Closing the Positioning Gap in Hard Rock Mining

By: Advanced Navigation

In modern mining, the central challenge beyond extraction is one of knowledge.

While the physical movement of material is the visible output of a mine, the data associated with that movement is what drives profitability. Specifically, achieving certainty about which stope or draw point a specific tonne of ore originated from (and exactly where it went) is the bedrock of effective ore grade control, production reconciliation, and long-term mine planning.

However, for decades, this chain of custody has been undermined by the fundamental physics of the underground environment. In the absence of GNSS (Global Navigation Satellite Systems), underground positioning has remained an exercise in estimation rather than precision.

As the mining industry looks to the future, the reliance on legacy tracking methods has created a data gap.

Closing this gap requires a re-evaluation of how vehicles navigate in GNSS-denied environments, moving away from mechanical estimation toward sensor fusion technologies that can operate independently of physical infrastructure.

The Engineering Bottleneck and Why Underground Navigation Drifts

To understand the solution, one must first examine why underground positioning is challenging. In open-pit operations, a vehicle simply pings a satellite to verify its location. Underground vehicles must rely on dead reckoning by calculating current position based on a previously determined position.

Traditionally, this is handled by an inertial navigation system (INS). An INS uses gyroscopes and accelerometers to track motion. However, all inertial sensors drift over time. To counteract this, the INS needs an external velocity update to constrain the drift.

In the harsh environment of a hard rock mine, this approach faces severe mechanical limitations:

• Underground loaders (LHD) and haul trucks operate on loose, uneven, or wet surfaces. Wheel slip can range between 2% and 10% during standard operation. If an INS relies on wheel rotation to measure distance, and the wheels spin without traction, the navigation system believes the vehicle has moved further than it has. Over a kilometer, a 10% error equates to a 100-meter discrepancy.

• Visual odometry (using cameras) or LiDAR-based SLAM (Simultaneous Localization and Mapping) offer alternatives, but they face their own challenges. Underground tunnels often lack visual features (textureless walls), and heavy dust or water vapor can blind optical sensors. Furthermore, situations where a long tunnel looks identical for hundreds of meters can cause perception algorithms to fail in determining longitudinal progress.

Consequently, many mines have relied on installing extensive physical infrastructure, such as Wi-Fi tags or RFID beacons, to reset position data. However, this infrastructure is costly to maintain and is susceptible to damage from blasting.

A New Approach

To address the limitations of mechanical odometry, engineers have turned to aviation and military methodologies. The solution gaining traction in the sector involves decoupling velocity measurement from the vehicle’s drivetrain entirely.

This involves the integration of an INS with a Laser Velocity Sensor (LVS).

Compared to a wheel encoder, an LVS utilizes the Doppler shift of laser light. By emitting beams at the tunnel floor and walls, the sensor measures the frequency shift of the reflected light to calculate relative velocity. Consequently, because this measurement is optical rather than mechanical, it is immune to the effects of wheel slip, slide, or tire compression.

When an INS is fed accurate, three-dimensional velocity data that is independent of the vehicle's physical traction, the accumulation of error is drastically reduced. This sensor fusion allows a vehicle to track its position with high fidelity over long durations without requiring external infrastructure or satellite signals.

Case Study: BHP Deep Mining Challenge

This technology was stress-tested during the BHP Deep Mining Challenge, conducted at the Callio Mine in Pyhäsalmi, Finland.

At a depth of 1,440 meters, the test environment represented a "worst-case scenario" for navigation: no GNSS, no external infrastructure, and complex route geometries. The

objective was to determine if a vehicle could maintain accurate positioning over a 6-kilometer drive cycle using only onboard sensors.

The technology tested was a hybrid system from Advanced Navigation, fusing a Fiber-Optic Gyroscope (FOG) INS with an LVS. The results provided a benchmark for modern infrastructure-independent navigation:

1. Positional accuracy: The system demonstrated a position accuracy of approximately 0.55 meters over the 6-kilometer course. For mine operators, this level of precision distinguishes not just which tunnel a vehicle is in, but which specific side of the draw point it is loading from.

2. Heading stability: Despite the high latitude of Finland (where the Earth's rotation significantly impacts heading alignment), the system maintained heading accuracy below 0.1 degrees. This prevents "navigational creep," ensuring that the digital path of the vehicle aligns with the physical map.

3. Attitude data: The system recorded roll and pitch accuracy to 0.01 degrees. While primarily used for navigation, this data serves a secondary purpose of monitoring road quality.

From Estimation to Reconciliation

The transition from rough estimation to precision tracking has implications that extend beyond simple logistics. It fundamentally changes how a mine manages its resources.

The Digital Chain of Custody

The primary driver for high-precision positioning is ore reconciliation. When a loader collects material, the operator needs to know if that bucket contains high-grade ore from Stope A or waste rock from the adjacent wall.

Without precise tracking, this is often educated guesswork, leading to the challenge of processing waste rock or sending high-grade ore to the waste dump. High-fidelity tracking creates a timestamped digital record for every movement. This allows planners to link specific tonnage back to the exact extraction point, closing the loop between the geological model and the mill.

Enhancing Underground Safety

Accurate positioning is a prerequisite for advanced Collision Avoidance Systems (CAS). Current systems often rely on proximity detection that can be "noisy," generating false alarms that desensitize operators.

By utilizing high-fidelity velocity and position data, next-generation CAS can calculate trajectories with greater certainty. This data supports dynamic geofencing, where virtual exclusion zones around personnel or hazards can trigger alerts or vehicle interventions with high reliability.

The Future of the Connected Mine

The evolution of underground positioning is moving away from a reliance on external beacons toward more reliable and accurate navigation.

The industry is securing the foundational data layer required for the digital mine by solving the physics problem of wheel slip and drift through laser velocity sensing. This shift allows operators to move from reactive planning based on lagging data to proactive optimization based on real-time, verifiable facts.