The business case is more complicated than removing drivers

Mining 4.0 is now a practical operating agenda rather than a technology slogan. Remote operations centres, advanced analytics, autonomous equipment, artificial intelligence and connected systems are being integrated into mines across the world. Autonomous haulage sits at the centre of that transition because trucking is one of the largest, most repetitive and most measurable activities in a surface mine.

The appeal is clear. A truck that can operate through shift changes, execute repeatable driving behaviour and remain inside a tightly controlled dispatch system can improve utilisation and reduce variability. But a mine does not receive those gains automatically. Autonomous fleets introduce licensing costs, communications infrastructure, specialised technical roles, cyber-security requirements, server capacity and more intensive road-management standards. The correct comparison is therefore not driver cost versus no driver cost. It is the total operating system before and after autonomy.

Productivity depends on interaction, not just truck speed

Autonomous trucks are sometimes programmed to operate more conservatively than experienced human drivers. Their safety logic can also bring them to a stop when conventional equipment enters a protected operating envelope, or when an unexpected obstacle appears in the haul route. A mine that ignores those interactions can end up automating one element of a system while making the system itself less fluid.

The best projects therefore model the mine before they purchase the technology. Dynamic simulation can establish a credible baseline for cycle time, queueing, congestion and haul-road performance. This makes it possible to test whether the proposed autonomous operating model is likely to produce the productivity gains used in the investment case.

Seven management levers determine whether adoption works

First, management must be explicit about the problem autonomy is intended to solve. A safety objective, a labour constraint and a fuel-efficiency problem require different deployment choices and different performance measures. Second, the technical specification needs to be developed with the OEM or technology provider as a strategic partner rather than a conventional equipment vendor.

Third, the commercial contract should define implementation responsibilities, service levels, performance criteria and risk allocation in detail. Fourth, mine planning has to incorporate the geometry and operating requirements of autonomous fleets. Greenfield mines have the greatest advantage because roads, communications and operating zones can be designed around autonomy from the beginning.

Fifth, IT and operational technology have to be treated as production infrastructure. Communications loss, cyberattacks and sensor obstruction are production risks, not merely technology-department problems. Sixth, technical roles need to be identified before commissioning so that the mine is not trying to recruit or train critical people after the system goes live. Seventh, change management must be treated as part of the operating model. New technology changes responsibilities, performance measures and daily routines across the organisation.

The African implication

For African operations, this disciplined approach matters because the cost of implementation failure can be amplified by remote locations, imported parts, limited specialist labour and expensive downtime. The most credible pathway is often staged: automate one route or operating zone, establish the communications and support capability, compare actual performance with the baseline and then expand only when the value is demonstrated.

The opportunity is significant. The constraint is not whether autonomous haulage can work. It is whether the mine is prepared to redesign enough of its operating system for the technology to work as intended.

Mining becomes a connected system

Mining 4.0 applies the principles of Industry 4.0 to extraction: machines, infrastructure and people are connected through digital networks, while operational data is collected continuously and used to improve decisions. The result is a mine in which production planning, maintenance, safety and logistics become increasingly visible in real time.

Sensors can track temperature, vibration, fuel consumption, equipment health and environmental conditions. Industrial IoT platforms bring that information together so that operators can identify deviations before they develop into production losses. The mine therefore becomes less dependent on periodic inspection and more capable of continuous condition monitoring.

Autonomy is the most visible layer

Autonomous haul trucks, drill rigs and robotic inspection systems are the most visible expressions of the digital mine because they remove the operator from the machine. But the value of these machines depends on the information systems behind them. High-precision positioning, fleet management, communications and central control allow equipment to coordinate with the wider production plan.

Artificial intelligence and advanced analytics then use the data produced by those systems to detect patterns. Predictive maintenance is one of the clearest examples: instead of waiting for a component to fail, maintenance teams can intervene when operating data indicates a rising probability of failure.

Regional adoption is taking different forms

North American mines have been early adopters of autonomous haulage and remote operations, particularly in large open pits. In South America, major copper and iron-ore operations in Chile, Peru and Brazil have built increasingly sophisticated automation and digital monitoring systems. Europe’s mining sector has focused more heavily on underground automation, battery-electric equipment and technologies that improve environmental performance.

