Every year for the past decade, someone has declared that autonomous haul trucks are about to transform mining. Every year, the transformation is slower than predicted. But it is happening — just not at the pace or in the way that most people expected.

Here’s an honest look at where autonomous haulage stands in 2026, who’s doing it well, and what’s still holding it back.

The Current Fleet

Globally, there are approximately 1,200-1,500 autonomous haul trucks operating in production mining environments as of early 2026. That sounds like a lot until you consider there are roughly 50,000 haul trucks in service worldwide. We’re talking about 2-3% of the global fleet. Hardly a revolution.

The vast majority of those autonomous trucks are concentrated in a handful of large operations in Australia, Canada, and Chile. Australia leads the count, with BHP and Rio Tinto running the largest autonomous fleets in the Pilbara.

Rio Tinto’s Gudai-Darri mine in Western Australia is probably the most advanced example. It was designed from the ground up as an autonomous operation, which makes a huge difference. The pit layout, road design, loading infrastructure, and communications network were all built to support driverless trucks. The result is an operation that’s genuinely efficient — not a retrofit of autonomous trucks into a mine designed for human drivers.

What’s Working

Productivity. Autonomous trucks run longer hours than human-operated trucks. They don’t need shift changes, meal breaks, or bathroom stops. They don’t slow down at the end of a 12-hour shift due to fatigue. The productivity gain is real and measurable — typically 15-25% more material moved per truck per day compared to a manned equivalent.

Consistency. Autonomous trucks follow optimal driving patterns every cycle. They don’t take corners too fast, they don’t ride brakes on downhill runs, they don’t shortcut road maintenance zones. This consistency reduces tyre wear (a significant cost — a single haul truck tyre can cost $50,000-$80,000), extends component life, and reduces road maintenance requirements.

Safety. No human in the cab means no human exposed to the risks of operating a 300-tonne vehicle in a pit environment. Fatigue-related incidents — which have historically been a major cause of haul truck accidents — are eliminated entirely. The safety record of autonomous operations has been excellent, with no fatalities reported in autonomous haul truck operations to date.

What’s Not Working

Mixed fleets. Most mines that have adopted autonomous haulage run a mix of autonomous and manned vehicles. This creates interaction challenges. The autonomous trucks follow rules perfectly; human drivers don’t. Managing the interface between autonomous and manned traffic is complex, requires dedicated traffic management systems, and introduces risks that a fully autonomous operation avoids.

Retrofits. Converting existing manned trucks to autonomous operation is possible — Caterpillar, Komatsu, and Hitachi all offer retrofit kits — but the results are mixed. Retrofitted trucks generally perform below purpose-built autonomous trucks because the vehicle wasn’t designed for the additional sensors, computing hardware, and redundant systems that autonomous operation requires.

Wet weather. Autonomous systems rely on GPS, radar, and lidar for navigation and obstacle detection. Heavy rain, fog, and dust storms degrade sensor performance. In the Pilbara, cyclone season regularly forces autonomous fleets to stand down, sometimes for days. Human drivers can often continue operating in conditions that force autonomous systems to stop.

Loading. Most autonomous operations still use human-operated excavators to load the autonomous trucks. The loading interface is a known inefficiency — autonomous trucks queue at the loader, a human operator fills them, and the truck departs. Autonomous loading is being developed, but it’s several years behind autonomous haulage in maturity.

The Cost Question

Autonomous haul trucks cost 10-15% more upfront than equivalent manned trucks when purpose-built, and the conversion cost for retrofits is $1-3 million per truck depending on the platform. On top of that, you need significant infrastructure investment: high-precision GPS base stations, pit-wide communication networks, central control rooms, and dedicated support staff.

For a fleet of 40 trucks, the total investment (trucks, infrastructure, software, integration, training) can exceed $200 million. The payback comes from labour savings (fewer drivers), productivity gains, and reduced maintenance costs, but it typically takes 4-7 years to see a positive return.

The economic case is strongest for large, long-life operations in remote locations where labour is expensive and difficult to recruit. It’s weakest for smaller operations, shorter mine lives, and locations where labour costs are lower.

Who’s Getting It Right

The operations succeeding with autonomous haulage share common traits: they committed fully rather than running small pilot programs, they designed (or redesigned) their operations to suit autonomous trucks rather than forcing the technology to fit existing layouts, and they invested heavily in change management.

That change management piece gets overlooked. Moving to autonomous haulage means retraining maintenance crews, restructuring mine planning processes, and building new operational roles (fleet controllers, systems engineers). One company doing this well is focused on helping operations manage the technical and organisational complexity of these transitions.

What’s Next

The next step is autonomous loading, which would close the loop and create fully autonomous load-haul-dump cycles. Caterpillar has demonstrated autonomous loading in controlled environments, and several operations are running pilot programs. But the variability of digging conditions — different rock types, fragmentation patterns, truck positioning — makes autonomous loading harder than autonomous haulage.

Electric autonomous trucks are also on the horizon. Battery-electric haul trucks eliminate diesel costs and emissions while pairing naturally with autonomous operation (electric drivetrains are simpler to control autonomously than diesel-mechanical systems).

The Realistic Take

Autonomous haulage works. The mines running it are getting real value. But the industry-wide transformation that was predicted a decade ago hasn’t materialised because the barriers are economic and organisational, not technical.

For the handful of mega-operations with the capital, scale, and mine life to justify the investment, autonomous haulage is now standard practice. For everyone else, it remains aspirational. That gap will close as costs come down and the technology matures, but don’t expect 50% of the global fleet to be autonomous any time soon. We’re still in the early chapters of this transition.