GPS Fleet Tracking for Mining Trucks: Coverage, Fuel, Drivers and Platforms

The problem that starts the search

At month end the fuel bill doesn't match the number of trips. A subcontractor says his truck worked 10 hours, but the weighbridge recorded four loads. Nobody can prove what happened, because the truck was up a mountain 80 km from the office. This guide covers how to choose and install a GPS system that gives real answers in conditions like these.

Who this guide is for

Mining contractors, ore haulers and construction companies running trucks, excavators and service vehicles across the Lao provinces. If the big haul trucks inside your pit already run an OEM fleet management system, this guide covers everything else: the support fleet, subcontractor trucks, fuel bowsers and the long haul from mine to border crossing or processing plant.

What a tracker actually does

A GPS tracker (strictly a GNSS receiver, since most units also use GLONASS or BeiDou) works out position, speed and heading, then sends a small packet over the mobile network to a server. Many trackers in this class, including JimiIoT-style devices, talk to the server using a compact binary protocol of the GT06 family: a login packet carrying the device IMEI, periodic heartbeats, location packets and alarm packets.

You don't need to understand the bytes. You do need to know that your platform must support the exact protocol your device speaks, or you'll be locked into one vendor's server.

Coverage: the real problem in Laos

Along Route 13 and near provincial towns, 4G is usually fine. In valleys, inside the pit and on mountain roads in provinces such as Xaysomboun, Attapeu or the north, long dead zones are common. Two things matter.

Store-and-forward. The tracker must save positions to internal memory when there is no signal and upload them, in order, when it reconnects. Ask how many points the buffer holds, then test it: drive a route with known dead zones and check the track on the map has no straight-line jumps.

Network technology. In 2026 we'd avoid 2G-only units. Choose 4G (LTE Cat-1 or Cat-M1) with 2G fallback where the operator still runs 2G. If your route crosses two operators' coverage, ask about a SIM that can roam between Lao networks, or a dual-SIM unit.

For sites with no mobile coverage at all, satellite messaging terminals exist, but the data cost is a different conversation. Most fleets accept the delay and rely on store-and-forward.

Reporting interval and buffer size

Example: a truck runs 12 hours a day and reports every 30 seconds while moving. That's 120 points an hour, or 1,440 a day. If the longest dead zone on the route is 3 hours, the buffer must hold at least 360 points plus alarm events. Most units hold far more, but some cheap models fall short if a truck parks overnight in a dead zone. Also turn on reporting on heading change, so the track on the map doesn't cut across mountain bends.

Installation on heavy trucks

What fails in the field is rarely the tracker. It's the install.

• Heavy trucks run on 24 V. Check the input range (many units accept roughly 9–36 V or wider) and fuse the supply line.
• Drivers and mechanics use the battery isolator switch. A tracker with a small internal backup battery still reports the power cut and last position.
• Vibration and dust: mount the unit behind the dashboard on a solid bracket, not cable-tied to a wiring loom. Seal the connectors. Mine dust gets in everywhere.
• Antenna placement: the GNSS antenna needs sky view through non-metallic material. A unit buried under a steel dash loses its fix in the pit.
• Rainy season: mount the unit and connectors above the splash line for river crossings and mud.

Fuel: the reason most mines buy tracking

Position alone rarely pays back. Fuel does.

A 40-litre drop at 2 a.m., engine off, inside a zone nobody should be in: that is the report management actually reads. Calibrate every tank (fill in steps, record volume against the reading) or the numbers will be argued over forever.

A caution: some makes don't broadcast fuel level on the CAN bus at all, and some older trucks have no usable CAN data. Check each model in the fleet before deciding you can skip tank sensors.

Worked example: idle time

Say an excavator idles 3 hours a shift and its engine burns 4 litres an hour at idle (check the real figure in your engine manual). That's 12 litres a shift. Over 26 working days it's 312 litres a month per machine. With 20 machines and trucks like that, 6,240 litres a month go into moving no ore at all. Halving idle time is often the fastest visible saving from the whole system. For excavators, report engine hours, not kilometres.

Which sensors for which vehicle

Cameras and fatigue detection

Video telematics (an MDVR or 4G dashcam) can add a driver-facing camera that detects eyes closing, yawning and phone use and alerts in real time. For night shifts on mountain roads this is a genuine safety measure. But video uses many times more data than position: upload only short event clips and keep full footage on the in-vehicle memory card. Tell drivers about the cameras before they go in, and decide who may view the clips.

Driver ID and behaviour

Give each driver an RFID card or iButton key and have the tracker log who is driving. Combine it with harsh braking, overspeed and idling alerts. Idle time is underrated. An excavator left idling three hours a shift burns real fuel.

Geofences that earn their keep

Draw fences around the pit, crusher, weighbridge, fuel station, workshop and the villages along the haul road. Set lower speed limits inside village fences. Alert on entry and exit at night. Ten well-chosen fences beat two hundred that nobody looks at.

Platform: multi-tenant, roles and retention

Mining jobs usually involve the owner, a main contractor and several subcontractors. A multi-tenant platform lets each company see only its own vehicles while the owner sees everything. Check for:

• Role-based access (dispatcher, fuel controller, read-only management)
• Report export to Excel/PDF, plus an API for your ERP or weighbridge system
• Data retention: how long raw points are kept (we'd want at least 6–12 months of raw track for fuel disputes) and whether summaries are kept longer
• Where the server is hosted, and who owns the data if you change provider

Common mistakes

• Buying the devices first, then asking whether the platform supports their protocol.
• Skipping tank calibration, so the first fuel report gets rejected.
• Too many alerts, until supervisors mute all of them.
• Nobody owns SIM top-ups. When credit runs out, trucks "disappear" from the map.
• Not watching for tampering. Alert on main power disconnects and review that report weekly.

Pilot checklist

1. Pick 5 vehicles on the worst-coverage route, including trucks and an excavator.
2. Calibrate every tank and keep the calibration tables.
3. Draw no more than 10 geofences and agree who receives which alert.
4. After 2 weeks, compare trips in the system against weighbridge tickets.
5. After 4 weeks, measure data gaps, fuel accuracy and how often supervisors opened the dashboard.

Our recommendation

Pilot five trucks for one month on your worst route before buying a hundred units. Measure how big the data gaps really are, how accurate the fuel readings are after calibration, and whether supervisors actually open the dashboard. Our GPS fleet tracking page sets out the installation scope. If you also need dust or water-level readings at the mine, see IoT environmental monitoring.

Send us your route map and vehicle count, and we'll mark where we expect coverage gaps and propose a pilot.