Flying Blind: The New Era of Underground Inspections with Drone-Based SLAM LiDAR

Underground mining is one of the most dangerous industries, with incident rates exceeding those in construction or manufacturing. In the U.S. alone, 49 miners lost their lives in a recent year due to mining accidents. The hazards are numerous: unstable rock that can collapse without warning, falls from heights near shafts or stopes, exposure to explosives and heavy machinery, and extreme temperature and humidity swings deep below ground. These conditions make every inspection high stakes. Confined, dark tunnels and caverns force inspectors to work “blind,” often literally in the dark, risking life and limb to gather essential data on ground conditions or infrastructure integrity.

Regular inspections are not optional – they are mandated to prevent tragedies. Regulations like the Mine Safety and Health Administration (MSHA) rules in the U.S. require four annual inspections for every underground mine. Inspectors traditionally must venture into recently blasted stopes, unsupported tunnels, or hazardous voids to look for loose rock, check ground support, or survey blast results. Not only is this time-consuming and laborious, it puts personnel in harm’s way. A mine engineer from Vale describes how a typical cavity monitoring survey with a tripod-mounted Cavity Monitoring System (CMS) can take two people up to 4 hours – including hauling gear to the site, setting up near an open hole, and waiting for the scan to finish. Even then, the data quality often suffers from “shadows” (blind spots) because the laser scanner only sees from one position. Inaccessible nooks and crannies remain unmapped, and critical details can be missed. The pain points are clear:

• Safety Risks: Inspectors face falls, rockfalls, and exposure to dust or gases while working in confined, dark spaces. Post-blast areas can be especially unstable, posing grave danger in the minutes and hours after a blast.

• Difficult Access: Reaching a target inspection point deep underground might require navigating complex tunnel networks, climbing into stopes, or dangling equipment into voids. Some areas are simply unreachable by foot or scaffolding.

• Slow, Costly Surveys: Traditional manual or stationary laser measurements demand significant setup and shutdown time. As noted, a single stope scan can occupy a crew for half a shift. This leads to production delays and increased labor costs.

• Incomplete Data: Static scanners like CMS or terrestrial LiDAR have limited viewpoints. They often leave “occlusions” – unseen zones behind rock corners or pillars – resulting in shadowy point clouds that omit areas of interest. In mines, missing data can mean misjudging ground conditions or miscalculating ore volumes.

• Regulatory Pressure: Mine operators must document conditions and hazards to comply with safety regulations. Failing to inspect thoroughly isn’t just a safety risk – it can mean non-compliance with MSHA or other standards, which require that hazards be identified and mitigated.

Health and safety managers and geotechnical engineers know these challenges well. The question is how to perform inspections without sending humans into harm’s way and while obtaining better data, faster.

SLAM LiDAR on Drones: New “Eyes” in the Darkness

Fortunately, recent advances in drone technology and SLAM-based LiDAR mapping are transforming how underground inspections are done. The Emesent Hovermap ST‑X system exemplifies this shift. Hovermap ST‑X is a LiDAR sensor paired with an onboard computer running SLAM (Simultaneous Localization and Mapping) algorithms. Mount it on a drone, such as DJI’s enterprise-class Matrice 350 RTK or the latest Matrice 400, and you essentially give the drone a 360° LiDAR “vision” and autonomy brain. This enables the UAV to fly in total darkness and complex environments by mapping its surroundings in real time and navigating without GPS.

The Hovermap ST‑X boasts impressive specs for underground use. Its LiDAR can see up to 300 meters in range and captures over one million points per second, with multiple returns per pulse to penetrate dust or vegetation. The unit is built for harsh conditions – a lightweight (≈1.57 kg) yet rugged package that is IP65-rated for dust and water resistance. In other words, it’s made to survive in dripping-wet drifts or dusty stopes where standard electronics would fail. Emesent’s proprietary Wildcat SLAM and autonomy software act as the drone’s autopilot and collision avoidance system. Hovermap creates a live 3D map as the drone flies, and uses it to avoid obstacles in any direction. The system even streams a live point cloud to the operator’s tablet on the surface, so you can watch the scan in progress and ensure coverage in real time.

Crucially, this SLAM-driven autonomy means the drone can go beyond visual line-of-sight (BVLOS) – around corners or down vertical shafts – and still know exactly where it is. “With a SLAM-equipped drone, the scanner is freely movable in and around the environment and well beyond visual line-of-sight. This enables near shadowless scans of the environment without exposing workers to any hazards,” says Matt MacKinnon of UAS Inc., an underground drone inspection provider. In essence, the drone can “fly blind” so humans don’t have to, navigating by LiDAR alone. It can hover steadily in a pitch-black stope, map every crevice with sub-centimeter precision, and return safely – all while the crew remains outside the danger zone.

