Cat MineStar Command for hauling (Cat 793)

A system, not a truck

The naming matters, because it changes how you read everything else. Caterpillar sells the system, not an autonomous truck model — there is no “Cat 793AT” SKU. Cat MineStar Command for hauling is hardware, software and services, sold and installed through the Cat dealer network. It ships factory-ready on new trucks, retrofits onto existing fleets, and will even go onto a competitor’s trucks: Vale’s December 2025 order fits Cat kits to Komatsu machines.

The platform underneath it is the Cat 793, and that is on the record from Caterpillar’s own Group President for Resource Industries, Denise Johnson: “The Cat 793 haul truck remains Caterpillar’s most popular autonomous truck model.” The truck carries 240 tonnes, weighs 404,000 kg gross, runs a C175-16 making 1,976 kW (2,650 hp), and tops out at 60 km/h loaded.

Perception is a 64-laser lidar (905 nm, Class 1 eye-safe), a 76.5 GHz electronically scanning radar, GNSS accurate to under a metre with base-station corrections, four cameras, and wireless A-Stop transmitters on 919 MHz with 180 m+ range.

One detail worth filing away: Cat’s autonomous trucks keep the cab. Caterpillar’s own language is that the system “allows trucks to operate in either autonomous or manual mode, enabling the same performance and functionality of the standard truck model.” Komatsu went the other way, launching a genuinely cab-less autonomous haulage vehicle at MINExpo back in 2016. One vendor left the seat in and one took it out. That difference will probably read louder in ten years than it does now.

82%: the hardest displacement number in this catalog

Most robot entries can only describe job displacement qualitatively. This one doesn’t have to.

The peer-reviewed figure is that autonomous haulage takes required personnel from 4.5 per truck to under 0.8. For a 36-truck fleet, that is roughly 162 people down to roughly 29.

The coefficient matters because of who it lands on. Haul truck drivers in Australia earn AUD 150,000-180,000 — about USD 100,000-120,000. That is one of the very few jobs on earth paying six figures in US dollars without requiring a degree. Which points at something people routinely get backwards about automation: the machines did not start with the cheapest work. They started with the most expensive. The same paper offers the control case — Chinese haul truck drivers earn CNY 120,000-180,000 (about USD 16,500-25,000), and at those wages the autonomy business case simply doesn’t close.

A statistical trap worth flagging for US readers: the Bureau of Labor Statistics has no distinct occupation code for mining haul truck operator. The nearest proxy is SOC 47-5022, Excavating and Loading Machine and Dragline Operators, Surface Mining — 32,630 employed at a $55,500 annual mean as of May 2023. That badly understates what ultra-class haul truck operators actually earn, so treat it as a floor, not a figure.

Net job loss at site level: both sides on the record

This is the section that most needs restraint, because both sides have real material.

For displacement. The cleanest published number anywhere comes from Suncor COO Mark Little, to CBC in January 2018: “We have about 500 roles that will get eliminated through this and we’re going to add about 100. So the net change is about 400 positions.” It must be noted those were Komatsu trucks, not Caterpillar. The second data set is BHP Mitsubishi Alliance’s deployment at Daunia and Goonyella Riverside in Queensland — 18 permanent operator roles and 45 labour-hire operator roles affected, against 34 new roles created and advertised externally. BMA’s president said “there will be no job losses as a result of this decision.” A labour-hire worker, Peter Guinea, described it differently: “We were told they’d crunched the numbers and a certain amount of jobs would be lost per shift.” The loss landed almost entirely on the labour-hire tier.

Against displacement, and this is not just PR. Fortescue reported no redundancies across 100 truck conversions, with transfers and upskilling. Newmont Boddington explicitly committed to “reskilling and redeployment of haul truck drivers to other roles supporting the AHS,” and the site — around 2,000 direct employees — stood up a control room managing 200+ connected assets under a new job title, AHS control room superintendent. The heaviest counter-evidence is regional rather than site-level: Pilbara mining-focused FIFO workers numbered 13,384 in 2023, a 400% rise from 2011, over exactly the period Rio Tinto’s Pilbara haul fleet passed 90% autonomous. Haulage jobs automated and regional employment went up anyway.

Our read: the per-truck labour coefficient collapse is a hard fact. Site-level net job loss is not. And no independent body has audited headcount before and after at any Caterpillar site. That is as far as the evidence goes.

One safety record that belongs in the entry

Caterpillar’s safety claim is consistently narrow: no lost-time injuries “associated with automated truck operation.” Two 2025 Significant Incident Reports from Western Australia’s mine safety regulator are worth reading alongside it.

Report 226 (March 11, 2025): night shift, open pit. An autonomous haul truck struck a manned water cart. The water cart driver received minor injuries and the autonomous truck was significantly damaged. Travel paths intersected without ground delineation; a turnaround loop had been activated in the control system with no corresponding physical markings or communication; the truck, at about 40 km/h, could not avoid. Report 286 (March 7, 2025): a comms trailer failure cut network coverage and trucks stopped per protocol — but when light vehicles were deactivated the safety bubble collapsed, and the trucks reverted to autonomous mode and began moving while two operators were attempting to board them. No injuries.

Neither report names the autonomy vendor, so neither can be attributed to Caterpillar. What they do establish is that “zero injuries” does not hold as an industry-level statement.

Why we care for LostJobs

Komatsu’s FrontRunner is already in this catalog. Adding Caterpillar isn’t padding — it’s because telling mine automation as a one-company story gives the wrong impression. This is a duopoly, both halves are expanding, and they take contracts off each other. Komatsu won the 41-truck fleet at BHP’s South Flank, BHP’s flagship new Pilbara mine, despite BHP having done its earlier Pilbara autonomy with Caterpillar. Caterpillar won Vale in late 2025 and is retrofitting its kits onto Vale’s Komatsu trucks. For anyone who drives a haul truck for a living, “which brand does our mine run” has never been a question that buys safety.

The direction actually worth watching is downmarket. Command for hauling started on ultra-class metal and oil sands trucks. It is now on Cat 777s at the Luck Stone aggregates plant in Virginia and the Carmeuse limestone quarry in Michigan. Quarry operators are not in the Pilbara, do not earn AUD 150,000, and are usually not unionised. An automation system whose economics close at a $120,000 job will keep moving down the wage ladder as its cost falls — and each rung down has less bargaining power than the one above.

If you operate equipment at a mine, here is a test you can use directly. Don’t listen for whether the company says it will keep jobs. Listen for whether it says permanent employees or everyone working on site. The Queensland numbers already answer it: 18 permanent and 45 labour-hire roles affected, 34 new roles created and advertised externally. The permanent staff were placed, the labour-hire workers absorbed the difference, and the new roles were not reserved for the people displaced. That pattern is very likely to repeat at the next mine.