What a FRAT scores,
and where it stops.

See the highlighted block above the contents list. The rest of this article defines the tool, walks through the four moving parts that make one work, shows where FRATs are close to universal, and explains the two limits that decide whether a FRAT programme adds real safety value or quietly becomes a form nobody trusts.

A flight risk assessment tool, almost always shortened to FRAT, is a structured pre-flight checklist that turns the specific risks of one flight into a single number. Instead of asking a crew to hold a dozen separate concerns in their head at once (a tired captain, a marginal ceiling, a short contaminated runway, a night departure into rising terrain), the FRAT collects each of those factors, assigns it a weight, and adds the weights up. The output is one score, and that score maps to a band: proceed, proceed with a second look, or stop and get approval.

The point is not the arithmetic. The point is that the FRAT forces the crew to look at the same risk factors, in the same order, every single time, before every flight, and it produces a record that later numbers can be compared against. A FRAT is a worked example of Safety Risk Management, the second of the four components of an SMS, applied to the smallest possible unit of exposure: the individual flight. It is a close cousin of the aviation risk matrix. Where a risk matrix scores a hazard by severity and likelihood, a FRAT scores a whole flight by summing many weighted factors into one actionable total.

Both the FAA and the main industry associations publish reference FRAT models that operators can adopt and adapt. Most flight departments start from one of those, then tune the factors and weights to their own fleet, routes and crew mix. There is no single official FRAT; there is a well-understood shape, and a lot of room to calibrate it well or badly.

Every FRAT is built from four moving parts, and understanding them is the difference between running a tool and filling in a form.

• Factors. The questions themselves. A typical FRAT groups them into a handful of families: the crew (currency, recent experience on type, duty and rest, fatigue, significant recent stressors), the weather (ceiling, visibility, crosswind and gusts, icing, thunderstorm activity, day or night), the airport and route (runway length and surface condition, terrain, available approach type, lighting, how familiar the crew is with the field), and the aircraft and operation (any deferred equipment, the nature of the mission, time pressure from the schedule or the customer).
• Weights. Each answer carries points. A benign answer scores zero or close to it; a risk-adding answer scores more, and the more dangerous the factor, the heavier its weight. Better FRATs also account for factors that compound: a night departure is one thing, but a night departure into terrain in deteriorating weather is not the simple sum of three small risks, it is a much larger one, and a well-designed tool reflects that.
• Thresholds.The total falls into a band, and the bands are the whole reason the tool exists. A low total sits in the go band and the flight proceeds. A middle total sits in an elevated band that asks for something concrete: a documented mitigation, a conversation with a second pilot, a delay for better weather. A high total sits in a band that cannot dispatch on the pilot's authority alone.
• Escalation. This is the part that separates a real FRAT from paperwork. Above the top threshold the decision leaves the cockpit. The flight cannot go until a named person, usually the chief pilot, the director of operations or the accountable manager, has seen the score, understood the factors driving it, and either approved the flight with specific mitigations or cancelled it.

Escalation is the load-bearing part. A checklist that lets the same pilot who is under pressure to fly clear their own high number is a formality, not a control. The value is created the moment an elevated flight is routed to someone whose job is to say no.

FRATs appear across all of aviation, but they are close to universal in three worlds. On-demand and charter operators (in the United States, Part 135) fly to unfamiliar airports, at all hours, on short notice, often with a crew that has not flown that exact trip before; the FRAT is how they hold a consistent bar under commercial pressure. For those operators it is usually the most visible piece of their SMS. Business aviation runs on the same logic: a flight department serving a principal who wants to leave now needs a defensible, repeatable way to say “this specific trip is in the elevated band, and here is what we are doing about it.” And helicopter operations, above all air ambulance and offshore work, live and die by pre-flight risk assessment, because their exposure to weather, terrain and confined-area operations is so high.

What these three have in common is variability. An airline flying the same city pairs with the same fleet has designed most of its risk out through standardisation. An operator whose next flight might be anywhere, at any time, in whatever the weather is doing, cannot standardise the trip, so it standardises the assessment of the trip instead. That is the FRAT, and it is why the tool is densest exactly where flying is least routine.

The regulatory backdrop reinforces this. The FAA's SMS final rule, published in 2024, extended formal safety management system requirements to all Part 135 certificate holders, to commercial air tour operators under 91.147, and to certain Part 21 certificate holders, with full implementation and a declaration of compliance due no later than 28 May 2027. For thousands of these operators the FRAT is the most frequent, most daily piece of Safety Risk Management they run. That makes it the natural place to start building an SMS, and, as the next section argues, the wrong place to stop.

A FRAT is a decision gate for one flight. It answers one question well, “should this specific flight go, and under what conditions,” and then it is done. But a safety management system is four components working together, and a FRAT lives entirely inside the second one, Safety Risk Management, and only touches a corner of it.

Treating the FRAT as the SMS is one of the most common mistakes small operators make on the road to a 2027 deadline. The FRAT does not investigate the incident that happened anyway. It does not track a corrective action to closure. It does not tell you whether your reporting culture is healthy, whether last quarter's hazards were actually acted on, or whether the mitigations you are relying on still work. It is a leading control at the single moment of dispatch, and nothing more. A declaration of compliance rests on the whole management system being in place, not on one good checklist.

The second limit is human, and it is the one that quietly kills FRAT programmes: thresholds people do not respect. If the bands are set so low that half of all normal flights land in the elevated band, crews learn that “elevated” means nothing and start pencil-whipping the form to clear it. If they are set so high that nothing ever escalates, the tool is decorative. A FRAT that people respect has thresholds calibrated to the operation's real risk, an escalation that is fast and blame-free (a two-minute call, not a career risk), and a visible track record of management honouring the score by actually cancelling flights when the number says cancel. The score only carries authority if the organisation has shown that it will.

Here is where most FRAT programmes leave their best value on the table. The tool is filled in before every flight and then, in a lot of operations, the paper goes in a drawer or the number disappears into a spreadsheet nobody reads. Each score was a decision aid for one flight; collectively, the scores are one of the richest leading-indicator data streams a small operator owns, and almost nobody trends them.

Consider what the accumulated scores can tell you. How often are we dispatching in the elevated band, and is that number rising? Which routes, which airports, which times of day drive the high scores? Are the same two or three factors (fatigue, one particular airport, a seasonal weather pattern) behind most of our escalations? Is a specific crew consistently flying at the top of the band? None of these are questions about one flight. They are questions about the system, and they are precisely the questions Safety Assurance exists to answer.

That only works if the FRAT data lives in the same place as the rest of your safety picture. When a pre-flight score is captured as structured data in the same operational graph as your occurrences, your hazards and your indicators, the score stops being a throwaway number. A recurring pattern in the factors becomes a safety performance indicator you can monitor with real statistical process control (Western Electric rules that flag a true shift in the mean, not a colored arrow that turns red on one bad week). An escalation that keeps recurring on one route becomes a hazard with a bow-tie and a tracked barrier. A corrective action raised off that hazard is held open until it is proven effective. And all of it rolls up into a computed Safety Risk Profile rather than a folder of forms. That is the difference between a connected safety loop and a stack of disconnected tools.

eAviora is built for exactly that single-graph picture. It ships a pre-built aviation risk model (101 bow-tie models, 804 named barriers, 210 scenarios and a 610-indicator SPI library) so a small operator is not starting from a blank library, and it runs occurrences, hazards, corrective actions, audits, indicators and the computed Safety Risk Profile in one system rather than six. The pre-flight FRAT is where a good safety habit starts; a system of record with real analytics is where that habit compounds into oversight.