The pairing of Hovermap with DJI’s Matrice 350 and 400 series drones is a game-changer for mining. These drones provide the heavy-lift capacity, endurance, and reliability needed for underground missions. The Matrice 400, for example, is an IP-rated, enterprise drone. When equipped with Hovermap, the M350 can be flown in either pilot-assist mode (a human pilot flies manually but with omnidirectional collision avoidance as a safety bubble) or in full autonomous mode using Emesent’s Commander tablet app. In autonomous mode, the operator simply taps waypoints on a map of the tunnel, and the drone intelligently explores and maps the area, auto-avoiding walls and drawpoints. 

DJI’s newer Matrice 400 takes it even further. This next-generation flagship drone offers up to 59 minutes of flight time and a 6 kg payload capacity, along with advanced obstacle sensing powered by an integrated rotating LiDAR and radar system. It’s engineered for extreme environments (rated IP55 and operable from –20 °C to 50 °C) – perfect for the temperature swings in mines. The M400’s heavy lift lets it carry the Hovermap ST‑X (only ~1.6 kg) easily, potentially alongside additional sensors like cameras or gas detectors. With its enhanced stability and situational awareness system, the M400 can confidently navigate cluttered drifts or stopes, further reducing the chance of collisions. In short, these DJI Matrice platforms act as sturdy mules and Hovermap is the “eyes and brain,” together enabling unprecedented autonomous flight underground.

From Hours to Minutes: Real-World Results Underground

The true proof of this technology is in the mining operations that have embraced it. Case studies from active mines show dramatic improvements in safety, speed, and data quality:

• Barrick Gold (Bulyanhulu and Kibali Mines): These large underground gold mines in Africa struggled with incomplete cavity monitoring. Shadowing was a “perennial challenge” – the traditional pole-mounted CMS scanners could not capture the full extents of big stopes, leaving blind spots in the models and causing surveyors to miss chunks of data. This inaccuracy directly impacted mine planning, as monthly volume reconciliations were off due to missing scanned areas. Barrick introduced drone mapping with Hovermap to solve the issue. Now, after a blast, a drone can be deployed into the stope to perform a 360° scan with no occluded zones. The result is a high-resolution, shadowless point cloud of the entire stope, allowing precise volume calculations and geotechnical analysis. Importantly, surveyors no longer need to approach open stopes or send instruments in on booms, greatly improving safety. As George Fouche, Barrick’s Chief Mine Surveyor at Kibali, put it: “With the CMS, you’ve just got one position… it doesn’t scan past that edge. By contrast, the Hovermap is going down and picking up all the small places the CMS didn’t.” The mine can now identify and remediate hazards inside stopes that would have been invisible before, while keeping personnel out of those voids.

• Glencore – Mount Isa Mines (Australia): Mount Isa’s underground operations began using Hovermap as an alternative to the traditional CMS for stope surveys. The impact was immediate: data collection that used to take hours was reduced to minutes, and the point clouds produced were richer and “shadowless”, capturing the full geometry of stopes for both survey and geotechnical purposes. With accurate 3D models of the excavated voids, engineers can refine their blast designs and optimize ore recovery, knowing the true shape of each stope. Hovermap is also used at Mount Isa to scan development drifts for convergence (deformation) and check the integrity of ground support over time, tasks that previously required manual tape measurements or laborious total station setups. By enabling quick, remote scans, the mine improved both its productivity and safety performance.

• Alpha Safety (Drill-and-Blast Contractor, 2024): In a recent project, a mining contractor (Alpha Safety) needed to monitor drill and blast progress under strict safety regulations. They deployed a drone outfitted with Hovermap to scan blasted stopes and even measure drill blast holes in areas inaccessible to personnel. The Hovermap captured the over-break/under-break around holes and provided a comprehensive 3D model of the rock face for post-blast analysis. Impressively, the team scanned the blast holes quickly and accurately, delivering a full 3D report to their client while meeting strict regulatory requirements for data quality and safety. This meant the mine could resume operations faster after a blast, since the stability of the area was verified by the scan data, all without sending ground crews in to do scaling or face mapping until it was deemed safe.

