By Charles Pitts
Autonomous mining technology is entering a scale test. CiDi says its MetaMine platform now covers more than 1,900 autonomous mining trucks across nearly 40 mines, while projects at Vale’s Salobo copper complex, Western Australia’s Havana Pit and Sweden’s planned Viscaria mine show how different operating models are emerging.
The common thread is no longer the ability of a truck to navigate without a driver. The more consequential question is whether autonomy can be integrated with mine design, workforce planning, energy systems, processing capacity and mixed equipment fleets.
That distinction matters for operators and investors. A retrofit program can lower the barrier to adoption, but may create integration constraints. A greenfield mine can design roads, communications and production areas around autonomous equipment, but must commit capital before operating data is available. Underground mines face a separate set of challenges involving confined spaces, communications, emergency response and remote supervision.
The 2026 deployment picture
CiDi’s reported milestone is among the clearest indications that autonomous haulage is moving beyond isolated demonstrations. The Chinese technology company said cumulative shipments had exceeded 1,900 trucks by June 2026, with deployments at nearly 40 mines. Seven sites reportedly operate fleets of more than 100 autonomous trucks, while the largest single-mine deployment exceeds 220 vehicles.
The figures are company-reported cumulative shipments rather than an independently audited global operating-fleet count. Reuters previously reported more than 1,700 autonomous vehicles across roughly 30 mines and quarries, suggesting that the difference may partly reflect reporting dates and definitions.
CiDi’s figures nonetheless point to a major shift in commercial scale. The company says its MetaMine system supports overburden removal, coal and waste hauling, dumping, crusher feeding and energy replenishment. It has also reported a remote-intervention ratio of approximately one supervisor for every 100 trucks in some operating environments.
That ratio should not be interpreted as a universal labor standard. It depends on mine layout, traffic complexity, fleet uniformity, communications reliability and the number of exceptions requiring human intervention. It does, however, illustrate the direction of the business: fewer personnel operating individual vehicles and more workers supervising a connected production system.
Three models now competing for adoption
The principal deployment models can be separated into retrofit, greenfield and underground automation-first projects.
| Model | Representative example | Core advantage | Main constraint |
|---|---|---|---|
| Retrofit | EACON’s six Komatsu HD1500 trucks at Havana Pit | Extends the life of existing equipment and allows phased conversion | Integration with legacy controls, mixed fleets and existing traffic systems |
| Greenfield or major expansion | New autonomous haulage capacity at Vale’s Salobo copper complex | Production, roads and dispatch can be planned around autonomous cycles | High upfront capital and dependence on compatible OEM equipment |
| Automation-first underground | Viscaria’s planned Sandvik fleet in Sweden | Mine layout and operating procedures are designed for autonomy from the start | Complex underground communications, emergency response and workforce transition |
The table also highlights why there is unlikely to be a single path to automation. A large established open-pit mine may prioritize incremental conversion, while a new underground operation can make autonomy part of the initial mine plan.
Vale Salobo connects autonomy to copper growth
At Vale Base Metals’ Salobo copper complex in Pará, Brazil, 19 Komatsu 930E-AT trucks are operating driverlessly under Komatsu’s FrontRunner Autonomous Haulage System.
The deployment is significant because the technology is linked to a defined production objective. Vale is targeting an increase of approximately 7% in total mine movement, while its Salobo III coarse-particle flotation project is expected to add about 6 million tonnes per year of ore-processing capacity.
The operating logic is straightforward: additional plant capacity requires a more consistent flow of ore from the pit. Autonomous haulage can support that flow through centralized dispatch, stable cycle times and fewer interruptions during shift changes.
The Salobo case also exposes a central interoperability issue. The mine has a mixed truck fleet, but the autonomous operation is currently centered on Komatsu units. Caterpillar 797 and 794 trucks remain conventionally operated, according to industry reporting.
That creates a choice for mine planners. They can standardize on one OEM’s autonomous system, introduce a second platform or use a third-party system designed to operate across equipment brands. Each option affects training, control-room architecture, maintenance, cybersecurity and safety certification.
An autonomous fleet may improve performance within its own operating envelope while still leaving the wider mine dependent on manual equipment. Mine-wide productivity therefore depends on how well the autonomous and conventional systems interact.
Remote operations centers shift mining work from vehicle cabins to fleet supervision, dispatch and exception management.
EACON tests the brownfield route at night
EACON’s project at Havana Pit, part of Norton Gold Fields’ Mulgarrie operations in Western Australia, represents a different adoption path.
The project uses six retrofitted Komatsu HD1500 rigid dump trucks equipped with EACON’s ORCASTRA autonomous haulage technology. Thiess is the mining contractor, while Norton Gold Fields is owned by Zijin Mining Group.
The fleet progressed from autonomous day-shift operations into night-shift production. That step matters because low-light operations place additional demands on perception, localization, obstacle detection and traffic management.
