Overcoming GNSS Blind Spots for Resilient Positioning in Autonomous Mining

By: Advanced Navigation

For mine operators and autonomy teams, the shift to Autonomous Haulage Systems (AHS) has promised unprecedented productivity, enhanced safety, and operational precision. However, a common frustration persists across the industry: a massive, multi-million-dollar haul truck, executing a carefully planned route, suddenly stops dead in its tracks. There is no physical obstruction or mechanical failure. Instead, the vehicle’s safety protocols have triggered a halt due to a temporary loss or degradation of its Global Navigation Satellite System (GNSS) signal. Consequently, ensuring the absolute reliability of positioning is a priority for maintaining operational efficiency in the complex, dynamic topography of a modern mine.

The Cascading Cost of Signal Degradation

Autonomous fleets depend heavily on a continuous stream of centimeter-level accuracy, typically delivered via Real-Time Kinematic (RTK) corrected GNSS. Yet, the physical environment of a mine, characterized by deep pit sections, steep high walls, and expansive metallic infrastructure, frequently obscures satellite lines of sight. When an RTK fix is temporarily lost in these zones, the positioning accuracy can suddenly downgrade from centimeters to meters. Because meter-level accuracy falls below the strict safety thresholds required for autonomy, the AHS stack defaults to the safest programmed action: stopping the vehicle. While these interruptions may be brief, their system-wide impact is substantial. When a truck is stopped and considered 'lost,' other autonomous vehicles operating in the same work area may also be forced to halt — since each truck must know its position relative to nearby autonomous equipment. Depending on the AHS implementation and zone configuration, a single lost vehicle can cascade into a multi-truck stoppage, causing significant delays and often preventing that shift from hitting its production target. Cumulatively, these micro-stoppages can account for hundreds of hours of unbudgeted downtime annually, severely undercutting the fleet’s Return on Investment and hindering overarching production targets.

Beyond Sole-Reliance on GNSS

Addressing this vulnerability requires moving away from an over-reliance on standalone GNSS and adopting a more robust, system-level architecture. To build true positioning resilience, autonomy teams are increasingly focusing on the integration of high-performance Inertial Navigation Systems (INS) deep within the autonomy stack. A high-performance INS typically combines dual antenna, multi-constellation GNSS receivers with precision MEMS gyroscopes and accelerometers, all governed by tightly-coupled sensor fusion algorithms. During normal operations, these systems seamlessly blend satellite and inertial data to produce accurate positioning. However, the true value of this architecture is realized when external GNSS signals are compromised. So what happens when the truck actually enters a GNSS-degraded zone? The INS core immediately takes over. Utilizing the last known position and a continuous, internal stream of inertial data, the system can “dead reckon” the vehicle’s path. A well-calibrated INS can bridge satellite outages for several minutes with minimal drift — enough to keep the truck moving safely through a blind spot rather than stopping the entire fleet. This capability allows the truck to maintain its speed and trajectory safely through the blind spot. Once clear skies are restored, the system rapidly reacquires the GNSS fix, seamlessly correcting any micro-drift without interrupting the vehicle’s mission.

Interoperability and Integration in the Autonomy Stack

For these positioning solutions to be truly effective, they must be architected for deep interoperability within existing AHS frameworks. This is where open architecture matters most. Navigation hardware needs to be vehicle-agnostic and speak standard protocols — Ethernet, RS422/232, CAN bus — so it can plug into any AHS framework without proprietary lock-in. This ensures that high-integrity position, velocity, acceleration, and orientation data feed directly into the central control and dispatch systems without creating data silos or latency. By streamlining this integration at the system level, engineering complexity is reduced, and deployment timelines are accelerated. Furthermore, feeding continuous, reliable data into the dispatch system enhances operational predictability across the entire circuit. Vehicles remain visible and taskable at all times, fortifying the fleet against common environmental interruptions.

The Future of Mining Autonomy

Ultimately, resolving the GNSS blind spot is not merely about adding new sensors to a truck, but rather about engineering a resilient autonomy architecture. Organizations that prioritize robust sensor fusion, precision INS solutions, and seamless interoperability can reduce unnecessary stoppages, convert idle time back into productive cycles, and ensure their autonomous fleets operate consistently at peak potential.