By Penny Langford
Sandvik’s Sami concept moves autonomous mining technology 2026 beyond driverless haulage and toward a more complete vision of the crewless surface mine. The fully autonomous, battery-electric drill has no operator cabin, carries its own drilling tools and uses robotic handling to complete key parts of the drilling cycle.
Unveiled at Sandvik’s Future of Mining event in Tampere, Finland, Sami is a concept rather than a commercial product. Its significance lies in how several technologies are combined: autonomous navigation, robotic tool changes, digital-twin integration, natural-language control and AI-based fleet coordination through Sandvik’s Sandi agent.
The development comes as mining companies and equipment suppliers move from isolated automation pilots toward integrated fleets. Sandvik is working with Rio Tinto on autonomous open-pit drilling, while Vale Base Metals is expanding a fleet of AutoMine-ready surface drills for copper operations in Brazil. Epiroc is extending its Deep Automation platform across drilling, bolting and autonomous haulage, and Hexagon is adding intervention systems designed to stop or control mobile equipment when safety limits are breached.
The direction is clear: automation is becoming a mine-planning and operating-model decision, not simply an equipment upgrade.
Sami combines autonomy, electrification and robotic handling
Sandvik describes Sami as a next-generation surface drilling concept built on a boom-drill platform. Unlike conventional rigs, the machine has no operator cabin and is intended to operate autonomously under remote supervision.
Its design includes onboard drill bits, collar pipes and down-the-hole hammers. A robotic manipulator can perform bit changes, install collar pipes and replace hammers without requiring a worker to approach the rig for every tool-handling task.
The concept also measures hole depth and deviation after drilling. That data could connect drilling execution more closely with blast design, mine planning and quality control. In a mature autonomous operating model, the drill would not merely complete a pattern; it would continuously provide feedback on whether the planned pattern is being delivered accurately.
More Read
Battery-electric propulsion is another central feature. Removing the diesel engine and exhaust system could reduce local emissions, noise and vibration while changing maintenance requirements and energy planning. The benefits will depend on battery capacity, charging infrastructure, duty cycle and the availability of reliable site power.
Remote operators monitor mine maps, equipment status and production data.
Sandi shifts control from the machine to the mine
The most important part of the Sami concept may be the connection between the drill and Sandi, Sandvik’s AI agent.
Sandvik says Sandi assigns drilling tasks and coordinates activity across the wider mining fleet. Sami then plans and executes its work cycle while adapting to nearby equipment, people and changing site conditions. Operators interact with the system using natural language and intervene when human judgment is required.
That architecture represents a shift from machine-level automation to mine-wide coordination. A conventional automated drill can execute predefined movements. A connected autonomous fleet must decide which machine should perform which task, in what sequence and under which operating constraints.
The continuously updated digital twin is critical to that model. It provides a shared representation of mine geometry, equipment locations, work areas, drilling plans and changing conditions. The value comes from connecting operational data rather than keeping drilling, haulage, maintenance and planning systems separate.
Sami’s capabilities are also intended to scale across Sandvik’s wider surface drilling portfolio, including rotary blasthole drills. Sandvik’s public statements emphasize that the concept itself is not yet a product, but that many of its underlying technologies are expected to influence future commercial solutions.
Sandvik is already testing the operating model
The concept is supported by commercial activity around Sandvik’s AutoMine platform.
In January, Sandvik said it would supply Vale Base Metals with 16 surface drills equipped for autonomous operation at the Salobo and Sossego copper operations in Brazil. The order includes nine DR416i rotary blasthole drills and seven Leopard DI650i down-the-hole rigs, alongside services and rock tools.
Deliveries began in late 2025 and are scheduled to continue through 2029. Sandvik’s AutoMine Autonomous package and AutoCycle capabilities are designed to allow multiple surface drills to be operated from a remote control room.
Sandvik is also working with Rio Tinto to integrate i-series surface drill rigs with Rio Tinto’s Autonomous Drilling System. The program begins with testing at Sandvik’s facility in Finland before field trials in Western Australia, with the longer-term objective of multi-rig and multi-site operation through Rio Tinto’s Perth Operations Centre.
These projects matter because they test interoperability, production performance and remote operations in live mining environments. Sami provides a forward-looking design target; Vale and Rio Tinto provide pathways for parts of that technology stack to be industrialized.
Epiroc’s Deep Automation shows a parallel route
Epiroc is pursuing a related but distinct automation strategy through Deep Automation.
In 2026, Epiroc expanded Deep Automation to include underground drilling and bolting solutions. The platform brings together drill planning, data flow, automated operations and precise execution, with a progression from remote operation toward higher levels of autonomy.
Epiroc has also demonstrated 3D LiDAR-based autonomy for mining trucks operating between underground and surface environments. The LiDAR system is used for localization, mapping and obstacle detection, including automatic stopping when an obstacle is identified.
That distinction is important. Publicly available Epiroc information links 3D LiDAR primarily to autonomous truck haulage. It does not establish that the same sensor package is being used on Epiroc drilling rigs. The broader lesson is that autonomy depends on several specialized layers: planning and control for drilling, perception for navigation, safety intervention for mobile equipment and fleet software for coordination.
Hexagon adds an independent intervention layer
Hexagon’s next-generation Vehicle Intervention System, or VIS, illustrates the safety architecture needed as autonomy expands.
VIS is positioned as an EMESRT Level 9 vehicle intervention system. It can inhibit propulsion, apply retardation or braking and enforce defined operating rules when a vehicle enters an unsafe condition. Applications can include preventing movement into a hazard, controlling overspeed on ramps and placing equipment into a safe state if a critical system fault is detected.
