Most mines add automation gradually, layering it onto infrastructure and workflows that were never designed for it. Gudai-Darri, which Rio Tinto brought into production in the Pilbara in 2022, was built the other way around. The company designed it from first principles as what it called its most technologically advanced mine, and the project marked the first time Caterpillar, rather than long standing partner Komatsu, became the primary autonomous haulage supplier at a major Rio Tinto Pilbara site.

The scale of the undertaking is easy to understate from a distance. Since ground was broken in April 2019, more than 14 million work hours went into the mine before it produced its first tonne of ore, work that moved over 20 million cubic metres of earth, placed 35,000 cubic metres of concrete, and installed 10,000 tonnes of steel. Construction supported more than 3,000 jobs at its peak, and the finished operation now runs with around 600 permanent roles. Gudai-Darri was designed with an expected life beyond 40 years and an original annual capacity of 43 million tonnes, built specifically to underpin Rio Tinto's flagship Pilbara Blend iron ore product, and it became the company's seventeenth mine in the region and its first new Pilbara greenfield development in more than a decade.

Rio Tinto's chief technical officer at the time, Mark Davies, framed the mine's completion as a statement about what the company had learned across two decades of automation elsewhere in the Pilbara rather than a one off achievement. "In building this new hub we have brought together the best of our innovations, including autonomous trucks, trains and drills, as well as the world's first autonomous water trucks, to make Gudai-Darri our most technologically advanced iron ore mine." That list matters because none of it existed as a single package before Gudai-Darri. Each piece had been proven separately, in some cases over more than a decade, at other Rio Tinto sites first.

Choosing Caterpillar over a decade long partner

The decision to bring Caterpillar in came with a substantial existing track record behind it, and it was not an obvious choice given how closely Rio Tinto had worked with Komatsu since its original Pilbara automation push began in 2008. By the time Gudai-Darri was being planned, Caterpillar's Command for hauling system had already moved more than a billion tonnes safely across the Pilbara with roughly 200 autonomous trucks in commercial service, and Caterpillar reports that mines running the system have recorded productivity increases of more than 30 percent alongside reductions in the frequency of safety incidents of as much as 50 percent.

Gudai-Darri opened with 23 autonomous Cat 793F haul trucks and three autonomous MD6310 drills. The trucks use onboard sensors for real time ore tracking, giving operators live visibility of dig face progression as material is extracted, while data informed modelling from the drills builds a more accurate picture of ground conditions than periodic manual surveys could provide. The mine also became the site for the world's first fully autonomous water trucks, developed jointly with Caterpillar, which use an onboard system to detect dry and dusty haul road conditions and automatically trigger water application, cutting both dust and water consumption through digital tracking rather than a driver's judgment call. A 34 megawatt solar farm, built from around 83,000 photovoltaic panels, supplies roughly a third of the mine's average electricity demand, and a fully automated ore sampling laboratory transfers production samples from a conveyor to an automated production cell by robot, giving the operation real time visibility of stockpiled ore grade rather than waiting on periodic lab results.

When the Caterpillar partnership at Gudai-Darri was first announced, Chris Salisbury, then Rio Tinto's Iron Ore chief executive, described it as a natural extension of the company's existing automation strategy rather than a new direction. "Technology is rapidly changing our mining operations as we harness innovation to make our operations safer, smarter and more productive," he said. "We're pleased to be partnering with Caterpillar and WesTrac to help make Gudai-Darri the most technology enabled and innovative mine in our Pilbara iron ore network."

A mine designed to grow without a second construction project

What makes Gudai-Darri worth revisiting in 2026 rather than treating as a finished 2022 story is what has happened to it since. Rather than needing a second capital intensive construction phase to expand, Rio Tinto found it could grow the mine's output using the flexibility already built into its design. Matthew Holcz, Rio Tinto's managing director of Pilbara mines, explained the logic behind the approach the company settled on: "What we have learnt during the rapid ramp up of Gudai-Darri has given us the confidence to find better ways to increase capacity at our newest and most technologically advanced mine. Rather than taking a capital intensive approach to replicate existing infrastructure, we have now identified a low capital pathway to creep capacity to 50 million tonnes a year."

The numbers behind that plan are striking next to what a comparable greenfield expansion would typically cost. Rio Tinto put the capital intensity of the capacity increase at around 10 dollars per tonne, a fraction of what building new processing infrastructure from scratch would require. The autonomous haul fleet itself grew alongside that plan, from the original 23 trucks to 26 Cat 793F units, with the drill fleet expanding to three MD6310s and two additional MD6250s. None of that growth required a second version of the 2019 to 2022 construction effort. It required incremental additions to a system that was already built to absorb them, which is precisely the advantage of designing for automation from day one rather than retrofitting it onto fixed infrastructure years later.

The wider Pilbara picture in 2026

Gudai-Darri does not operate in isolation, and Rio Tinto's broader Pilbara results through 2026 give useful context for how much weight the mine now carries. In its first quarter 2026 production update, the company reported its second highest Pilbara iron ore output since 2018, up 13 percent year on year, even after tropical cyclones Mitchell and Narelle disrupted shipping and cost the company an estimated 8 million tonnes of lost shipments, roughly half of which it expected to recover. Rio Tinto chief executive Simon Trott, who has since moved from leading the Iron Ore division into the group's top role, tied the quarter's performance to a broader operational push the company calls its stronger, sharper, simpler way of working, under which the first 650 million dollars of annualised productivity benefits had, by that point, been fully implemented across the business.

Autonomy's footprint across the wider Pilbara has also kept expanding well beyond Gudai-Darri specifically. By mid-2026, Rio Tinto was operating 40 autonomous drills across seven mine sites in the region, all coordinated remotely from the same Perth Operations Centre, more than 1,500 kilometres from the mines themselves, that has anchored the company's automation strategy since the original Mine of the Future program began in 2008. The company has also moved into the next generation of haulage technology entirely. In June 2026, Rio Tinto joined BHP and Caterpillar in launching a trial of two Cat 793 XE Early Learner battery electric haul trucks at a demonstration site in the Pilbara, an industry first collaboration between two rival mining companies sharing a single trial. The trucks underwent safety validation and controlled testing at Caterpillar's Tucson Proving Ground in the United States before arriving in Western Australia, and after three months of on site testing the trial was reported to be showing meaningful progress on one of the more difficult open problems in large scale mining: cutting greenhouse gas emissions from haulage without sacrificing the productivity autonomy has already delivered.

What Gudai-Darri actually demonstrates

The practical lesson for anyone planning a new mine, rather than retrofitting an old one, is direct and now backed by four years of operating evidence rather than a single opening day announcement. Automation is measurably easier to deploy, and cheaper to expand later, when it is designed into a mine's infrastructure, communications, and workflows from the very first construction drawings rather than layered onto an operation built for human drivers. Gudai-Darri did not just open as an automated mine in 2022, it grew as one in 2026, adding a fifth of its original capacity for roughly 10 dollars a tonne specifically because the underlying systems, the fleet management software, the autonomous drill fleet, the remote operations link back to Perth, were built with headroom rather than built to a fixed specification.

That is a genuinely useful reference point for any greenfield project currently being planned across Africa's copper, iron ore, or gold belts, where capital is typically far more constrained than it is for a major like Rio Tinto. A mine that can grow its output through incremental additions to an existing automated system, rather than through a second multi year construction program, changes the entire capital planning conversation for a financier or a mine manager deciding how much automation infrastructure to build in from the start. The answer Gudai-Darri suggests is that the infrastructure itself, once automated, tends to pay that upfront cost back more than once.