Common FDM events explained

Open a flight data readout and the first thing you see is a list of events. To anyone outside the safety office the word sounds like an accusation. It is not one. An event is a measurement: a point in the flight where a recorded parameter moved outside the range your operation decided was worth a second look.

This article goes through the events most programmes watch, grouped by flight phase, and explains what each is and why it earns its place. You will not find threshold values here: the ones that matter are those your operator sets for its own aircraft, routes and procedures.

What is an FDM event?

A flight data monitoring programme reads the digital data your aircraft already record and looks for moments where a flight departed from what was expected. Those moments become events. Two words get used as though they mean the same thing. An exceedance is a value that went past a defined limit, such as a speed above a placard limit or a temperature above an engine maximum. An event is broader: it includes exceedances, but also things entirely permissible and still worth counting, such as a go-around or a terrain warning.

Alongside both sits routine measurement: how a flight was flown, recorded even when nothing was breached. Events tell you what went wrong on a handful of flights; routine measurement tells you how the rest went, and whether the margin between normal operations and the limit is quietly thinning.

How does an event get triggered?

An event definition is a small set of conditions, and most of them need the same ingredients:

• A parameter. The recorded value being watched: airspeed, vertical acceleration, pitch attitude, radio height, an engine temperature.

• A threshold. The value that parameter must pass before anything is recorded.

• A flight phase. The same pitch attitude means one thing on rotation and something else in the cruise, so a definition is tied to a phase.

• A duration. A momentary spike is often noise; a condition held for a period is a finding.

• A severity level. The same event is commonly graded minor, major or critical by how far past the threshold it went.

Get those settings wrong and the programme fails in one of two directions. Thresholds set too high hide the events you needed to see; set too low, they bury an analyst in detections that mean nothing. That balance is one of the risks in our article on the top risks to an FDM programme, and it is why an event set is reviewed periodically rather than written once. In AeroSight, the events that are monitored constitute a dynamic set, continuously evolving to adapt to the company procedures in place.

Which events come from takeoff and climb?

The takeoff is short, close to the ground and irreversible past a certain point, which is why so much of an event set lives here.

• Rejected takeoff at high speed. The takeoff was abandoned above a set speed, where the margin left for stopping on the remaining runway is much smaller. Low-speed rejects are routine; high-speed ones are worth understanding.

• Early or late rotation. The nose was raised before the planned rotation speed, or well after it. Early costs performance; late eats runway.

• High pitch rate or pitch attitude at rotation. The nose was raised faster or higher than normal. This is the tail strike condition on departure, watched closely on types whose geometry makes it a real risk.

• Excessive bank angle. Roll beyond a set angle. It matters most close to the ground, on departure and approach, where there is little height in hand.

What is watched in the climb, cruise and configuration changes?

The cruise tends to produce fewer events than the runway, and they tend to be structural, usually about speed.

• Flap or slat overspeed. The aircraft was faster than the limit speed for the setting selected at the time. These surfaces are not built for high speed, and an overspeed can call for an inspection under the manufacturer's procedures.

• Landing gear overspeed. The gear was extended, retracted or left down above its limit speed, with doors and actuators carrying the load.

• Maximum operating speed exceedance. The aircraft went above VMO or MMO, the maximum operating speed and Mach number for the type. Turbulence and a descent flown too eagerly are the usual contexts.

These map directly onto what a programme is set up to find, which our best practices page describes as the detection of SOP violations and deviations from structural aircraft limits.

Which events come from approach and landing?

Approach and landing account for a large share of a typical programme's events, and are where safety managers tend to spend their time. Many are connected: a fast, high approach becomes a late configuration, then an unstable approach, then a long landing.

• High rate of descent at low height. The aircraft was descending faster than expected close to the ground. Height and rate together make this significant, not either alone.

• Glideslope or vertical path deviation. The approach was flown above or below the vertical path, changing where the aircraft arrives and with how much energy.

• Speed high or low on final. Approach speed outside the target band. Fast is the more common finding, and the one that carries into the landing.

• Late landing configuration. Gear or landing flap was selected later than the procedure calls for, leaving less time to settle before the approach must be stable.

