By Aliesha Aden
On most mine sites, the biggest delays don’t come from failures. They come from the decisions around them.
Not the obvious ones, like responding to a breakdown, as those are clear enough to demand action and everyone needs to align quickly. It’s the smaller, daily decisions that shape performance over time. The ones made in control rooms, during handovers, or in the space between “something’s changed” and “what do we do about it.”
These decisions rarely feel significant in isolation. But across a site, every day, they quietly determine whether operations stay stable or drift toward disruption.
Decisions that shape throughput
Across most sites, the same handful of decisions come up again and again. They’re rarely documented, often made too quickly, and they don’t always get revisited. But they sit at the centre of how a site actually performs.
Decision 1: Do we act now, or keep it running?
Early signs are rarely definitive. A vibration trend starts to shift. Temperature creeps outside its usual range, and something doesn’t look quite right, but it’s not clearly wrong either.
Act too early and you risk unnecessary intervention, lost production, and wasted effort. Act too late, and you’re dealing with unplanned downtime, safety risks, and escalating costs.
So teams do what makes sense in the moment by weighing experience, current production pressure, and what else is going on across the plant. Often, the decision becomes: let’s keep an eye on it.
Decision 2: How serious is this, really?
Even when something’s flagged, the next question is rarely straightforward:
- How urgent is it
- Is it critical, or just noise
- Does it need action today, this week, or during the next shutdown
The challenge is that the same data can lead to different conclusions depending on who’s looking at it. An experienced reliability engineer might recognise a developing fault immediately, whereas, a less experienced operator might see the same signal and deprioritise it.
Across shifts, teams, or sites, that inconsistency adds up. What gets actioned in one context might be ignored in another.
Decision 3: What actually needs to be done?
Detecting that something has changed is only the first step. The harder part is understanding what’s likely happening and what action will actually resolve it.
Is it a bearing issue, misalignment, lubrication, or something else entirely? Does it require immediate shutdown, planned intervention, or further investigation?
Without clear direction, teams fall back on what they know about:
- Inspect the asset (involves more risks)
- Monitor it further (costing time and resources)
- Or apply a general fix and hope it holds (delaying or adding to the problem)
Sometimes that all works. But it introduces uncertainty, risks, and costs that will land eventually.
Decision 4: Can this one wait until the next shutdown?
This is one of the most common and most difficult calls on site. Every operation is balancing risk against production. Stopping an asset has a cost, but so does letting it run with a building problem.
The decision to defer action is rarely taken lightly. But it is often made without a clear, shared understanding of how quickly the issue is progressing, knowing what the consequences of failure would be, or how confident the team is in the diagnosis. So it becomes a judgement call. And like all judgement calls, it varies.
Decision 5: Are we seeing this everywhere or just here?
Mining operations are rarely confined to a single asset or site. The same equipment, operating under similar conditions, exists across multiple locations. But the way issues are identified, interpreted, and resolved is often localised.
Knowledge sits with individuals or teams. Lessons learned in one place don’t always translate to another. Which means the same problems get solved multiple times, in slightly different ways, with slightly different outcomes.
Why these decisions are harder than they should be
None of these decisions are new. Most sites make them every day. The challenge is how they’re made.
They rely heavily on individual experience, fragmented data, time pressure, and incomplete context. Even with more sensors, dashboards and alerts, the burden of interpretation still sits with people. As systems improve visibility, they often increase the volume of information that needs to be interpreted, not reduce it.
So, while sites know more than ever about what’s happening, translating that into consistent, confident action remains difficult.
What this looks like in practice
Across mining, metals, and heavy industry sites, we consistently see the same pattern play out:
- Thousands of potential issues identified, but only a portion are clearly prioritised
- Teams act on most recommendations, but still carry a backlog of “watch and wait” decisions
- Downtime is avoided not by reacting faster, but by acting earlier and with more confidence.
In many cases, the difference isn’t whether a problem was detected. It’s whether the team had enough clarity to act at the right time, with the right level of urgency.
Across our mining, metals, and heavy industry deployments, more than 6,100 prescriptive actions have been generated, with 96% actioned by teams on site, contributing to over 17,700 hours of downtime avoided.
What changes when decisions become structured
Improving these decisions isn’t about adding more data. It’s about making them clearer, more consistent and easier to act on by:
- Understanding what a change likely represents
- Prioritising based on impact
- And creating a way for teams to learn from outcomes
When that happens, the conversation on site shifts from “somethings changed, what do you think?” to “this is what’s likely happening. This is the risk. This is what we should do next.”
Over time, most operations don’t lose performance in a single moment. They lose it gradually, through small delays, inconsistent calls, and decisions made without full clarity. Even for the most seasoned teams, this result is not because they don’t care, or because they lack capability, but because the decisions that matter most are often the hardest ones to make well. And until those decisions become clearer, faster, and more consistent, throughput will always depend on who’s making the call and when.