an entire underground mine can be overwhelming and counterproductive. Instead, LoopX focuses on using LiDAR to provide real-time insights into mine operations, including localization and tracking of mining equipment, as well as assessing the efficiency of material handling processes.

At GroundedAI, we leverage point cloud data to improve geological block models and ground control management plans. We’ve seen firsthand how a digital mapping approach that directly integrates with industry-leading software can transform operational insights. Mines using our underground mapping solutions have reduced data capture to reporting timelines by over 50%, freeing up geologists and engineers to focus on interpretation and decision-making rather than data wrangling. The result? Faster turnaround on geological models, more accurate ground control decisions, and ultimately, more efficient mining operations.

What’s next? Mining’s data-driven future

The question is no longer whether we have enough data. It’s whether we know what to do with it.

The next challenge is breaking down data silos, consolidating all the information captured, and taking action on newly available insights. Mining operations have spent millions in investing in cutting-edge data capture solutions, but the real value lies in software that brings these datasets together. The companies that succeed in the next phase of mining technology will be those that seamlessly integrate with existing platforms and systems, allowing mines to move away from raw data to operational insights effortlessly.

Mining is no longer a data-poor industry. The winners will be those who master the art of translating unruly datasets into actionable intelligence – delivering insights to the right person, at the right time, in the right format.