The blast bench was one of the last manual steps in an open-pit mine. Dyno Nobel wants to take it out of human hands

After five years of development, Dyno Nobel has unveiled a system that measures, primes, loads, stems and records every blasthole autonomously, run from a control room far from the bench. The prototype will be trialled at a copper mine in Utah.

Between late September and early October 2026, Dyno Nobel announced a fully operational prototype of an autonomous system for the drill and blast bench. According to the company, it took five years of research and development. With no operator in the area, the system performs the tasks a crew currently does by hand on the bench: measuring each hole, priming — that is, installing the initiator and booster — loading bulk explosive, stemming and recording the result. Every component carries sensors, and the system was designed to talk to the mine's systems and to other autonomous solutions already in operation. The first trial will be at Copper One, Mariana Minerals' copper operation and refinery in south-eastern Utah, in the United States, which describes itself as a mine designed for autonomy from the outset. Mariana will integrate its operating system with Dyno Nobel's. The schedule calls for a staged rollout through mid-2027, starting in the US and then moving to other countries. Dyno Nobel's chief technology officer, Dirk Van Soelen, says the company has built a complete process capable of deploying initiating systems and bulk explosives on a bench at scale.

Automation arrived first where work is repetitive and the environment is predictable: trucks on defined routes, drill rigs following a digital blast plan, crushers under advanced control. The loaded blast bench lagged behind for a practical reason. Every hole is slightly different from the design — it may have partly collapsed, have water at the bottom, or have been drilled short. Today, the person who sorts this out is the blaster, who measures the hole with a tape, decides how much explosive to load and adjusts the stemming. It is judgement work, done on foot, in the most dangerous place in the mine. What Dyno Nobel says it has solved is precisely the judgement part: the system measures before loading, and only then decides the charge. The word that matters most in the announcement is "record". In a manual blast, information on what was actually placed in each hole rarely leaves the crew's heads or a spreadsheet. In an autonomous blast, every hole becomes a record with real depth, explosive mass and initiator position. That changes blast engineering in three ways. First, it allows design to be compared with execution, hole by hole, and explains why a shot produced oversize boulders or overbreak on the wall. Second, it lays the groundwork for so-called mine-to-mill: blast fragmentation affects shovel productivity and the energy consumed by the crusher and the mill, and without reliable loading data that optimisation remains guesswork. Third, it opens up areas that are currently idle because of risk — unstable benches, zones with old underground workings — because no one has to set foot there. The trial partner itself sums up the logic: in a mine designed to be autonomous end to end, blasting is the first link, and if it stays manual, the whole chain after it inherits the variability. An honest caveat: this is a prototype that still has to face dust, rain, irregular holes and the routine of a producing mine. The real test will be the trial in Utah.

Blasting accounts for a small share of a mine's total cost, but it influences almost everything that follows — from the shovel's load to the energy spent on grinding, one of the largest bills at a copper or gold plant. Taking people off the loaded bench reduces exposure to one of the most critical risks in the operation. And turning every hole into data is what lets geotechnics, drilling, blasting and processing work from the same information. For mining engineers and blasting technicians, the message is clear: the role does not disappear, but it moves from the bench to the control room, and starts to require data literacy, blast plan programming and integration with other autonomous systems.