I’ve spent the last few months visiting mine sites across Western Australia and Queensland, talking to operations managers, maintenance supervisors, and the people who actually work alongside autonomous trucks every day. The picture they paint is quite different from the one you’ll get at a mining conference.

Here’s my honest assessment of where autonomous haulage stands in early 2026.

The Scoreboard: Who’s Running What

The Pilbara remains the global centre of autonomous haulage. The big three are all expanding:

Rio Tinto operates the largest autonomous fleet in the world. Their Komatsu FrontRunner truck fleet now covers more than a dozen mine sites. Well past the pilot phase—this is operational at scale.

BHP has expanded autonomous haulage at South Flank and is integrating it into Jimblebar. Their Cat 794AC platform rollout has been slower and more methodical than Rio’s.

Fortescue continues autonomous operations at Christmas Creek and Cloudbreak, investing heavily in fleet electrification alongside autonomy. Tackling both simultaneously is ambitious, maybe overly so.

Outside the Pilbara, things get thin. Roy Hill runs autonomous trucks at smaller scale. A handful of gold and copper operations are trialling systems, but most haven’t moved past proof-of-concept.

What’s Genuinely Working

Credit where it’s due. In the right conditions, autonomous haulage delivers:

Safety. Zero fatalities from autonomous truck operations across the Pilbara since inception. Given that haul truck incidents are historically a leading cause of mining fatalities globally, this matters enormously.

Consistency. Autonomous trucks don’t get tired at 4am. They maintain consistent speeds, follow optimal paths, and avoid attention-lapse errors. The productivity uplift—typically 15-20% over manned operations—is significant.

Utilisation. No shift changes, no crib breaks, no fatigue management pullouts. An autonomous truck runs roughly 22-23 hours per day versus 18-19 productive hours for a manned truck.

These gains are real. But they come with context that often gets glossed over.

Where the Press Releases Don’t Match Reality

Multi-commodity operations are struggling. Autonomous haulage works best in large iron ore pits with simple geometries. Complex ore bodies, multiple pit levels, or variable bench heights—common in gold, copper, and nickel operations—cause the fleet management system (FMS) real problems.

Interoperability remains a headache. Most sites run mixed fleets. A Komatsu autonomous truck needs to interact with Cat excavators, Hitachi loaders, Sandvik drills, and human-driven light vehicles. Each OEM’s proprietary platform doesn’t play nicely with competitors. One operations manager described this as “the single biggest barrier to scaling beyond iron ore.”

Maintenance isn’t cheaper. The autonomy stack—LiDAR arrays, radar units, GPS receivers, edge computing hardware—adds a new maintenance burden. Sensor arrays get sandblasted in the Pilbara. Computing units overheat. Calibration requirements are constant.

According to CSIRO research, autonomy component maintenance typically adds 8-12% on top of the base truck maintenance budget. That partially offsets productivity gains, especially early on.

The Workforce Reality

Autonomous haulage hasn’t eliminated haul truck operator jobs overnight. It’s changed them. Former operators have moved into remote monitoring, field support, and maintenance. The skills required are different—more screen-based, more technical, more diagnostic.

The transition has been smoother at sites that invested in genuine retraining. It’s been rocky where retraining was a tick-the-box exercise. Some experienced operators have left the industry entirely, and that institutional knowledge of local pit conditions isn’t easily replaced.

My Honest Assessment

Autonomous haulage in the Pilbara iron ore sector is mature and proven. It works, it saves money over time, and it’s dramatically improved safety.

Beyond the Pilbara? Still early innings. The technology needs to handle complexity better, OEMs need to get serious about interoperability, and the industry needs honesty about costs and workforce transition.

If you’re a mid-tier miner considering autonomous haulage, don’t believe the timeline your OEM salesperson gives you—double it. Budget for sensor maintenance, communication infrastructure, and genuine retraining. Talk to operators who’ve been running autonomous fleets for more than two years, not just the ones that switched on last quarter.

The technology is coming for the whole industry eventually. But “eventually” and “next year” are very different things.