These differences matter for Africa. Mining 4.0 is not one technology package. Each region is combining automation, connectivity and electrification according to its geology, labour market, energy system and regulatory environment. African operators therefore have room to select the parts of the digital mine that solve the most urgent local constraints rather than reproduce another region’s architecture in full.

The new risks are digital

Greater connectivity creates new dependencies. Cyber-security, data integrity and communications availability become production issues. Skilled technicians, network engineers and data specialists become as important to the operation as traditional equipment maintainers. Remote sites can also struggle with the communications infrastructure needed to support real-time systems.

The trajectory is nevertheless clear. Mines are becoming more automated, more instrumented and more software dependent. The competitive advantage will come not from owning the most digital equipment, but from integrating these technologies into an operating model that makes better decisions faster.

Myth one: autonomy simply removes jobs

The most persistent concern is that driverless trucks automatically mean large-scale job losses. In practice, many mines have used automation to shift people into maintenance, fleet control, data analysis, training and planning. Driving roles decline, but the autonomous operating system creates other technical and supervisory functions.

That does not make the workforce transition painless. Mines still need credible retraining programmes and clear communication. The important distinction is that automation changes the pattern of work rather than reducing the operation to machines without people.

Myth two: the capital cost makes autonomy uneconomic

Upfront cost is only one part of the investment case. Autonomous fleets can reduce idling, smooth driving behaviour, lower collision exposure and increase utilisation. Retrofit platforms also allow mines to automate existing trucks instead of replacing the fleet at once.

The correct question is therefore not whether autonomy is expensive. It is whether the full lifetime savings in fuel, maintenance, productivity and safety exceed the cost of sensors, software, infrastructure and specialist support.

Myths three and four: real mines are too difficult, and people are safer

Modern AHS is designed for dust, rain, snow, degraded roads and limited visibility. Sensor coverage and vehicle-to-vehicle communications can provide consistent awareness even when operating conditions deteriorate. Yet safety still depends on strong procedures, exclusion zones and training.

Autonomy removes fatigue and distraction from the cab and reduces human exposure in high-energy areas. It does not remove every hazard. The risk moves from individual driving behaviour towards system configuration, communications, sensors and operating discipline.

Myths five and six: autonomy is only for giant mines and single-OEM fleets

Modular retrofit systems have changed the scale threshold for adoption. Smaller mines can automate selected fleets or routes without building the kind of site-wide system associated with the largest Pilbara operations.

Mixed fleets are also becoming a central design requirement. Modern OEM-agnostic platforms are intended to work across multiple truck makes and payload classes while coordinating conventional and autonomous equipment inside the same mine. That flexibility matters in Africa, where fleets are often built over long replacement cycles and rarely consist of one brand or one generation of equipment.

The practical conclusion is that autonomy is becoming less dependent on owning a perfect greenfield fleet. The technology still requires discipline, but the range of mines capable of adopting it is expanding.

Geology can overwhelm a perfectly sensible fleet strategy

Kamoa-Kakula deliberately diversified its underground equipment between Sandvik and Epiroc to avoid depending on a single supplier. That strategy worked on its own terms: equipment supply was not the critical bottleneck during the mine’s recovery from flooding.

But supplier diversification could not protect the operation from the seismic and hydrological event itself. Dewatering, underground rehabilitation and geotechnical assessment became the dominant constraints. The case is a reminder that technology strategy needs to sit inside a much wider risk map.

The Syama lesson is even harder

Resolute Mining’s Syama underground project pursued an ambitious automation architecture, but the system ultimately proved incompatible with the way the mine evolved. The automation programme was abandoned and the recovery depended on conventional fundamentals: mine planning, processing-plant reliability and recapitalisation.

This is an important correction to technology optimism. A mine can deploy a sophisticated system and still create poor economics if the system does not match the geology, mining method or operating workflow.

Country risk must be analysed at site level

African mining projects often operate in jurisdictions that attract broad political or security headlines. Those risks matter, but management still needs to distinguish national conditions from the specific exposure of the mine. The relevant questions are whether roads, borders, energy supply, contractors or local communities can interrupt the operation.