These examples are echoed by many others. BHP’s Olympic Dam mine was an early adopter of Hovermap, using it to map previously inaccessible old stopes and even a large “mega-cave” subsidence area, allowing monitoring of ground movement from a safe distance.

In South Africa, after a serious fall-of-ground (FOG) incident, a Hovermap drone was sent into the collapsed area to assess conditions so the mine could plan a safe re-entry and resume production. The common theme is that drone-based SLAM LiDAR has enabled mines to get vital insights faster and without exposing people to danger. What once required miners to venture into unstable cavities with scanners or measuring tools can now be done by a robot. And the data is often better than what was obtainable before.

New Era vs. Old Ways: A Brief Comparison

It’s worth highlighting how drone-based SLAM LiDAR inspections stack up against traditional approaches:

• Worker Safety: Perhaps the biggest advantage – no human needs to enter the hazard zone. Drones can fly into post-blast stopes or unsupported tunnels long before it would be safe for a person. This eliminates the risk of falls, rockfalls, or exposure to toxic gases during inspections. By keeping workers out of harm’s way, mines directly reduce the chances of accidents (a priority given that underground mines account for a large share of serious incidents). As one industry article quipped, now “Hi Ho, Hi Ho, off to work drones go!”, implying robots head into danger so miners don’t have to.

• Coverage & Data Quality: Mobile LiDAR scanners produce more complete data than stationary ones. A recent independent study in a 120 m long mine tunnel compared old-school terrestrial laser scanners (TLS) with modern SLAM LiDAR units. Emesent’s Hovermap ST‑X outperformed the static scanners on overall point cloud accuracy, and the mobile scans had far fewer gaps because the scanner could move around and see all angles. No more big blind spots behind pillars or under ledges – the drone can circle or hover at different points to get a true 3D capture. The result is “near shadowless” models, as seen in the Barrick case where Hovermap picked up all the small voids the single-point CMS could not. For geotechnical engineers, this means better inputs for stability analysis; for mining engineers, it means more precise volume and dilution calculations.

• Speed and Efficiency: Time is money in mining, and here the drone+LiDAR combo shines. Traditional cavity monitoring might halt production for 1–4 hours for a single scan. In contrast, a Hovermap drone can often scan the same stope in 10–15 minutes of flight, and multiple stopes can be done in a shift because setup is minimal. Mount Isa’s team reported cutting survey times by an order of magnitude. Moreover, fewer personnel are needed – often a single pilot/operator can do what used to require a crew of two or more. This efficiency not only saves labor costs but also minimizes disruption to operations (e.g. less downtime waiting for clearance to re-enter a production area).

• Autonomy & Reach: Drones can go where people or tripod scanners simply cannot. They can fly beyond line-of-sight, down winding declines or up raises, far past the point a remote control signal would normally drop – Hovermap’s onboard autonomy takes over navigation. In practical terms, this capability has opened up mapping of vertical shafts, ore passes, and stopes that have no access from the bottom. Mines have used Hovermap to map up ore chutes and down raises to check for blockages or estimate backfill volumes, tasks that were impossible or highly dangerous before. The “virtual safety shield” around the drone prevents it from crashing even if comms are lost, so it can boldly explore unstable areas and find its way back.

• Versatility: The Hovermap unit itself is versatile – it can be quickly detached from the drone and used as a handheld or vehicle-mounted scanner for other duties on site. For example, after doing an aerial scan of a stope, staff can mount the same LiDAR on a pole to map an underground drive on foot, or on a rover (like Boston Dynamics’ Spot robot) to inspect an old inaccessible tunnel. This multi-use flexibility means a single investment in SLAM LiDAR covers many surveying needs (mapping stopes, drives, shafts, surface stockpiles, infrastructure, etc.), whereas older systems often had specialized single-purpose equipment.

To be clear, traditional inspection methods still have their place – for instance, a well-trained geotechnical engineer’s visual observation is invaluable, and high-precision static scanners might still be used for certain detailed tasks. However, the gap is closing fast. In many scenarios, drone-based LiDAR not only matches traditional methods but exceeds them. The Czech tunnel study mentioned above underscored that Hovermap’s mobile scan was as accurate as the best tripod TLS (Trimble & FARO) and far quicker. The mining industry is seeing that what used to be “flying blind” – sending in drones without GPS – is now a reliable, repeatable process yielding rich 3D data.