Hexagon’s intervention technology is designed to control or stop mobile equipment in defined hazards.
For autonomous mines, a system such as VIS can serve as an additional safety layer rather than replacing the machine’s own autonomy controls. That separation is valuable because it allows intervention logic to operate independently from route planning or production optimization.
Hexagon says VIS has been deployed in production mines since 2018 and is designed for both new equipment and brownfield retrofits. Its role is particularly relevant in mixed fleets where autonomous and manually operated vehicles share roads, loading zones and maintenance areas.
EACON’s 3,500-plus figure is a fleet milestone
EACON adds scale to the haulage side of the autonomous mining technology 2026 landscape. The company says its ORCASTRA autonomous haulage solution has been deployed on more than 3,500 trucks globally, with those vehicles collectively traveling more than 150 million kilometers.
The figure refers to the number of trucks using EACON’s system. It does not describe a 3,500-tonne truck class.
EACON’s model is built around OEM-agnostic autonomy that can be factory-fitted or retrofitted to different truck models and powertrains, including diesel, hybrid and battery-electric vehicles. That approach is relevant to mines that cannot justify replacing an entire fleet or that operate equipment from multiple manufacturers.
The company’s earlier Australian work with Thiess and Norton Gold Fields involved autonomous haulage on Komatsu HD1500 trucks at the Havana Pit. Such deployments test whether autonomy can be extended from controlled daytime operations into longer operating windows, including night shifts.
Autonomous mining technology 2026: deployment indicators
| Technology or company | Reported scale or capability | Strategic relevance |
|---|---|---|
| Sandvik Sami | Fully autonomous battery-electric surface drilling concept | Demonstrates cabless drilling, robotic tool handling and AI fleet coordination |
| Sandvik and Vale Base Metals | 16 AutoMine-ready surface drills | Commercial pathway for autonomous copper drilling in Brazil |
| Sandvik and Rio Tinto | Multi-rig, multi-site integration program | Tests interoperability between AutoMine and Rio Tinto’s autonomous drilling system |
| Epiroc Deep Automation | Underground drilling, bolting and autonomous haulage solutions | Links planning, execution and remote operations across mine functions |
| Hexagon VIS | EMESRT Level 9 intervention system | Adds independent braking, propulsion inhibition and rule enforcement |
| EACON ORCASTRA | 3,500-plus trucks, according to EACON | Shows the scale of OEM-agnostic autonomous haulage deployment |
Figures are company-reported and refer to different measures, including concepts, orders, integration programs and deployed systems.
Base, bull and bear framework
| Scenario | Operating assumptions | Likely outcome |
|---|---|---|
| Base case | Autonomous drilling expands through brownfield pilots and selected fleet deployments; remote supervision remains essential | More autonomous drills and trucks operate in controlled zones, but mixed fleets and manual intervention remain common |
| Bull case | Digital twins, AI agents, electric equipment and independent safety systems become interoperable across mine sites | Drilling, haulage and maintenance are coordinated through integrated control centers, increasing utilization and reducing worker exposure to high-risk areas |
| Bear case | Network outages, cybersecurity incidents, charging constraints, workforce resistance or weak project economics slow deployment | Automation remains concentrated in repetitive haulage and drilling applications with limited mine-wide integration |
The base case is the most likely near-term outcome. Mining operations rarely move directly from conventional equipment to a fully crewless site. Adoption is more likely to proceed through staged deployment, beginning with repetitive routes, defined drilling zones and remote supervision.
The next test is integration, not demonstration
Sami points toward a future in which the drill is one autonomous node in a larger operating system. Sandi coordinates tasks, a digital twin provides situational context, perception systems identify hazards and intervention systems impose safety limits.
For operators, the critical questions will be practical:
- Can autonomous drills deliver consistent hole quality under changing ground conditions?
- Can multiple OEM systems exchange data without creating new control-room bottlenecks?
- Can battery-electric equipment be charged without reducing production availability?
- Can safety systems intervene predictably in mixed traffic?
- Can workforces transition into remote operations, maintenance and systems roles?
- Can the total mine cost per tonne improve after infrastructure and training are included?
The answer will vary by commodity, mine geometry and fleet scale. A repetitive copper or iron ore operation may offer a clearer path than a rapidly changing pit with complex traffic. A mature mine may prefer retrofit autonomy, while a greenfield project can design roads, power, communications and control centers around autonomous equipment from the beginning.
The significance of Sami is therefore not that a crewless surface mine has arrived. It is that the building blocks are beginning to align: autonomous drilling, electric power, AI coordination, digital twins, 3D perception and independent safety intervention.
LinkedIn snippet
Sandvik’s Sami concept brings autonomous drilling, battery-electric power, robotic tool handling and AI fleet coordination into one surface-mining vision. The machine is not yet a commercial product, but its architecture shows where mining automation is heading: from isolated machines to integrated, remotely supervised operating systems.
X snippet
Sandvik’s Sami concept points beyond driverless haulage: a cabless, battery-electric surface drill with robotic tool handling and AI coordination through Sandi. The next autonomy race is about integrating drilling, haulage, safety and mine planning: not just removing operators from cabs.
Sources and further reading
- Sandvik introduces Sami autonomous surface drilling concept
- Sandvik and Rio Tinto advance autonomous open-pit drilling
- Sandvik and Vale Base Metals expand autonomous surface drilling
- Autonomous mining technology: fleets, milestones and outlook
- Autonomous mining technology: 1,000-truck milestone reshapes fleets
- EACON autonomous haulage project and deployment updates
