This article is part of the Ventilation Automation Pyramid series. Read the full overview here.
Level 2 gave the ventilation system the ability to act on its own but only according to what was supposed to happen. A schedule knows when the blast crew is supposed to fire, when a shift is supposed to start, when a heading is supposed to go quiet. It has no way of knowing when something happens that wasn’t on the calendar.
Level 3 closes that gap. The ventilation system stops asking what time is it and starts asking what just happened.
From the Calendar to the Sensor Network
A schedule is a plan. Underground mines do not always follow the plan.
Equipment fails outside of maintenance windows. Gas levels rise without anyone calling it in. A door or bulkhead gets left open. None of these events care what the schedule says, and none of them wait for an operator to notice and react.
Level 3 introduces event-triggered automation the ability for the ventilation system to detect a discrete event from a sensor or system signal and execute a predefined response automatically, without waiting for a scheduled trigger or a human command.
The system is not just executing a plan. It is now watching for exceptions to the plan and reacting to them directly.
What Counts as an Event
Where Level 2 reacts to time, Level 3 reacts to discrete signals coming from the systems already monitoring the mine:
- Gas or CO sensor exceedance a reading crosses a defined threshold in a specific zone
- Equipment fault or breakdown a fan, damper, or VFD reports a fault state
- Door or bulkhead position change a ventilation door opens or closes outside of expected sequence
- Communication loss a control node or sensor goes offline unexpectedly
- Power loss to a ventilation asset an unplanned trip on a feeder or starter
Each of these is something that can be detected automatically, and each one carries a known, definable response. That combination detectable signal, defined response is what makes an event automatable at this level.
A Concrete Scenario: A CO Sensor Exceedance
Consider a CO sensor in a heading where diesel equipment has been running continuously through the shift.
Without event-triggered automation: The CO sensor crosses its alarm threshold. The alarm appears on a screen in the control room, assuming someone is watching it at that moment. The operator has to notice the alarm, identify which zone it corresponds to, determine which fans and dampers affect that zone, and manually adjust them all while potentially also coordinating with the crew in the area. Minutes pass between the exceedance and the corrective action, and those minutes depend entirely on operator attentiveness.
With Level 3 event-triggered automation: The same exceedance is detected by the control system the moment it crosses threshold. The system automatically increases airflow to the affected heading, confirms the fan and damper response, and notifies the operator that an automated response has been triggered along with the zone and the readings that caused it. The operator’s role shifts from detect and react to verify and confirm a response measured in seconds rather than minutes, and one that happens identically whether the control room is busy or quiet.
What Automatic Response Actually Changes
Response Time Independent of Operator Attention
The single biggest gain at this level is speed. A sensor-to-actuator response loop closes in seconds. A human-mediated response notice, interpret, act, almost never does, no matter how good the operator is.
Coverage for Events That Schedules Cannot Anticipate
Level 2 handles the predictable. Level 3 handles the rest. Equipment faults, gas exceedances, and unplanned door openings are, by definition, not on a calendar. This is the level where the system starts protecting against the things nobody scheduled for.
Reduced Reliance on Constant Operator Vigilance
Operators are still essential at this level, but their role changes. Instead of needing to continuously scan every reading on every panel, they need to respond when the system tells them something already happened and was already addressed. This is a meaningfully different and more sustainable cognitive task over a 12-hour shift.
A Verifiable Safety Response Layer
Every triggered event and every automated response gets logged with a timestamp, the triggering condition, and the resulting action. For gas exceedances in particular, this creates a defensible record that the appropriate ventilation response occurred immediately, independent of whether a person happened to be watching.
Groundwork for Demand-Driven Ventilation
Event-triggered automation is reactive by design: it responds after a threshold is crossed. But the underlying capability, tying live sensor data directly to ventilation actuation, is the same capability that full demand-driven ventilation will need at Level 4. Level 3 proves that the sensor-to-actuator pathway works before asking it to run continuously.
The Limitations That Level 3 Does Not Solve
Event-triggered automation responds after a threshold is crossed. It does not anticipate that the threshold is about to be crossed, and it does not optimize ventilation continuously based on the full picture of what is happening underground.
Event-triggered automation responds to discrete threshold crossings, in both directions. A CO exceedance increases airflow to the affected zone, and when readings return below the acceptable threshold, that recovery is itself an event that returns the system to its previous ventilation state. The system is not stuck in a response posture indefinitely.
What Level 3 cannot do is reason about the space between events. It does not see that CO levels have been trending upward for the past 20 minutes and gradually increase airflow in anticipation. It does not know that a heading is producing more diesel exhaust than usual because three pieces of heavy equipment converged there at the same time. It acts correctly when thresholds are crossed, but it has no awareness of the conditions building toward that crossing.
That gap, continuous awareness of mine conditions and proactive airflow adjustment before thresholds are ever reached, is the work of the level above.
Closing Thoughts
Level 3 is often the point where ventilation automation starts to feel less like a convenience and more like a safety system. A scheduled blast sequence is operationally useful. A gas exceedance triggering an automatic airflow increase in seconds, logged and verified without depending on someone watching a screen at the right moment, is a different category of value entirely.
This is also the level where the conversation in the control room tends to shift. Instead of asking did anyone catch that alarm, the question becomes what did the system already do about it.
Previous in the series: Level 2 – Scheduled and Time-Based Ventilation Automation.
Next in the series: Level 4 – Location-Driven Demand Control.
Working out where your own operation sits on this pyramid, or what the next level would actually take? That’s the kind of evaluation I do. Get in touch. No pitch.
Leave a Reply