Aligning with Safety Standards and Future Outlook

Adopting drone-based SLAM LiDAR for inspections isn’t just a technical upgrade; it’s also a step toward stronger safety compliance and risk management. Regulatory frameworks and industry standards are increasingly supportive of these technologies:

• MSHA and Government Regulations: As noted, MSHA requires regular inspections and has thousands of federal inspectors making sure mines comply. The use of drones can help mines meet these requirements more thoroughly. In fact, some jurisdictions explicitly allow drone inspections as part of compliance. For example, the state of Alabama modified its rules to count an aerial drone survey as a partial mine inspection for certain surface mining operations. Around 42 U.S. states have their own mine safety agencies, and many are updating policies to incorporate new tech. By using an autonomous drone to inspect a risky area, a mine can demonstrate proactive safety measures to regulators – essentially showing that they’re going above and beyond traditional practices to ensure a safe workplace. This can be invaluable during safety audits or incident investigations.

• ISO 19434 and Safety Frameworks: ISO 19434:2017 is an international standard for classifying mine accidents by cause and type. It explicitly notes that most serious mining accidents occur in underground operations. Falls of ground, ground collapses, and machinery-related accidents rank high in these classifications. By removing people from the most hazardous inspection situations (like going under unsupported ground to check rock conditions), drone-based inspections directly target some of the top causes of accidents identified in frameworks like ISO 19434. In essence, this technology is a controls measure that fits into the hierarchy of hazard control: it eliminates exposure. Implementing such measures not only aligns with the spirit of global safety standards but can also feed into an operation’s internal safety KPIs (e.g. reducing person-hours in high-risk zones). Some forward-looking mines are even including drone scans as part of their documented ground control plans and safety management systems, which might help in achieving certifications or meeting guidelines of bodies like the Initiative for Responsible Mining Assurance (IRMA).

• Data Integrity and Audits: From a health & safety manager’s perspective, having detailed 3D documentation of an underground workplace is a huge advantage. It provides an objective record of conditions that can be referred to in case of any incident or for routine audits. High-resolution maps of rockbolt placement, for example, or measurements of convergence over time can be extracted from these point clouds. This makes demonstrating compliance (say, with support installation standards or tunnel clearance requirements) much easier and more credible. The digital twin of the mine becomes a living safety document.

Looking ahead, the “flying blind” era of underground drones is likely just the beginning. As more mines adopt drone-based SLAM LiDAR, we can expect even tighter integration with mine operations. Autonomous drones may be permanently stationed underground in charging docks, ready to launch after every blast to scan the heading. The data they capture can feed directly into mining software for near-real-time decision-making – everything from adjusting the next drill design to alerting engineers of a new geotechnical hazard. This continuous mapping could also play a role in future predictive safety: for example, AI algorithms might analyze successive point clouds for subtle movements or stress signs in the rock, providing early warning of collapses.

Moreover, the partnership between companies like Emesent and DJI highlights how the drone industry and mining industry are converging. DJI’s release of the Matrice 400 – with features clearly tailored for tough, GPS-denied environments – shows that major manufacturers see a big opportunity in underground applications. We’re likely to see drones with even more autonomy, better dust-proofing and lighting, and specialized sensors (like thermal cameras or gas detectors) flying alongside the Hovermap LiDAR. All of this will further reduce the need to send humans into dangerous situations.

Conclusion: Safer Mines Through Smarter Eyes in the Sky

In the new era of drone-based SLAM LiDAR inspections, mining operations no longer have to choose between safety and good data – they can have both. What started as a daring idea (flying a drone in a pitch-black tunnel, far from any GPS signal) has matured into a proven practice that top-tier mines around the world are adopting. Geotechnical engineers now get rich, shadow-free 3D models of the rock mass to guide design and monitoring. Health and safety managers can breathe easier knowing critical inspections are done without putting boots on unstable ground. And mine managers see the benefits in productivity and compliance metrics alike.

“Flying blind” is now a phrase with a very different meaning: a SLAM-equipped drone can fly blind in the dark, yet map everything around it – effectively giving us vision where we had none. In an industry that has always been on the cutting edge of technology (from mechanization to automation), drone-mounted LiDAR is the latest tool making mines safer and more efficient. Underground inspections are taking off – literally – and the implications for mine safety and productivity are tremendous. As the case studies show, drones carrying Emesent’s Hovermap on DJI platforms are already helping mines big and small to navigate the unknown, mitigate risks, and make better decisions based on solid data. In this new era, we’re not just flying drones underground; we’re also flying past the old limitations, finally seeing clearly in the dark depths of our mines.