China is replacing humans with robots amid the threat of a commodity supercycle. What could possibly go wrong
- Новости
- Economy
- China is replacing humans with robots amid the threat of a commodity supercycle. What could possibly go wrong
Brief content
The global metallurgy industry is facing the threat of a new super cycle — a prolonged rise in the cost of resources due to the depletion of rich deposits and rising costs of launching new projects. The problem is particularly acute in the copper market: due to a sharp increase in demand due to the electrification of transport and the construction of data centers, metal shortages could reach 25% of global needs by 2035.
To reduce operating costs, mining giants around the world, from Australia to Canada, are introducing unmanned vehicles. China has become the absolute leader, where, as part of state policy, about two thirds of the world's fleet of autonomous mining dump trucks has been assembled — 3,843 vehicles in July 2026.
Automation perfectly reduces transportation costs in large open-pit mining operations, such as iron ore mining. However, it does not protect against market collapses in overproduction (as has happened with lithium in recent years) and is not able to compensate for impoverished copper ores or stagnating demand for coal.
Replacing the operating costs of staff salaries with capital investments pays off in countries with expensive labor, such as Australia. In China, with low salaries for drivers and high prices for robots, profitability without government subsidies is not obvious. In addition, with a market reversal and falling prices, expensive autonomous systems can become a financial burden, and a massive increase in production can provoke a new collapse in quotations.
The global economy is on the verge of a new commodity supercycle: rich deposits are being depleted, and demand for copper and other resources for energy transition and data centers is breaking records. In response, the mining industry is changing the rules of the game — instead of racing for new deposits, companies are betting on robots. China has become the world leader, having concentrated about two thirds of its fleet of autonomous mining dump trucks with strong government support. But whether drones will save the market from shortages and high prices, or only trap miners into huge expenses when demand drops, Izvestia investigated.
Has the commodity supercycle really started
The commodity supercycle is not just another spike in oil, coal, or metal prices. This is the name given to a long-lasting, sometimes decades-long wave of changes in real commodity prices, which includes both a multi-year upswing and a subsequent recession. Growth usually begins when demand for resources increases faster than the extractive industry is able to increase supply. High prices encourage investments in new fields, but it takes years to launch projects. When supply finally catches up with demand, the market can turn around.
The last major commodity supercycle began in the early 2000s. Its main engine was the industrialization of China, which required huge amounts of iron ore, copper, coal and oil. The scale of the growth was impressive: the average annual copper price increased from about $1,560 per ton in 2002 to $6,960 in 2008. In July 2008, oil reached $133 per barrel on average per month.
Then the global financial crisis brought down the quotes: by December, copper had fallen to about $3,070 per ton, and oil to $41 per barrel. However, the super cycle did not end there. Already in 2011, the average annual copper price reached $8,823 per ton, after which a protracted decline began. By 2016, it had dropped to $4,868, almost 45% below the 2011 peak. In general, the World Bank's industrial commodity indexes lost more than 35% from the beginning of 2011 to the end of 2014.
Over the past 130 years, the world has experienced four commodity supercycles.
Today, the global economy is once again facing the prerequisites of a prolonged rise in the cost of resources. But now the problem is not only growing demand or a drastic supply restriction. The mining economy itself is changing. Rich and easily accessible deposits are gradually being depleted, new projects require more and more capital investments, and their implementation dates stretch over years. Even high prices do not guarantee that manufacturers will be able to quickly increase supplies.
The problem is particularly noticeable in the copper market. Electrification of transport, expansion of power grids, energy development and construction of data centers are increasing the demand for metal. Meanwhile, the copper content in the ore at many deposits is decreasing. To get the same amount of metal, you have to recycle more rock mass, consume more energy and water. The International Energy Agency estimates that by 2035, the potential gap between copper demand and supply from existing and announced projects could amount to about a quarter of global demand.
However, it is still premature to talk about a new global supercycle. Different markets are moving in different directions. The prerequisites for a structural deficit are accumulating in copper, while lithium has already experienced overproduction and a collapse in prices. The long-term prospects of coal are limited by the uncertainty of demand, and iron ore is highly dependent on the state of Chinese construction and metallurgy.
Robots versus expensive raw materials: who is betting on automation and how
Mining companies have several ways to respond to the rising cost of raw materials. You can search for new deposits, develop deeper (and, as a result, more expensive) deposits, invest in waste recycling, or wait for high prices to make previously unprofitable projects attractive. But there is another way to change the mining process itself, reducing dependence on human labor with the help of autonomous technology. This requires additional capital investments (CAPEX) in equipment, management systems, and infrastructure, but it potentially reduces current operating costs (OPEX), primarily for personnel, downtime, and maintenance of transportation operations.
