Australian autonomous haulage in May 2026 is past the novelty stage and into the operational maturity stage. The big three iron ore operators in the Pilbara are running large autonomous fleets and the question has shifted from “will it work” to “how do we run mixed manual–autonomous operations economically”.

The Komatsu and Caterpillar autonomous truck counts in Pilbara operations are well above where they were two years ago. Most large open pit iron ore operations are running over 60% of haulage fleet hours autonomously. The productivity numbers have been broadly stable through 2024–2026, with most operations reporting 15–25% productivity uplift on autonomous fleets compared with manual on a like-for-like haul.

The interesting work in mid-2026 is on three secondary problems.

First, mixed-fleet operations. Most operations are still running a portion of manual fleet alongside the autonomous fleet, and the integration cost is higher than the slide decks said it would be. Manual trucks moving through autonomous truck operating zones still trigger conservative safety logic that knocks productivity. The operators getting the most out of their fleets are restructuring the haul road layout to keep autonomous and manual fleets on separate haul roads where possible.

Second, dispatch and short-interval control. The dispatch systems running autonomous fleets need a clean stream of operational data — equipment state, queue lengths, dump availability, road conditions — to make good decisions. Operations that have invested in operational technology integration are getting meaningfully more out of their autonomous fleets than operations that are still running on the dispatch system the vendor shipped.

Third, the workforce model. The autonomous fleet has not eliminated jobs. It has restructured them. The Pilbara operators now have larger remote operations centres in Perth running multiple sites and smaller on-site operational teams focused on equipment availability, maintenance, and exception handling. That has changed the skill mix on site significantly.

Outside the Pilbara the picture is more varied. Coal operations in the Bowen Basin have done less autonomous haulage. The economics of coal at current prices have not justified the capital. Gold operations are mostly still manual. The mid-tier metals operations are watching the Pilbara closely but not committing.

The argument that is still being run in 2026 is about whether autonomous loaders and autonomous drilling can match the operational maturity that autonomous haulage has reached. Autonomous drilling has progressed faster than expected and is now standard at many open pit operations. Autonomous loaders are a harder problem — the load cycle interacts with rock conditions in ways that are harder to engineer around than the haul cycle.

For operations planning the next investment cycle, the read in mid-2026 is that the autonomous haulage business case is now well-established for large open pit operations with long mine lives and stable haul profiles. The case is much weaker for smaller operations, shorter mine lives, and operations where the haul profile changes frequently. The vendors are continuing to push the technology down-market but the economics still struggle below a certain operational scale.

The next 12 months will probably be defined by the operational technology integration question rather than the truck count. The operations that get the dispatch and OT stack right will keep extending the productivity gap on the operations that are running good trucks on a poor data backbone.