Proximity detection technologies compared: radar, UWB, RFID, LiDAR

Why the technology choice matters

Collision prevention is only as good as its ability to reliably sense a hazard in your environment — and mining environments are hostile to sensing. A technology that works flawlessly on a clean surface haul road can misbehave in a dusty, curving underground heading full of reflective steel. Picking the wrong sensing approach is how projects end up with false alarms operators learn to ignore, or missed detections that defeat the whole point. So before comparing vendors, it's worth understanding what each underlying technology can and can't do.

The main technologies

• RFID / RF tags. Each person and vehicle carries a tag; readers detect tagged entities reliably and identify who/what they are. Strength: robust identification, works in dust and dark. Limit: it only sees tagged things — an untagged person or obstacle is invisible — and basic versions give zone presence rather than precise distance.
• Ultra-wideband (UWB). A tag-based RF technology that adds precise, real-time distance measurement. Strength: accurate ranging, performs well underground, good for tight vehicle-pedestrian distances. Limit: still tag-dependent (untagged hazards unseen), needs reader infrastructure.
• Radar. Detects objects whether tagged or not, measuring range and closing speed. Strength: works in dust, smoke and darkness, sees untagged hazards. Limit: coarser at telling a person from a rock or a wall, and reflective surfaces and confined geometry can create clutter.
• LiDAR. Builds a detailed 3D picture of surroundings. Strength: high spatial detail and obstacle mapping. Limit: degraded by heavy dust, water and poor visibility — a real constraint in many workings — and higher cost.
• Camera / AI vision. Classifies what it sees (person vs vehicle vs structure). Strength: rich classification, good on surface in good light. Limit: dust, darkness and glare hit it hard underground; needs good lighting or augmentation.

Why fusion usually wins

Because each technology covers another's blind spot, robust systems combine them. A common pattern: RF/UWB tags for reliable identification and precise distance between known people and machines, plus radar (and, on surface, camera/LiDAR) to catch untagged hazards and obstacles. The fusion gives you both "I know exactly where my tagged people are" and "there's something in the path I didn't expect" — which neither alone provides. The right blend depends entirely on whether you're surface or underground, your dust and lighting, your geometry, and your fleet.

How to choose

1. Start from your environment. Underground (dust, curves, metal, dark) and open-pit (range, weather, speed) demand different blends. Match the technology to the conditions, not the brochure.
2. Demand proof from comparable sites. Ask for detection-performance evidence from workings like yours — not a clean-demo render.
3. Insist on intervention. The sensing must feed an EMESRT Level 9 system that warns and can slow/brake the machine. Detection alone doesn't comply.
4. Check false-alarm behaviour. A system that cries wolf gets switched off in operators' heads. Ask how it suppresses clutter and nuisance alarms in your environment.
5. Plan integration and logging. It must integrate with the vehicle's controls and log every event for compliance — and ideally share a platform with person location and hygiene monitoring.

At addanode we approach collision prevention by fitting the sensing to your environment and the EMESRT Level 9 requirement — not by selling one favourite technology — and integrating it on the in-house addaNet platform with event logging for compliance. Because we build the hardware and software in-house and support it locally, the system is tuned for real South African conditions, surface and underground.

Regulatory references are for orientation; confirm current requirements against the latest Mine Health and Safety Regulations and your appointed advisers.

Frequently asked questions