Australia has become a pioneer of this strategy. Back in 2008, Rio Tinto (an Anglo-Australian group that mines iron ore, copper, and bauxite) launched the first autonomous mining dump trucks in the Pilbara iron ore fields. Gradually, other mainland companies joined it, including BHP and Fortescue. Today, unmanned trucks transport millions of tons of ore there, and the equipment is controlled from remote control centers. Automation reduces downtime, optimizes routes, improves safety, and reduces dependence on personnel working in remote areas.
Similar processes are developing in Canada and Chile. Canadian companies are introducing autonomous technology in oil sands fields and in mines of solid minerals. In Chile, one of the earliest examples was Codelco's Gabriela Mistral copper mine, where unmanned dump trucks have been used since 2008. Thus, automation has long ceased to be an experiment: for the largest mining companies, it has already become part of the industrial infrastructure.
The process is developing most rapidly in China today. According to the GlobalData analytical company, in July 2024 there were 562 autonomous dump trucks in the country, and a year later there were 2,090. During the same period, for example, the Australian fleet grew from 927 to 1024 vehicles. In July 2026, GlobalData evaluated the Chinese fleet of autonomous and self-driving dump trucks (not all of them were actually in use at that time) There are already 3,843 units — about two thirds of the world's total in the corresponding category.
One of the most significant projects was launched in May 2025 at the Yimin coal mine in northern China. They have launched a fleet of 100 fully electric unmanned dump trucks using 5G-A connectivity and autonomous control systems. According to the developers of such solutions, automation allows not only to reduce personnel costs, but also to reduce downtime, increase safety and optimize traffic flows. At the same time, the cost of robotic machines is one and a half times higher: according to an industry review published in 2025 in the materials of the Hong Kong Stock Exchange, in 2024 a conventional mining dump truck in China cost on average about $ 210 thousand, and an autonomous one — about $ 350 thousand.
It is important that the effect of the introduction of such developments in China goes beyond saving for individual companies. Automation of mining has become part of the state industrial policy.
Back in 2020, eight Chinese departments adopted a program for the development of intelligent coal mines, which includes not only the introduction of unmanned vehicles, but also the creation of their own production of mining robots, equipment and digital systems. In 2024, the authorities extended this strategy to other types of minerals, setting a goal to form a full-fledged technology, equipment and management system for automated mining by 2030. Thus, Beijing considers robotics not only as a way to reduce the cost of raw materials, but also as a direction for the development of its own industry.
However, in the context of the victory of industrial robots over the commodity supercycle, the scale of implementation does not mean economic superiority. It is important not only the number of unmanned vehicles, but also how much they reduce the cost of a ton of raw materials extracted.
Not all raw materials are the same: Where robotization really reduces production costs
The most obvious example of the effectiveness of the introduction of robotic technologies is iron ore. Its mining is carried out in giant open pits, where huge volumes of rock mass are transported along relatively predictable routes. These are almost ideal conditions for unmanned vehicles. According to Rio Tinto, in 2017, the company's autonomous dump trucks worked an average of 700 hours more than usual, and the unit cost of loading and transporting rock was 15% lower. These figures relate to specific operations of the company and do not imply a similar reduction in the total cost of production.
Here, robotization can really significantly increase the efficiency of mining. But it does not solve the demand problem: if the global metallurgy industry (primarily Chinese) reduces purchases of iron ore, cheaper transportation in the new super cycle will not save producers from falling prices.
The lithium market is also subject to sharp fluctuations in market prices, as producers of this raw material react relatively quickly to demand by launching new projects. There are several mining technologies for this metal: it is obtained from both hard rocks and salt brines. In the first case, autonomous dump trucks and drilling rigs can reduce the cost of mining ore. In the second case, automation of chemical processes, control of the composition of solutions and extraction efficiency are much more important.
However, even technologically advanced production is not protected from market price collapses. For example, after the rapid growth in 2021-2022, the lithium market is facing overproduction. The IEA now expects a more sustained shortage to occur only in the next decade.
In 2021-2022, the rapid growth of electric vehicle production has dramatically increased the demand for lithium for batteries.
The situation is more complicated with copper. The main problem of the industry is the gradual depletion of rich deposits. According to the IEA, the average copper content in ore in the world's active deposits has decreased by about 40% since 1991, and the capital intensity of existing mine expansion projects has increased by 65% since 2020. Against this background, the agency warns that if current and announced mining projects are taken into account, the potential gap between primary copper supply and expected demand may reach 25% by 2035. Robotization of copper mining is able to partially offset the cost increase by increasing productivity, but it does not solve the main problem: reducing the copper content in the ore requires processing more and more volumes of rock, so automation alone is not enough to eliminate the risk of shortage.
