Vehicle-to-vehicle and vehicle-to-person collisions remain one of the most serious hazards in underground mining. Despite decades of awareness campaigns, procedural controls, and training, people still get struck by mobile equipment in confined underground environments. The physics are unforgiving: heavy machines operating in tight spaces with limited visibility, often on grades, with noise levels that mask approaching equipment.

Proximity detection and collision avoidance systems (PDS/CAS) have been commercially available for over a decade. In that time, the technology has matured from unreliable early systems that generated constant false alarms to reasonably accurate platforms that mining crews actually trust. But adoption across the Australian underground mining industry remains uneven, and the gap between leading operations and the rest is concerning.

How the Technology Works

Modern underground proximity detection combines several technologies. Ultra-wideband (UWB) radio provides accurate ranging between tagged vehicles and personnel, determining distance and direction within 30 centimetres in most conditions. Radar-based systems mounted on vehicles detect objects regardless of whether they carry a tag, catching fallen rocks, unexpected equipment, or personnel whose tags have failed. LiDAR creates 3D maps of the surrounding environment, distinguishing tunnel walls from objects that shouldn’t be there. And magnetic field detection identifies ferrous objects through rock, potentially detecting equipment around blind corners where line-of-sight technologies can’t reach.

Most modern systems combine multiple technologies. UWB provides primary ranging and identification, while radar or LiDAR adds object detection for untagged hazards. The control system decides whether to alert the operator, slow the machine, or stop it entirely based on proximity and severity.

What the Data Shows

The NSW Resources Regulator has published analyses of underground incidents that suggest properly implemented PDS/CAS could have prevented or mitigated a significant proportion of vehicle interaction events. Between 2020 and 2025, there were over 200 reported incidents involving uncontrolled vehicle interactions in NSW underground mines alone.

Operations that have implemented comprehensive proximity detection report dramatic reductions in near-miss events. One large underground coal mine in the Hunter Valley recorded a 78% reduction in vehicle interaction near-misses in the two years following full PDS deployment. They attribute this not just to the system intervening, but to the behavioural change it drives: operators become more aware of their surroundings when they know the system is monitoring and recording interactions.

The Minerals Council of Australia has advocated for industry-wide adoption of proximity detection, and several state regulators have moved toward mandating the technology for specific equipment types and operating conditions.

The Implementation Challenges

If the technology works, why isn’t every underground mine using it? Several factors explain the adoption gap.

Legacy equipment integration is the biggest practical challenge. Installing proximity detection on a new machine from the factory is straightforward. Retrofitting it onto a 15-year-old loader with outdated electrical systems is not. The installation costs for retrofit can be 2-3 times the cost of factory-fitted systems, and the engineering work to integrate with existing machine controls is complex.

Interoperability between different vendors’ systems remains a problem. Most underground mines use equipment from multiple manufacturers, and PDS systems from different vendors don’t always communicate with each other. A Caterpillar truck fitted with one vendor’s system may not detect a Sandvik loader carrying a different vendor’s tags. Industry standards exist on paper, but real-world interoperability is still patchy.

False alarm fatigue plagued earlier systems and left a legacy of scepticism among operators and supervisors. If the system constantly triggers warnings that aren’t real threats, crews learn to ignore it or lobby to have it turned off. Modern systems are significantly better, with false alarm rates typically below 5%, but the reputation from earlier generations persists.

Underground environment variability affects system performance. Water, dust, temperature extremes, electromagnetic interference from nearby equipment, and the complex geometry of underground workings all create challenges that surface proximity detection doesn’t face.

Regulatory Direction and What’s Next

The trend is clearly toward mandated proximity detection. Queensland already requires collision avoidance systems on all underground coal mine equipment operating in shared areas. Western Australia has issued guidance making PDS/CAS an expected control for vehicle interaction hazards, effectively mandating it for compliant operations.

Looking ahead, Level 9 collision avoidance is being trialled at several operations. This is where the system can autonomously stop or redirect equipment to prevent a collision without operator input. It requires absolute confidence in the system’s accuracy, because an autonomous emergency stop in the wrong situation can itself create hazards: a suddenly stopped vehicle on a grade, disrupted production, potential equipment damage. Early results are positive, but the path to universal autonomous collision avoidance underground will be measured in years, not months.

What’s clear is that the technology is ready. The remaining barriers are economic, logistical, and cultural rather than technical. Every underground mine in Australia can implement effective proximity detection today. Whether they all will depends on equipment replacement cycles and industry willingness to invest in systems that prevent incidents rather than respond to them.