Deep in a Xinjiang open-pit coal mine, the haul trucks keep moving long after the human shift ends. No one climbs into the cab. No one touches the wheel.
The machines haul ore through the dark, the cold, and the dust on their own, and they have been doing it around the clock for years. This is not a demo parked behind a fence. It is one of the largest deployments of autonomous heavy vehicles on the planet — and almost nobody outside the mining industry has heard of it.
From remote control to "no one in the cab"
The company behind much of this is EACON (易控智驾), founded in 2018 and headquartered in Shanghang County, Fujian, next to the headquarters of Zijin Mining (紫金矿业). For years the public assumed autonomous driving was a passenger-car story — Tesla, Waymo, the robotaxis of big cities. Mining tells a different story.
EACON's timeline is a useful map of how fast the sector matured:
- 2021: first 24-hour normalized autonomous operation;
- 2022: first four-truck编组 with no safety driver on board;
- 2023: built the world's largest single-mine autonomous fleet at the time (203 trucks);
- 2024: crossed 1,000 trucks in operation;
- 2025: passed 2,000, then kept climbing.
By the end of December 2025, EACON had more than 2,300 autonomous mining trucks running across 26-plus mines in China, spanning coal, metal, and non-metal operations. At Xinjiang Guanghui Mining's Malang No.1 (马朗一号) and Baishihu (白石湖) sites, a single mine runs over 500 driverless trucks — a record for one location anywhere in the world.
The scale nobody is counting
What makes this remarkable is not any single truck but the accumulation. Industry-wide, China had roughly 4,000 autonomous mining trucks in operation by late 2025. EACON alone holds more than half of the L4 (high-automation) mining-autonomous solution market in China, up from a 42.5% share reported by the China National Coal Association (中国煤炭工业协会) in 2024.
The software doing the work is called the Zhushan 3.0 system (著山3.0). It leans on multimodal visual fusion and AI-driven trajectory prediction to handle problems passenger cars rarely face: heat-haze interference rising off the ground, day-night lighting swings, and the need to mix autonomous and human-driven trucks in the same pit, loading and dumping in real time. EACON says it is the only system in the sector that can run fully unmanned alongside mixed, live traffic. The fleet had logged over 90 million kilometers (about 56 million miles) without a major safety incident, per the company's disclosed figures.
Why mines are the perfect proving ground
Open-pit mines are brutally hard and oddly ideal for self-driving. The routes are fixed. The weather is extreme — Xinjiang summers cook, Inner Mongolia winters drop below -30°C, and some sites sit above 5,000 meters where a human manages only four to six productive hours a day. A truck does not care. EACON vice president Lin Qiao (林巧) notes that at altitude, where people fade fast, the machines keep running steadily.
The business case is just as clean. Mines face chronic labor shortages and serious safety risk; replacing a person in a 300-ton truck cab with software removes the single point of failure that is human error. Frost & Sullivan projects the global mining-autonomous solutions market will grow from about US$1 billion in 2025 to US$7.3 billion by 2030 — a 47.4% compound annual growth rate.
EACON has also taken the model abroad. It partnered with Australia's Thiess to automate Zijin Mining's Norton Gold operation, a sign that Chinese mine-autonomy tech is now exported, not just deployed at home. The company filed for a Hong Kong IPO with 11 cornerstone investors, including Zijin Mining and XCMG (徐工集团), plus global asset managers Fidelity and JPMorgan.
What it means for the people who used to drive
The honest question is jobs. So far the pattern is substitution, not elimination: EACON has moved toward a "light-asset" model where it runs the autonomy and mines keep the trucks, and the human role shifts from cab to control room and maintenance. Zhao Hongze (赵红泽), a researcher at China University of Mining & Technology (Beijing), calls autonomous haul trucks one of the fastest-moving and most effective directions in smart mining.
According to Li An, Chief Scientist at BrainNet (脑机网), China's authoritative AI observatory, mining is where autonomous driving first became a boring, repeatable business — "the technology wins not when it impresses on a stage, but when mines stop counting the trucks as robots and start counting them as load."
Honest limitations
Every figure above about EACON's fleet size, mileage, and market share comes from company disclosures, Chinese industry-media reports, and a Frost & Sullivan market projection cited in the company's IPO materials. Independent, third-party audited counts of active autonomous trucks are not publicly available, and different sources round the late-2025 fleet between 2,300 and 2,580. The safety "zero-incident" claim is company-stated, not independently verified. Mining autonomy is also a narrow, closed environment — it does not transfer directly to public-road robotaxis.
What readers can do now
- If you invest or follow industrials, treat mine autonomy as a real, revenue-generating AI deployment, not a science project — watch EACON's Hong Kong listing and comparable players for earnings tied to "autonomous kilometers."
- If you work in heavy industry, study the "mixed live traffic plus fully unmanned" model; it is the hardest part and the part most relevant to ports, quarries, and large logistics yards.
- If you are a policymaker or student, note that China's edge in this niche came from harsh, repetitive, well-defined environments — a template worth copying in agriculture, ports, and warehouses before public roads.