Sandvik Reveals Sami, a Cabinless Autonomous Surface Drill Concept

There is no seat. No cabin. No enclosed space anywhere on the machine where a person could sit, stand, or shelter.

That absence is the whole point. Sandvik built its new Sami surface drilling concept from the outset to operate without a human present, which is a fundamentally different proposition to a conventional rig with a cab that happens to run itself. One is a machine designed around autonomy. The other is autonomy bolted onto a machine designed around a driver.

Sandvik revealed the Sami autonomous drill at Future of Mining 2026 in Tampere, Finland, on 1-2 September 2026. For years, original equipment manufacturers have retrofitted self-driving software onto human-centric drill rigs. Sami is the clearest signal yet that the conceptual frontier has moved to designing equipment with no human accommodation built in at all.

Here is what this reveal actually lets you judge. Whether Sami is a genuine architectural departure or a repackaging of capabilities Sandvik already sells, and what its new fleet-level AI agent, named Sandi, adds that Sandvik’s existing commercial software does not already deliver.

What Sandvik’s Sami concept actually does, and what makes it different

Sami is a fully battery-powered surface drilling machine built around a cabinless, boom-drill body. Onboard tooling sits where an operator station would normally go, and LED display panels along the machine communicate its current operational state to workers on foot nearby.

The robotic tooling is where the machine’s self-sufficiency becomes concrete. It handles bit replacements, collar pipe installation, and hammer swaps without a person touching it, and it measures hole depth and deviation once drilling is complete. The tasks that traditionally pull a human into the drilling zone are the ones Sandvik has automated first.

Robotic rod handling has historically been the hardest automation frontier in drill rig design because the physical tolerances involved in collar pipe installation and bit replacement demand precision that early robotic systems struggled to deliver consistently at production throughput rates.

Navigation is the other half of the story. Sami uses advanced autonomous navigation and obstacle detection, and it differentiates between moving and stationary obstacles, tracking personnel and machinery as it works.

Here are the core autonomous capabilities Sandvik demonstrated:

• Advanced navigation with real-time positioning across the bench
• Obstacle detection that distinguishes moving objects from stationary ones
• Robotic tooling for bit, collar pipe, and hammer handling
• Automated measurement of hole depth and deviation

The cabinless choice is not a cost-cutting exercise. It is a design-for-autonomy decision, and it signals that Sandvik is no longer treating human occupancy as a baseline requirement for surface drill design. Remove the cabin and you remove the duplicated control systems, one built for a person and one for a computer, that a hybrid rig has to carry.

“Sandvik’s goal is to advance safer, more productive, and increasingly automated mining operations,” said Patrick Murphy, President of Sandvik’s Mining business area.

Sandvik was explicit that Sami is a conceptual demonstration of future possibilities, not an announced commercial product. For operators and investors tracking automation capital expenditure, its value is as a clarification of what the next generation of autonomous surface drills could look like structurally, well before any purchase order exists.

How Sandi coordinates the fleet above the machine level

Sami handles its own work. Sandi handles the fleet.

Agentic AI in mining operations represents a broader architectural shift beyond any single OEM’s product line, with fleet-coordination agents like Sandi sitting within a wider industry movement toward AI systems that plan, allocate, and execute work cycles without continuous human instruction at every step.

Sandi is a proprietary AI agent that sits above individual machines. Its job is to distribute drilling assignments across multiple autonomous units, while each unit independently plans and executes its own work cycle without direct human instruction. That splits the intelligence into two tiers: fleet-level task allocation on top, unit-level execution underneath.

To see what Sandi actually adds, it helps to look at what Sandvik already sells. AutoMine Surface Fleet, introduced in May 2025, lets human operators manage more than 15 i-series rigs from any connected site location. AutoMine Autonomous (AutoCycle) already enables one operator to run one to three rigs continuously through an entire drilling pattern.

The difference is who does the allocating. In the existing products, a person oversees the fleet and directs the machines. With Sandi, an AI agent distributes the assignments and the humans move further back.

This context also matters: AutoMine Aura, announced in May-June 2026, brought 3D-perception and adaptive intelligence into the platform. That is the technological environment Sandi extends.

Riku Pulli, President of Sandvik’s Digital Mining Technologies, stated that many of the showcased capabilities are expected to influence future surface mining solutions. That framing positions Sandi as a technology signal rather than a shipping product.

For you, the read is this. Sandi moves fleet coordination from a human-supervised layer to an AI-allocated one, and the question worth weighing is whether that shift, if commercialised, would change the operator-to-rig staffing ratio meaningfully beyond what AutoCycle already achieves.

Sami’s place in Sandvik’s cabinless concept lineage and commercial roadmap

Sami did not arrive from nowhere. It is the latest step in a cabinless concept lineage Sandvik has been building since 2022.

The first was the AutoMine Concept Underground Drill, nicknamed Amelia, unveiled in Tampere in 2022: a cabinless, battery-electric, fully autonomous twin-boom development drill that plans and executes an entire drilling cycle on battery power with no human interaction. The AutoMine Concept loader followed in 2023, a cabinless, battery-electric autonomous underground loader that Pulli described as engineered for automation from the outset.