Autonomy can reduce some personnel exposure and improve remote supervision, but it does not remove country risk. In some cases, greater dependence on imported electronics, networks and specialist support can create new supply-chain vulnerabilities.

Change management is operational risk management

As roles shift from driving towards control, maintenance and technical support, mines need a deliberate plan for training, recruitment and workforce communication. Poorly managed transitions can create resistance, skills gaps and unsafe interaction between conventional and autonomous equipment.

Autonomy therefore belongs in the same risk conversation as geology, energy, logistics and workforce capability. The technology is only one layer of the operating system.

Capital starts beyond the vehicle

Nevada Gold Mines illustrates the high-capital end of AHS deployment. Large autonomous fleets need more than trucks: private communications networks, control systems, systems integration and site infrastructure become part of the investment. Conventional mine radio networks are rarely sufficient for low-latency autonomous operations.

Gudai-Darri demonstrates the opposite advantage. When the digital and autonomous architecture is designed into a mine from the first construction drawings, future expansion can be cheaper because the underlying systems already have capacity.

The most consistent returns are operational

OEMs report improvements in tyre life, maintenance requirements and productivity when autonomous trucks operate inside their design envelope. Consistent driving removes some of the variability associated with acceleration, braking and cornering. The fleet also avoids conventional shift-change delays and can operate through periods when driver availability would otherwise reduce utilisation.

Safety can carry its own economic value through fewer high-energy interactions, lower incident exposure and more predictable operations.

Labour economics are not a simple subtraction

A conventional 24-hour truck may require several operators across shift rotations. An autonomous fleet can be supervised by a much smaller number of controllers, but the system creates demand for network specialists, maintenance technicians and systems support.

The result is a structural change in labour rather than a clean removal of labour cost. Mines that ignore the cost of the new technical organisation will overstate the financial return.

African projects should sequence the investment

Few African mines need to begin with a Pilbara-scale system. A more disciplined approach is to automate the most expensive or dangerous part of the transport cycle first, measure the result and then decide whether further capital is justified.

The return is generated by the operating system around the truck. A driverless machine alone does not create an autonomous mine.

Why two suppliers made sense

Kamoa-Kakula operates in the DRC’s Katanga Copperbelt, a landlocked mining region where logistics interruptions and imported equipment lead times matter. The operation deliberately built parallel fleets around Sandvik and Epiroc equipment rather than placing all mobile-equipment risk with one vendor.

That approach creates additional training, parts and maintenance complexity, but it reduces the danger that one supplier problem immobilises the entire mine. In an environment where long supply chains can amplify disruption, redundancy has strategic value.

The event the fleet strategy could not hedge

Seismic activity and flooding interrupted underground operations at Kakula in 2025. The recovery challenge centred on water management, geotechnical stability and underground access. Neither Sandvik nor Epiroc was the decisive constraint.

The distinction matters. Supplier diversification is insurance against supplier dependence, parts shortages and vendor-specific disruptions. It is not insurance against geology.

Resilience needs several layers

A serious mine-risk framework therefore separates equipment, energy, logistics, geotechnical, hydrological and market exposures. Each requires a different mitigation strategy. A second OEM can improve fleet resilience; redundant power can improve energy resilience; alternative transport corridors can improve logistics resilience.

Trying to solve every risk through equipment procurement creates false confidence. Kamoa-Kakula shows why operational resilience needs to be designed as a portfolio of safeguards rather than a single diversification decision.

The wider lesson

The mine’s recovery also illustrates why autonomous and digitally advanced operations still depend on physical fundamentals. Pumps, ground support, access, ventilation and geotechnical knowledge remain decisive.

The future mine will be more digital, but it will not be less geological.

A mine designed around automation

Resolute Mining and Sandvik set out to make Syama one of the most advanced underground gold operations in the world. The original concept placed automation at the centre of the underground mine rather than adding it after production had begun.

That approach offered obvious advantages: remote operation, reduced underground exposure and the possibility of more continuous production. But it also made the mine plan and the automation architecture deeply interdependent.

