The Mine Loses Its Memory Before Production Begins

Picture a mine and you picture production: the trucks, the crusher and the ore moving. Production is the longest phase of the asset’s life. It can run for decades. But it is also the most captive. Whether the mine runs well was largely decided in the years before it started, while it was being studied, designed and built.

Between each of those phases is a gap, not only in the activity, but also in ownership, accountability and the disciplines involved. Call that gap a seam. A mine has many seams, and any of them can leak.

This piece follows one of the least examined: the gap between the model of what the mine is supposed to be and the systems that eventually run it. This is not solely a mining problem. The same seam runs through other capital-intensive projects, including offshore oil and gas developments.

Every mine you read about is “going digital” by converging IT and OT. The business systems and plant-floor control systems are finally talking to each other. It is a real programme and a worthwhile one, but it quietly assumes that the mine already exists. It begins with two domains running a live asset. It says nothing about the years spent building that asset or the domain responsible for defining it.

That domain is the one that builds the mine. IT connects people to data. OT connects people to machines. Both connect people to something that already exists. ET, or Engineering Technology, is different. It defines what is supposed to exist in the first place. Throughout the build, from the first study to the start of production, ET dominates. OT has nothing to run yet. IT usually enters in a narrower role, establishing systems for functions such as procurement and finance.

Domain	Primary concern	Typical systems / data
IT — Information Technology	Providing data access to humans	ERP, finance, HR, CRM, enterprise data, cloud, corporate networks
OT — Operational Technology	Running the physical process	PLCs, DCS, SCADA, sensors, actuators, process control, autonomous equipment
ET — Engineering Technology	Defining and modelling what the asset/process is supposed to be	CAD, engineering models, process simulations, P&IDs, asset models, equipment specs, digital twins, engineering document/data systems

The diagram at the top of the article simplifies an industry model of five to seven stages into six phases and five handoffs. This allows us to follow one seam rather than all of them. Phases one through four belong largely to ET. Its job is to define everything the mine needs to operate efficiently before there is a mine to run. Commissioning is where OT comes alive. Operations must then work with whatever it inherits.

In a recent conversation, a mining executive with more than 35 years of industry experience described a serious mismatch between the tools used by the disciplines at each stage.

Every stage uses different disciplines, tools and data formats. As a result, each handoff is treated as a file to be passed. Something is lost each time that file crosses a seam, often without anyone noticing until a downstream team tries to run the mine using what survived.

Consider the handoff between studies and design. During the studies phase, geologists build a working 3D model of the orebody in specialist software. Points in the model may carry information about ore grade, rock type and the level of confidence in the geological interpretation.

The design team then opens that model in engineering software intended for a different purpose: designing the mine and its processing plant. Making the geological model readable by the design software can require exporting it into formats that cannot preserve all of its contents.

The export flattens the model. Picture a layered graphics file saved as a single flat image. It may still look right, but the separate layers of information underneath are gone. In effect, you keep the geometry but lose the geology.

Notice where this happens. It occurs inside the ET domain, long before operations or the company’s business IT function comes near the asset.

The format, however, is only the symptom. At the same seam, the studies contract closes and a new team from a different discipline takes over. Each contractor delivered what it promised. Nobody was required to carry the model forward.

Continuity is nobody’s deliverable.

The way projects account for pre-production work can reinforce the problem. Before a mine or offshore facility begins operating, expenditure is scrutinised and pressure to minimise project costs is intense. Spending money to preserve information for a later team or phase can therefore be difficult to justify, particularly when continuity is not part of the contracted deliverable.

Follow the seams forward and the small losses compound into the one that matters most. At commissioning, when ET hands the asset to OT, a chasm can appear where the live model should be. The design says one thing, the constructed mine says another, and valuable geological information may have been stripped away several phases earlier.

Offshore projects can encounter the same problem when a floating production, storage and offloading vessel leaves the shipyard for installation and hookup.

None of this requires sabotage or ignorance. Each team can perform its assigned work with real expertise. The loss is created quietly during engineering and felt years later in operations.

That is when an expensive digital twin fails to reconcile with the physical asset. In industry terms, it does not “tie out.” The model intended to help run the mine turns out to be a shell: the shape of what was built without the knowledge once contained inside it. It is a foundation crack that becomes visible only after the wall settles.

If the phrase “digital twin” made you wince, hold onto that reaction. EY has reported that 92 per cent of oil and gas organisations were either investing in digital twins or planning to do so, while only 14 per cent considered their initiatives successful. The gap between those figures helps explain the problem.

Many digital twins are created during commissioning as replicas of plants that already exist. By then, the physical asset may have diverged from the engineering information that originally defined it. Reconstructing a model at that point is not continuity.

Continuity is more than a cleaner handoff or a better file converter. It means preventing the model from dying, being repeatedly recreated or being diluted to fit the tool used in the next phase.

The goal is one model that matures. It begins as a resource model during studies, develops into a design model, becomes an as-built model and eventually supports operations. It is reshaped at each stage, but its accumulated knowledge is not discarded and reconstructed.

“Never recreated” does not mean “never changed.” The value lies in maintaining a living model that can expand, contract and be tested against physical reality as the asset develops. A model that cannot change is only a fossil under glass.

That continuity depends on reliable data exchange and connectivity. If the model is to remain a shared source, every phase must be able to read from it and contribute information back to it. Sending a static file to the next team is not enough.

The information must also use interoperable structures rather than depend entirely on one vendor’s proprietary file type. The Open Mining Format is one attempt to make geological and mining data easier to exchange between different software systems. Where systems still cannot communicate directly, integration is precisely the problem that IT is equipped to solve.

Connectivity must also be designed for the project’s actual environment. A new mine may be remote, temporary construction facilities may move, and terrestrial infrastructure may be limited or absent. The communications architecture may therefore need a combination of terrestrial and satellite services from the earliest project phases. Treating the network as an operational system to be added at the end is already too late.

That points to where the real work is. The answer is not to appoint a new owner for the engineering model. It is for IT to take on a broader role. Instead of entering periodically to establish procurement, finance and office systems, IT must provide the integration and connectivity that keep the model alive across the full life of the mine, potentially for 30 years or more.

That does not give IT ownership of the model or of the engineering decisions contained within it. Those responsibilities remain with the relevant disciplines. IT owns the connective capability that allows the model to pass between them without being flattened, abandoned or rebuilt from fragments.

Who should be paid to preserve continuity across each contractual handoff is a larger question than this piece can settle. But the technical distinction is clear: the disciplines own their work; the connective layer must ensure that their knowledge survives.