The short version
Autonomous haulage systems (AHS) replace the driver of a mining haul truck with a combination of high-precision GPS, on-board lidar and radar, and a central traffic-management system that assigns every truck a path, a speed and a task. The truck does not "see and decide" like a self-driving car on a public road — it operates inside a tightly controlled, fully mapped zone where every machine, light vehicle and person is tracked, and the central system owns the traffic model.
Who runs it today
The technology is no longer experimental. Caterpillar’s MineStar Command and Komatsu’s FrontRunner together operate well over a thousand autonomous trucks across iron ore, oil sands and copper operations, with the largest fleets in Western Australia’s Pilbara. Rio Tinto, BHP and Fortescue run substantial autonomous fleets; Rio’s AutoHaul rail system moves the ore to port autonomously as well. The consistent operator-reported results: double-digit productivity gains from consistent cycle times, and the removal of people from the highest-risk interactions on site.
How the system actually works
Each truck runs redundant positioning (GNSS plus inertial), obstacle detection (lidar/radar) and a drive-by-wire control stack. A central server plans every truck’s route segment-by-segment and holds a live model of everything moving on site — autonomous trucks, manned dozers and graders, light vehicles fitted with transponders. If anything enters a truck’s planned corridor unexpectedly, the truck slows or stops; if positioning confidence degrades, it stops. The design philosophy is conservative: the system trades a little speed for the guarantee that it never acts on stale or uncertain data.
What autonomy vendors audit before deployment
This is the part most operators underestimate. Before an AHS deployment, vendors audit the site’s data discipline: survey accuracy and currency (the mine model must match the real pit), haul-road design and condition management (grades, widths, berms, super-elevation kept within specification), traffic management maturity, and network coverage. A site that cannot keep its survey model current or its haul roads within design tolerance is not autonomy-ready — regardless of budget.
The honest economics
AHS suits large, long-life operations with big fleets and repetitive cycles: the infrastructure cost (network, control room, fleet retrofit or replacement) amortises over many trucks and years. For mid-tier mines and quarries, full autonomy is usually not the near-term play — but the prerequisites are, because they pay for themselves with drivers still in the seat: current survey truth, haul-road compliance monitored against design, disciplined traffic management and telemetry-driven maintenance all cut cost and risk today, and leave the site autonomy-ready when the economics arrive.
Where to start
Start with the data layer, not the truck: establish a live, accurate survey model of the pit; monitor haul-road condition against your approved design limits; and get machine telemetry into a system people actually use. Those three moves capture a meaningful share of autonomy’s safety and cost benefit at a fraction of the cost — and they are exactly what the vendors will audit if you ever make the call.