The System
That Needs Watching

There is a version of automated flight that looks like good airmanship from the outside but is not. The crew selects a mode, the aircraft begins to comply, and attention moves on. The automation appears to be doing its job. The crew appear to be monitoring. What is actually happening is that an assumption has been made — that the mode did what was asked — and the monitoring that would confirm or challenge that assumption has not occurred.

This is the gap that effectively monitors automation closes. It is not about knowing which modes to select, nor about the discipline of briefing automation strategy in advance. It is about the continuous, active discipline of watching the system after every input and throughout every phase — because automation is not fire and forget, and the crew that treats it as such has delegated responsibility to a system that does not carry it.

The First Question

Every automation input begins with the same question, and effective crews ask it explicitly every time: did it engage? It seems almost too simple to state. You pressed the button. The mode should have captured. Of course it engaged.

Except that it doesn't always. Mode capture requires conditions to be met — the aircraft within the capture envelope, the input timed correctly, the system in a state ready to receive it. When those conditions are not met, the mode does not capture. The aircraft continues in its previous state. And if the crew have already moved on — already assumed engagement and redirected their attention — the non-engagement goes undetected until its consequences become visible. By then the situation has developed, often in a direction that requires more workload to resolve than the original monitoring would have cost.

The discipline of confirming engagement is therefore the foundation of everything else. It takes seconds. It requires looking at the flight mode annunciator and confirming that what is displayed matches what was selected. That single habit — so simple that it barely merits discussion in training — is the first gate in effective automation monitoring. Everything downstream depends on it.

Action, Mode, Response

The action-mode-response sequence is the disciplined framework for every automation input. Action — the selection made. Mode — confirmation that the annunciator reflects the intended mode. Response — verification that the aircraft is actually doing what the mode should produce. All three. Every time.

The response step is the one most often skipped. A crew that checks the mode annunciator and sees the correct mode displayed has confirmed that the system believes it is in that mode. It has not confirmed that the aircraft is responding as expected. A speed mode that captures but does not converge on the target. An altitude capture that arrests the rate of climb short of the cleared level. A lateral mode that tracks a course other than the one intended. In each case the annunciator may show the correct mode while the aircraft's actual behaviour tells a different story. Response monitoring is what bridges that gap — the cross-check between what the system says it is doing and what the aircraft is actually doing.

This is where automation monitoring connects directly to Aviate, Navigate, Communicate. Aviate is not suspended because the automation is engaged. It is continuously verified through the monitoring that effective automation use requires. The crew that monitors the automation is aviating — actively, continuously, with full awareness of the aircraft's state. The crew that has delegated aviating to the automation and redirected their attention elsewhere has not stopped aviating. They have simply stopped noticing whether it is happening correctly.

Mode Degradation — The Hardest Part

If confirming engagement is the foundation of automation monitoring, understanding and detecting mode degradation is its hardest element. Modern flight management systems are designed to protect the flight envelope and maintain safe operation — and part of that design is the automatic reversion to a lesser mode when the current mode can no longer perform its function safely.

The system does not ask permission. It does not announce the change in a way that demands immediate attention. It transitions — and the crew, if they are not monitoring, may not know it has happened until the aircraft's behaviour reveals it. An autothrust mode that reverts from speed protection to a thrust limit mode during an energy exceedance. A vertical mode that reverts from VNAV PATH to VNAV SPD when the path can no longer be flown at a safe speed. A lateral mode that reverts to heading select when the FMS loses the ability to compute the programmed track.

Each of these reversions has implications. The aircraft is now doing something different from what the crew last confirmed. The workload picture has changed. The flight path may be diverging from the intended profile. And the crew, if they are not watching, are flying an aircraft they believe is in one state when it is actually in another — which is precisely the SA breakdown that monitoring exists to prevent.

Monitoring to Enable Intervention

The purpose of monitoring automation is not observation for its own sake. It is to maintain the situational awareness that makes timely intervention possible. A crew that monitors effectively is a crew that is always in a position to intervene — because they know what the automation is doing, they understand what it should be doing, and they can detect the divergence between those two states early enough to act before it becomes consequential.

Intervention is the downstream product of monitoring. The crew that does not monitor has surrendered the ability to intervene at the optimal moment. They will intervene eventually — when the divergence becomes large enough to force their attention — but by then the situation has developed further than it needed to, and the intervention is reactive rather than considered. Reactive interventions under workload carry their own risks. The crew that monitors early intervenes early, when the correction is small, the options are wide, and the workload of correction is low.

There is also a category of situation where the correct intervention is no intervention — where the automation, given time, will resolve the situation better than the crew acting manually. Recognising that category requires the same monitoring discipline. The crew that does not understand what the automation is doing cannot make a reliable judgement about whether to trust it. The crew that monitors continuously can make that judgement with confidence — and the confidence to let the automation work when it should work is itself a product of effective monitoring.

Monitoring as Part of the Scan

Effective automation monitoring does not exist in isolation from the broader instrument scan — it is integrated into it. The flight mode annunciator is part of the scan. The autoflight status is part of the scan. The cross-check between what the automation is commanding and what the aircraft is doing is part of the scan. A crew with a disciplined scan that excludes automation status is not scanning effectively — it is scanning most of the picture and assuming the rest.

The integration of automation monitoring into the scan is particularly important during high workload phases, when the temptation to narrow attention is greatest. It is precisely during those phases — the demanding approach, the departure into complex airspace, the non-normal situation — that automation monitoring matters most. The crew that maintains their automation scan discipline during high workload is the crew that will catch the mode reversion at the moment it occurs rather than three minutes later when the aircraft's behaviour can no longer be ignored.