Dyno Nobel to Trial First Autonomous Blast System at Utah Copper Mine

A five-year research programme has produced what Dyno Nobel describes as the first fully autonomous end-to-end blast hole inspection and loading system in the world, and its first commercial trial is not headed for a laboratory or a test pit. It is headed for a live copper operation in southeastern Utah.

For the first time, according to the company, no personnel would need to be on the bench during measuring, priming, loading, stemming, and documenting a blast. Those tasks would run from a remote control centre at a safe distance. The announcement was made to investors in New York by Dirk Van Soelen, Chief Technology Officer at Dyno Nobel. As of 1 October 2026, no independent public confirmation of the announcement had appeared, so the Dyno Nobel-specific claims here are attributed to the company rather than presented as established public record.

The timing is what makes this more than a technology press release. The chosen trial site, Copper One, is already the world’s first autonomy-first mine and refinery, running autonomous drills, driverless haul trucks, and robotic inspection units under a software layer called MarianaOS. Blasting has been the last major manual step in that chain.

Copper One is targeting 50,000 metric tonnes of refined copper a year, and its operator, Mariana Minerals, raised a $310 million Series B in August 2026 to accelerate domestic production. What follows explains why this particular site, at this particular moment in U.S. mining, is the logical place to attempt closing that gap.

What Dyno Nobel’s autonomous system actually does

The clearest way to understand the system is to picture the bench running with nobody standing on it. Every task that a crew would normally perform by hand is instead executed by sensor-equipped machinery, coordinated remotely.

According to Van Soelen, the system handles the full operational chain for each blast hole:

1. Measuring the hole to confirm depth and condition
2. Priming it with the initiating charge
3. Loading the explosive product
4. Stemming the hole to contain the blast energy
5. Documenting the entire sequence as a data record

Every element of the bench system carries sensors throughout, which changes what blasting produces as a by-product. The manual version of this work left little behind in the data record. The autonomous version logs a repeatable, documented file for each hole, feeding directly into blast design and mine planning rather than disappearing once the shot is fired.

“The system is intended to transform one of mining’s most safety-sensitive processes into a data-driven, reliable, and consistent operation,” Van Soelen told investors, according to the company’s account of the New York presentation.

That integration logic is not new to Copper One. It is the same principle already running beneath the Sandvik AutoMine drilling deployment there, where penetration rates and geological signals flow into the planning layer. Dyno Nobel says the blast system is built to slot into existing mine infrastructure and to work alongside other autonomous solutions already deployed at customer sites, positioning it as additive rather than a reason to rebuild a site from scratch.

The Sandvik AutoMine deployment at Copper One already demonstrated that autonomous drilling could produce reliable geological data feeds suitable for downstream planning models, which is the same data-discipline logic Dyno Nobel is now proposing to extend into the charging and loading phase.

The company has described a staged commercial rollout running through to mid-2027, with U.S. trials commencing first and other countries to follow. Dyno Nobel also says it worked with leading technology and robotics firms during development, though it has not publicly named those partners, and none have been independently confirmed.

Why Copper One is the right first site for this trial

Copper One was not selected as a convenient volunteer for someone else’s prototype. It was architecturally designed to need exactly this kind of system, which makes the trial read as the final piece of a deliberate puzzle rather than a bolt-on experiment.

The site sits in Lisbon Valley, San Juan County, roughly 40 miles southeast of Moab in southeastern Utah. It runs on MarianaOS, an integration layer that connects Sandvik AutoMine drills, Boston Dynamics Spot robots, driverless haul trucks, and a private Celona 5G network, with human operators supervising from digital control rooms.

The economic case here is not theoretical. Autonomous drills and robotic haul trucks have already cut mining costs by approximately 50% and refining costs by approximately 30% at Copper One, according to Deseret News reporting from 27 April 2026. That tells you autonomous blasting is being evaluated as the next extension of a cost-reduction model that is already demonstrably working, not as a speculative add-on.

The strategic backing is visible too. The $310 million Series B announced on 3 August 2026 was framed as accelerating critical minerals production across Copper One in Utah and Lithium One in Texas. Mariana restarted Copper One within roughly four months of acquiring the previously idled mine in late 2025, and marked the moment with a ceremonial “First Blast” attended by Utah Governor Spencer Cox in April 2026.

The Copper One brownfield restart, which Mariana completed within roughly four months of acquiring the previously idled site, established the autonomy-first infrastructure that makes a fully autonomous blasting trial operationally coherent rather than speculative.

Juan Lozano, Co-Founder and CTO of Mariana Minerals, identifies blasting as the initial step in the site’s autonomous operational chain. Make that step autonomous and you have an upstream trigger that sets the rest of the production sequence in motion.

Closing the autonomy gap with MineOS integration

Mariana is building MineOS integrations so that the Dyno Nobel system’s sensor data flows into the same planning and optimisation layer already handling drill data, haul fleet routing, and processing plant control.

The effect is to make the blast step data-visible inside the mine’s operational model for the first time. Where blasting previously sat as a gap in an otherwise connected autonomous data chain, the trial is designed to fill it.

The safety and productivity case behind autonomous blasting

The safety argument starts on the bench itself, which is among the most hazardous places in the whole drill-and-blast cycle. People handling explosives next to unstable ground or over hidden voids are exactly what the system is designed to remove from the equation.