In the coal industry, it is not so much about preparing for a shortage of raw materials, as about reducing costs, improving safety and maintaining profitability. This is especially true for deep mines and remote mines, where it is expensive to maintain personnel. Workers must be transported to the fields, provided with housing, food and medical care, as well as compensated with increased salaries for difficult working conditions, health risks and prolonged separation from family. In deep mines, the costs of ventilation, safety systems, and transporting people underground are added to this. Automation allows you to reduce not only the payroll, but also some of the associated costs.
But the prospects for the coal market itself are ambiguous. According to the IEA, global demand for coal in 2025 amounted to about 8.84 billion tons, almost unchanged from the previous year. The short—term dynamics remains unstable: after the growth of global demand for coal to 8.84 billion tons in 2025, the IEA predicts a new record of about 8.94 billion tons in 2026. However, the agency attributes some of the additional demand to temporary energy and weather factors, so this growth does not guarantee long-term market expansion.
At the same time, the energy transition and competition with other sources of electricity create a risk of demand stagnation. It turns out to be a paradox: China is actively robotizing an industry that may face not a shortage, but an excess of production capacity. In the coal industry, automation is becoming more of a survival tool than a way to capitalize on a new supercycle.
Thus, it is reasonable to say that automation can certainly make the extraction of various metals more efficient, but it is not always able to make the raw materials themselves cheaper.
The economics of automation: when do robots make a profit, and when do they become a burden?
Automation of mining looks attractive primarily because it allows you to replace part of the ever-increasing operating costs (OPEX) with capital investments (CAPEX). However, saving money on people does not mean that technology becomes free. Robots need to be serviced, repaired, and software updated, and the initial investment needs to be recouped.
How profitable such a model is depends on the conditions of a particular deposit. Let's consider three scenarios.
The first is that robots pay off. This is most likely in large quarries with high equipment loads, predictable routes, and expensive human labor (as a rule, salaries are higher in countries with higher price levels — the United States, Canada, and Australia). Unmanned dump trucks can work longer, reduce downtime, and use fuel more efficiently. For example, the same Australian Rio Tinto previously reported that its autonomous machines work about 700 hours a year more than usual, and their operating costs are 15% lower. In such conditions, automation can provide a positive economic effect. However, the actual payback period depends on the cost of implementation, equipment loading, maintenance costs and the price of the attracted capital.
The second is that automation works "at zero." Productivity is increasing, staff costs are decreasing, but the savings are being absorbed by equipment depreciation, digital system maintenance costs, and the cost of capital raised. The company produces the same amount or even more raw materials, but its financial result practically does not change. Nevertheless, automation can still be justified: it reduces accidents and allows you to work in dangerous conditions. In this case, the company buys not so much additional profit as the stability of the production process.
Third, robots are becoming a burden. This happens when a company invests in automation at the peak of commodity prices, hoping for a multi-year increase in production, and then the market turns around. Demand is falling, equipment is idle, but loans need to be serviced, and infrastructure needs to be maintained. An additional risk is created by rapid technological obsolescence: an expensive system may require modernization long before the end of its estimated service life. As a result, an automated mine turns out to be technologically modern, but financially less stable than an enterprise with lower fixed costs.
The Chinese model is particularly interesting in this regard, where automation is being implemented on the largest scale. According to the National Energy Administration of the People's Republic of China, by May 2024, investments in the intellectualization of coal mines exceeded 112 billion yuan (about $15.5 billion at the average exchange rate for 2024). The authorities claimed that productivity had increased by more than 20% at the upgraded sites, and the number of shift workers had decreased. At the same time, the government actively encourages robotics: it provides tax incentives, subsidies and preferences for expanding production. However, a study published in 2025 in the journal Mining, Metallurgy & Exploration points to an important feature: the salaries of dump truck drivers in China are several times lower than in Australia, so saving on personnel alone does not justify expensive automation. It turns out that Chinese robotics is already showing productivity growth, but it is still unclear how quickly these investments pay off without government support, according to open data.
Finally, there is a risk that goes beyond a single deposit or even a single country. If automation allows multiple producers around the world to simultaneously reduce production costs and increase production, the additional supply may simply lead to a drop in global prices. Then the companies will receive a technological gain, but will lose part of the expected profit due to changes in market conditions.
Will robotics help to defeat the commodity supercycle? Historically, such cycles have been completed not by a single technology, but by a combination of several processes.: High prices stimulated investments, new capacities gradually entered the market, consumers began to save raw materials or look for substitutes. Therefore, the main result of the technological race may not be the disappearance of commodity cycles, but a change in their nature: some resources will become more accessible, others will remain scarce, and companies that have managed to recoup investments before the next market reversal will benefit.
Переведено сервисом «Яндекс Переводчик»