Night work also tests the practical value of autonomy. A driverless fleet that can operate safely beyond daylight hours may increase equipment utilization while reducing workers’ exposure to fatigue, darkness, dust and isolated haul roads. But the safety case must account for interactions with light vehicles, maintenance crews, manually operated equipment and changing road conditions.
Retrofit programs can be attractive to mines that already own large truck fleets or rely on contractors. They may allow operators to automate selected haulage circuits without replacing every vehicle at once.
The trade-off is engineering complexity. Existing trucks may have different braking systems, electronic architectures, payload characteristics and maintenance histories. The autonomy supplier must integrate with those systems without compromising the base vehicle’s reliability or emergency controls.
The Havana Pit project therefore serves as a test of whether brownfield autonomy can be repeated across different fleets and sites, rather than remaining a bespoke engineering exercise.
Viscaria makes autonomy a mine-design decision
Sweden’s Viscaria copper project near Kiruna is taking the most deliberate route: an automation-first underground mine plan.
Viscaria has selected Sandvik to provide an underground fleet and automation systems built around the AutoMine Multi-Lite platform. The planned equipment package includes four DL432i automated longhole drill rigs and four Toro LH621i underground loaders.
The project is designed around autonomous production from the first machine delivered, rather than beginning with a conventional fleet and converting it later. Equipment deliveries are scheduled to begin in 2027, with the mine targeting a production restart in 2028.
Underground autonomy changes the design problem. In an open pit, positioning systems and communications can often be deployed across broad, visible haul roads. Underground, the system must operate through changing headings, confined intersections, ventilation zones and areas where satellite positioning is unavailable.
Viscaria’s plan uses dedicated production areas and pre-programmed routes between ore faces and dump points. Surface-based operators will supervise the fleet, with automation handling repeatable machine cycles and human intervention reserved for exceptions.
That model may improve utilization and reduce exposure to underground hazards, but it also makes mine layout and digital infrastructure critical. A poorly designed roadway, inadequate communications link or unclear emergency procedure can constrain the entire production system.
Underground automation requires mine layouts, communications and production areas to be designed around machine autonomy.
Labor and safety risks remain central
Autonomy changes mining labor; it does not eliminate it.
The number of conventional haul-truck driving positions may decline, while demand rises for control-room operators, sensor technicians, network specialists, data analysts, software integrators and safety-assurance personnel. The transition can create higher-skilled roles, but those positions may not be located where existing workers live or may require new certifications.
Retraining and redeployment can reduce disruption, but workforce planning must begin before equipment arrives. Operators also need to define who has authority to intervene, stop a fleet, approve route changes and restart operations after an incident.
Safety benefits are potentially significant. Removing workers from truck cabs can reduce exposure to collisions, dust, vibration, fatigue and extreme weather. Yet autonomous systems introduce new hazards, including sensor failure, communications loss, software faults and incorrect responses to unexpected obstacles.
Mixed fleets are particularly demanding. Autonomous trucks must recognize and respond to manually driven trucks, light vehicles, pedestrians and maintenance equipment. The system must also cope with blasts, road construction, water, dust and temporary route changes.
Battery and interoperability risks
Electrification adds another layer of complexity to autonomous mining.
Battery-electric trucks require charging or battery-swapping infrastructure, grid capacity, thermal-management systems and energy scheduling. Charging windows must be coordinated with production targets, maintenance and shift planning. At remote mines, the availability and resilience of the power system may be more important than the truck’s nominal battery capacity.
Autonomy can improve energy efficiency by controlling acceleration, braking and routing more consistently. But a vehicle that must stop because of a depleted battery, charging queue or communications outage can disrupt the whole dispatch network.
Interoperability is the other major risk. Proprietary platforms can be optimized for a particular truck and mine environment, but they may limit the operator’s ability to mix equipment brands or change suppliers later. Open data interfaces and common traffic-management protocols could reduce that risk, although implementation remains uneven across the industry.
What decision-makers should measure
Fleet size is an important milestone, but it is not sufficient to evaluate an autonomous mining project. Operators should track:
- Autonomous operating hours and utilization
- Tonnes moved per available hour
- Remote-intervention frequency
- Mixed-fleet productivity
- Tire, brake and component life
- Energy consumption per tonne
- Network uptime and recovery time
- Safety incidents and near misses
- Training, redeployment and workforce retention
- Cost of integrating autonomous systems with processing and maintenance
The strongest projects will demonstrate repeatable performance across full production cycles, not only successful pilot runs.
CiDi’s reported 1,900-truck scale shows that autonomous haulage is becoming a large commercial market. Vale’s Salobo deployment shows how autonomy can be tied directly to copper throughput. EACON’s night-shift operations demonstrate the brownfield path, while Viscaria’s plan suggests that future mines may be designed around autonomy before the first production blast.
The next competitive advantage will not come from driverless vehicles alone. It will come from coordinating trucks, drills, loaders, energy systems, processing plants and people as one operating network.
Sources: Reuters on CiDi’s overseas expansion, CiDi company updates, Vale’s autonomous truck program, EACON Mining, and Viscaria. Further reading is available through Skillings’ mining technology coverage.