When the mine changed, the system no longer fitted

The underground mining method and operating conditions evolved in ways that made the original automation concept increasingly difficult to apply. The problem was not that autonomous mining technology was inherently incapable. It was that the specific architecture being implemented no longer matched the mine.

Resolute ultimately abandoned the programme. The decision is significant because mining technology projects often suffer from escalation of commitment: once large sums have been spent, management can become reluctant to stop.

Recovery came from fundamentals

Syama’s later improvement depended on work that was less technologically dramatic. Mine planning was revised. Plant reliability had to improve. Capital structure and operational discipline became central.

That recovery provides a useful counterweight to the industry’s strongest success stories. Technology cannot substitute for a workable mine plan or reliable processing system.

A due-diligence test for African mines

When evaluating an automation proposal, management should ask whether the technology has been proven in geology, mine geometry and production workflows close enough to its own operation to make the comparison useful. A case study from a very different mine can demonstrate technical capability without demonstrating local economic fit.

Syama’s lesson is not to avoid automation. It is to demand a stronger fit between technology and mine design before committing the operation to it.

Why Kansanshi was a credible test site

Hitachi’s full-battery EH4000 AC3 was commissioned at Kansanshi after an extended trial. The truck’s scale made the project notable, but the more important detail was the mine around it. Kansanshi already had trolley-assist infrastructure and access to an electricity system dominated by hydropower.

Those conditions reduced two of the biggest barriers to battery haulage: how to deliver enough energy to a very large truck and whether the electricity itself provides a meaningful emissions advantage.

The vehicle changes the mine’s energy system

A battery-electric haul truck cannot be evaluated only against the purchase price of a diesel machine. Charging strategy, grid connection, trolley infrastructure, battery life and operating cycle all become part of the economics.

A poorly designed charging system can simply move the bottleneck from refuelling to electrical availability. The mine needs enough power at the right places and times without disrupting production.

The trial created operational evidence

The Kansanshi programme accumulated thousands of kilometres of running and moved tens of thousands of tonnes before commissioning. That evidence matters because battery performance in a mine depends on grade, payload, ambient conditions and duty cycle.

Real operating data makes it possible to compare energy consumption, availability and maintenance behaviour with the diesel baseline.

A more useful African lesson

Kansanshi should not be read as proof that every African open pit is ready for battery-electric ultra-class haulage. It shows which enabling conditions make the technology unusually credible.

Mines with reliable low-cost power, trolley infrastructure or a strong decarbonisation case begin from a different position from operations dependent on expensive or unstable electricity. The correct strategy is therefore site-specific: electrification should solve an energy and operating problem, not simply satisfy a technology trend.

Autonomy as a design principle

Rio Tinto’s Gudai-Darri iron-ore mine was built around autonomous haul trucks, autonomous drills, automated water carts, digital sampling and remote coordination with the wider Pilbara operating system. The technology was part of the mine architecture rather than an overlay.

That distinction matters because autonomy changes road design, communications, maintenance, control rooms and the interfaces between equipment. Building those systems in at the start avoids much of the disruption associated with brownfield retrofits.

Expansion reveals the value of headroom

Gudai-Darri was originally developed around annual capacity of roughly 43 million tonnes. Rio Tinto later identified a pathway to increase capacity towards 50 million tonnes without repeating the original greenfield construction programme.

The underlying digital and autonomous infrastructure had been designed with room to expand. The mine could therefore increase output through targeted additions rather than building an entirely separate operating system.

Capital intensity becomes the key comparison

Rio Tinto put the capital intensity of the expansion at around US$10 per tonne. The number is important because it illustrates the option value created by infrastructure that is sized beyond the immediate production requirement.

Spending more on enabling systems during construction can be difficult to justify, particularly when a project is already capital intensive. But if communications, roads, energy and control systems have no headroom, later expansion can become disproportionately expensive.

A greenfield lesson for Africa

Large new African copper, iron-ore and gold projects have a rare opportunity to make these decisions before production begins. Not every mine will justify full autonomy on day one, but designing the physical and digital architecture so that autonomy can be added later may be much cheaper than redesigning the operation after capacity is committed.

Gudai-Darri’s broader lesson is therefore about optionality. A mine designed for future technologies can expand more easily than one designed only for its first production year.