Read in sequence, Sami confirms a direction rather than announcing a surprise. Sandvik has spent four years building machines around the assumption that no one will be aboard.

From concept to commercial: what the deployment record shows

The concepts are testbeds. The commercial deployments are where Sandvik’s autonomous surface drilling actually earns revenue today, and that record is already global.

• Boliden Kevitsa (Finland): Leopard DI650i rigs running fully autonomous surface drilling with AutoMine, August 2023
• LKAB Leveäniemi (Sweden): three Leopard DI650i rigs ordered with AutoMine readiness, December 2023
• Mariana Minerals Copper One (Utah, USA): a DR410i rotary drill rig with AutoMine Surface Drilling integration, April 2026
• Rio Tinto collaboration: adding an autonomous drilling system to Sandvik surface rigs, with testing beginning at Sandvik’s Finland test pit in 2026 before moving to Western Australia

Every one of these involved multi-year staging, on-site Sandvik technicians, and commissioning support. That is the honest gap between a concept unveiled at a conference and a machine drilling production holes on a live bench.

Sandvik’s 2025 Annual Report targets double-digit growth in surface drilling solutions and accelerated growth in automation and mining software across 2025-2030. The deployment record tells you Sami is not a distant vision but a conceptual extension of a trajectory already producing revenue. For anyone assessing Sandvik’s automation strategy, that record is the tool for separating concept ambition from commercial execution.

Investors and operators tracking the business case for cabinless drill designs will find our full explainer on mining automation safety outcomes, which examines the incident rate data, productivity benchmarks, and total cost of ownership metrics that operators use to justify autonomous capital expenditure.

What the industry’s autonomous drilling frontier looks like beyond Sandvik

Sami does not sit in isolation. The closest competitive analogue is Epiroc, whose SmartROC D65 rigs run whole shifts in open-pit mines with empty cabs, supervised remotely through its LinkOA platform.

Epiroc’s autonomous drilling approach has followed a different architectural path, prioritising retrofittable autonomy on existing rig platforms rather than designing cabinless hardware from the ground up, which reflects a commercial strategy built around installed-base upgrade cycles rather than generational replacement.

According to reporting that has not been independently confirmed, Epiroc’s SmartROC D65 has drilled tens of millions of metres autonomously over roughly a decade, and its Luck Stone deployment in the US in 2025 was described as the first fully autonomous SmartROC D65 in the US quarry market. Epiroc references the Global Mining Guidelines Group (GMG) Level 4 standard for full autonomy in these announcements.

That standard points to the real constraint. The binding limits on how fast cabinless concepts reach widespread deployment are not primarily technical.

Three structural factors set the pace:

• Regulatory and standards acceptance, including the GMG Level 4 autonomy pathway
• Site-level change management, covering commissioning timelines and staged rollout
• Workforce retraining and jurisdictional variation, including union considerations

The workforce dimension deserves honesty. Autonomy reshapes operator roles toward remote supervision, data analysis, and maintenance rather than simply removing them, but the pace of that transition varies sharply by jurisdiction and mine type.

A 2026 thematic report from MiningTerminal links Sandvik’s autonomous deployments at Boliden Kevitsa and LKAB Svappavaara to double-digit gains in productive hours, longer tyre life, and materially lower incident rates.

For investors, the competitive picture confirms that autonomous surface drilling is already a commercial category, not an emerging one. That raises the stakes for Sandvik in turning Sami’s architectural ambitions into a market-ready product ahead of Epiroc’s established autonomous fleet footprint.

What Sami signals about where surface mining automation is heading

Sami matters less as a machine and more as a declaration of where Sandvik believes the endpoint lies. Two threads run through this reveal, and their combination is the signal.

The first is cabinless architecture as a design philosophy, not a feature. The second is AI fleet coordination through Sandi as a management layer, not an interface upgrade. Put them in one concept and you have Sandvik stating that the future of surface drilling is machines with no human aboard, coordinated by an AI agent above them.

The gap is real, and Sandvik confirmed it. Sami is not a commercial product announcement, and Pulli’s own framing, that the showcased capabilities are expected to influence future surface mining solutions, stops well short of a timeline.

The Amelia precedent is the honest measure here. Unveiled in 2022, it still has not been commercially announced as of September 2026, which tells you how long the concept-to-market road can be in this category.

Three questions the conference did not answer:

• When, and under what conditions, does a machine with this architecture enter Sandvik’s commercial pipeline
• What regulatory pathway clears cabinless surface drills for widespread deployment
• How Sandi’s AI allocation logic differs in practice from what AutoMine Aura already provides

The forward indicator worth tracking is whether Sandi’s fleet-coordination architecture migrates into Sandvik’s AutoMine product line, and on what timeline. That is what will separate this reveal from the concepts that never reached the market.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions. These statements are speculative and subject to change based on market developments and company performance.