The specific hazards being addressed include:

• Personnel exposure during explosives preparation and loading on the bench
• Proximity to underground voids that can collapse without warning
• Unstable ground conditions that render sections of even active mines inaccessible
• Direct explosives-handling risk during priming and stemming

Dyno Nobel’s framing, per Van Soelen, is that even fully operational mines contain areas that have been off-limits because of hazardous conditions, and that an autonomous system can reach those areas safely. The resolution is to relocate all bench operations to a remote control centre, following the same human-oversight model already used for Sandvik AutoMine drills and Boston Dynamics Spot patrols at Copper One. A Boston Dynamics case study from 22 July 2026 describes Spot autonomously patrolling the pit, heap-leach pads, and SX-EW plant, collecting thermal, acoustic, and visual data while keeping workers out of hazardous inspection zones.

Dyno Nobel is not the only company pursuing autonomous explosives loading; Newmont’s deployment of Jevons robotics at a separate operation provides an early comparison point for how different technical approaches to bench automation handle the same core safety and precision challenges.

Here is where safety and productivity turn out to be the same design decision. The sensor-rich record the system produces does not just protect people; it improves blast design, fragmentation quality, and downstream processing efficiency. The Mariana-Sandvik partnership material already describes drill penetration rates and geological signals feeding models that improve blast design and material routing. Autonomous loading extends that same data discipline to the charging phase.

One dependency deserves emphasis for anyone weighing this at scale.

Celona describes its private 5G deployment at Copper One as “critical safety infrastructure,” noting that connectivity gaps could pose risks to autonomous equipment operations, per its press release of 23 April 2026.

That framing matters because the safety model relies on deterministic, low-latency communication. Connectivity is not background detail at a fully autonomous blasting site. It is a load-bearing requirement, and any operator studying this model should treat it as such.

What a successful trial at Copper One would mean for U.S. critical minerals supply

Step back from the bench and the trial sits inside a larger question: whether autonomous blasting can help close the gap between domestic critical mineral demand and the supply capacity that currently exists.

The structural drivers behind U.S. autonomous mining investment are stacking up:

• Labour constraints in rural mining regions where skilled workers are scarce
• Persistent critical mineral supply gaps against rising demand
• Capital flowing into data-rich mines positioned as AI-enabling infrastructure
• The prospect of unlocking deposits and bench sections previously inaccessible for safety reasons

The labour point is not a footnote. Deseret News reported on 27 April 2026 that Copper One was revived “after human labour ran dry,” which tells you autonomy here is partly filling a gap that cannot be filled with workers at all in many rural U.S. regions. That reframes autonomous blasting as a question of industrial policy, not just mining technology.

Critical minerals supply resiliency has moved from policy discussion to capital allocation decision, with autonomous production models like Copper One representing one of the primary structural responses to the gap between domestic extraction capacity and projected demand for copper and battery materials.

The capital is following the same logic.

Fortune, on 3 August 2026, connected the wave of investment in autonomous, data-rich mines capable of producing tens of thousands of tonnes of copper a year to the capital required to “power AI.”

Mariana’s $310 million Series B was explicitly framed as accelerating critical minerals production, with the autonomous model being extended from Copper One to Lithium One in Texas. The specific capability autonomous blasting adds is access: the ability to safely extract from deposits or bench sections long considered too dangerous to work, expanding the recoverable resource base of existing operations without new greenfield development.

Against a production benchmark of 50,000 metric tonnes a year, the Copper One trial becomes the reference data point for how fast this technology can travel to other critical mineral operations. Dyno Nobel’s staged rollout through to mid-2027, U.S. first, means the outcome here carries read-through well beyond one Utah copper mine.

What the Copper One trial resolves, and what it leaves open

None of this is settled yet. The Dyno Nobel system is at prototype and early-trial stage, the commercial rollout runs through to mid-2027, and no independent public confirmation of the announcement existed as of 1 October 2026. The honest read is that the technology sits early on the maturity curve, however strong the surrounding logic looks.

The trial will need to answer three questions in particular:

• Whether the system performs reliably at a live commercial copper operation rather than a controlled test environment
• How cleanly it integrates with MarianaOS at the data level
• Whether the regulatory framework accommodates fully autonomous blast loading without new compliance structures

That third point is the most consequential and the least resolved. No publicly available MSHA or state regulatory guidance on autonomous explosives handling was surfaced in independent research as of 1 October 2026. That is not a confirmed barrier, but it is the single most important open question for any operator considering replicating the model, and the Copper One trial will have to navigate it.

MSHA guidelines for autonomous mining systems require continuous two-way communication, immediate shutdown capability, and real-time situational awareness, setting the baseline compliance framework that any autonomous explosives handling deployment would need to satisfy under 30 CFR 56/57.

What tilts the odds is the setup. Copper One is described across multiple independent sources as the only autonomy-first mine and refinery at commercial scale, it already runs an integration layer handling multiple autonomous vendors, and the $310 million Series B gives the trial financial runway. That combination makes it the best available proving ground, and its outcome through 2026 and 2027 will set the industry reference point.

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.

Forward-looking statements regarding the technology, trial, and rollout are speculative and subject to change based on market and operational developments. The Dyno Nobel announcement and trial agreement are sourced from the company’s investor presentation and had not been independently confirmed in public sources as of 1 October 2026.