• Unstable approach. The operator's own stabilized approach criteria were not met by the height the procedures define. It is not one measurement but a combination of speed, configuration, rate of descent, path and thrust, which often makes it one of the more informative events in the set.

• Long landing or long flare. Touchdown beyond the planned touchdown zone, or a float before touching down. Both shorten the runway left for stopping.

• Hard landing. Vertical acceleration at touchdown above a set value. This one has an immediate operational consequence: the manufacturer's procedures may call for an inspection before the next flight, and the flight data is what tells maintenance what happened. Whether an inspection is required always rests with those procedures and your own engineering organization.

• Bounced landing. The aircraft became airborne again after first touchdown. The second touchdown is the one to look at.

• High pitch at touchdown. The tail strike case on landing, most relevant after a long flare or bounce.

• Go-around. A go-around is a normal procedure, not a fault. Programmes record it to see how often it happens, where, and how it was flown. The go-around that was not flown when it should have been is the more serious finding.

What about the warnings the aircraft raises itself?

Some events are not found in the raw parameters at all. The aircraft's protective systems recorded a warning, and the programme takes it as it is.

• Terrain warnings. A (E)GPWS or TAWS alert was recorded, graded by the kind of alert, since a caution and a pull-up warning are not the same thing.

• Traffic collision avoidance resolution advisory. The collision avoidance system commanded a climb or a descent. The event is recorded so the response can be reviewed, because following the advisory promptly and fully is what matters.

• Stall warning or stick shaker. The approach-to-stall warning activated: rare, and investigated when it appears.

• Windshear warning. The aircraft's windshear system produced an alert, either ahead of conditions or in reaction to them.

These say as much about the environment as about the crew: a cluster of terrain alerts at one airport usually points at an approach procedure rather than at the pilots flying it.

Which engine events does a programme watch?

Engine events sit slightly apart, being as much about the health of the aircraft as about how it was flown: exhaust gas temperature above the limit, shaft speed exceedances, overtorque on turboprop types, exceedances during start.

The same data feeds trend monitoring, where the interest is the slow drift of a parameter over months. AeroSight detects engine related events and offers automatic data export for engine condition monitoring. As with any finding that touches the aircraft itself, what to do about an engine exceedance is set by the manufacturer's procedures, not by the programme.

Why is a single event not a verdict?

An event is a hypothesis, not a conclusion, and a fair share do not survive validation.

The usual causes are dull: a parameter recorded badly or dropped out, a sensor misbehaving, a threshold set without allowing for how a type is flown, or an event that was real but entirely appropriate, such as an energetic descent flown to avoid weather. This is why AeroSight includes flight acceptance, where a reviewer can analyze detected events with their correlated parameters and reject inappropriate ones before the readout becomes part of the statistics.

The rate of undesired detections is therefore worth watching in its own right, a point covered in our article on getting the most of your FDM data. A programme producing events nobody acts on is not a careful programme; it is one whose thresholds need attention. Events are also only useful in aggregate: one high rate of descent tells you about a flight, a run of them at the same airport tells you about an approach.

How do events feed the rest of the programme?

Detection is the beginning of the work, not the end. Validated events feed three things: the safety management system, where trends become hazards on a risk register and corrective actions are tracked to see whether they made any difference; crew feedback, which works only where the data is used non-punitively, an expectation written into the regulatory framework covered on our legislation page; and training, where recurring events point to the procedures worth revisiting rather than to the individuals involved.

The AeroSight Pilot Performance module presents trend data of this kind, showing trends for minor, major and critical events along with the most frequent events and the airports where they are triggered, as described in our article on the benefits of AeroSight Pilot Performance.

Where does this leave your event set?

An event set is not a fixed list that arrives with the software. It describes what your operation considers worth knowing about, and should change as your fleet, routes and procedures do. The events above are the common ground most operators start from, but which you watch, how tightly and with what severity grading is your safety department's decision.

AeroSight FDM provides a fast and automatic readout and exceedance detection process, periodical safety reports and advanced search in exceedance events. To talk through how your event set is defined today, get in touch.