Reference

Airline crew planning glossary: 45 terms, defined and cross-referenced

SkAI Tech · Published on 10 August 2026

This glossary defines the 45 terms that structure airline crew planning: planning and optimisation, regulation and flight time, fatigue, aircraft leasing, operating segments, and the job itself. Each entry gives a quotable definition, its French equivalent and its cross-references.

Entries are grouped by theme and individually addressable: every term carries its own anchor link, ready to be reused in an internal procedure, a training pack or a requirements document. Regulatory values are always attributed to the regime that carries them — ORO.FTL in Europe, 14 CFR in the United States — because none of them transfers unchanged from one to the other.

The 45 terms, by theme

Planning and optimisation

Crew scheduling
Crew rostering
Crew pairing
Flight assignment
Roster optimisation
Crew bidding
Roster stability
Roster pattern
Trips and rotations
Qualification constraint

Regulation and flight time

Flight Time Limitations (FTL)
Flight Duty Period (FDP)
ORO.FTL
Acclimatisation
Rest period
Standby duty
Home base
Air Operator Certificate (AOC)
Part 117
Part 135
Part 91
Type rating
Recency and currency
Block hours
Duty time
Augmented crew
Reserve crew

Fatigue

Window of Circadian Low (WOCL)
Fatigue Risk Management System (FRMS)
Biomathematical fatigue model
Crew fatigue

Aircraft leasing

ACMI
Wet lease
Dry lease
Damp lease

Segments and operations

Deadheading
Positioning flight
Medevac
HEMS
Disruption management
Crew recovery

The job and the organisation

Operations Control Centre (OCC)
Crew performance KPIs
Pilot shortage
Flight management system (FMS)

Crew scheduling (planification d'équipage)

Crew scheduling is the whole activity of assigning qualified crew members to an operator's flights while respecting flight, duty and rest limits, individual qualifications and the applicable labour agreements. It spans horizons from long-range manpower planning down to same-day reassignment on the operations desk.

At large carriers the activity splits into three distinct problems solved in sequence. Crew pairing builds flights into feasible trips that start and end at a home base. Rostering assigns those trips to named individuals alongside their leave, training and personal constraints. Day of operations then absorbs whatever reality delivers: delays, technical stops, sick calls.

At charter, business aviation, air ambulance and ACMI operators, those three phases collapse into a single job held by a handful of people. There is no fixed programme to start from: the schedule is built and rebuilt as requests arrive, sometimes hours before departure. The difficulty is not the number of flights but the number of rules that must hold simultaneously, and the rate at which the inputs change.

That combinatorial pressure is why crew scheduling defeats spreadsheets above a very modest size. Every assignment moves rolling counters for every crew member it touches, and every disruption invalidates part of the checking already done. When manual planning fails, carelessness is rarely the reason; the arithmetic has simply outgrown what any one person can hold in view at once.

Crew rostering (rostering, établissement des rosters)

Crew rostering is the stage of crew planning that produces each named crew member's individual schedule for a period, usually a month, combining trips, days off, standby duties, training, checks and leave into a single published, compliant plan that the crew member can plan their life around.

The roster is the finished product of the planning function. It is what the crew member reads, what payroll and HR draw on, and what the operator is judged by internally. Its quality is measured on three axes that rarely align: regulatory compliance, cost to the operator, and perceived fairness across the crew community.

No roster starts from a blank page. It starts from a skeleton that is already fixed: recurrent recovery rest, approved leave, training and check expiry dates, medical unavailability. Flights and standby duties are then fitted into whatever space is left. This is why late roster publication is expensive twice over: it creates industrial friction, and it shortens the window in which the operator can sell or commit capacity with confidence.

Rostering is SkAI Tech's core problem. The platform generates compliant rosters in an average of 12 minutes, running as an add-on to the operations system the operator already uses rather than as a replacement for it.

Crew pairing (appariement)

Crew pairing is the stage that assembles individual flights into feasible trips (sequences of flights, ground time, layovers and rest that begin and end at a home base) before any named crew member is involved. It is where most of the crew cost of a flight programme is decided.

Pairing reasons about anonymous positions: a captain type-rated on the fleet, a first officer, cabin crew where applicable. The question is purely structural: which flights can legally follow one another inside an FDP, with which layovers, which deadhead legs, and how the crew gets back to base. A poorly built pairing (an avoidable layover, a wasted duty day, a deadhead that exists only to fix an earlier decision) is paid for every time the pattern repeats.

Scheduled airlines keep pairing and rostering as separate optimisation problems, largely for model-size reasons inherited from decades of operations research practice. In on-demand segments the separation is far less useful: flights arrive too late and change too often for a frozen set of pairings to retain any value. Modern approaches handle pairing and assignment in one constraint model and simply re-solve as the programme moves, which is only practical when generation takes minutes rather than days.

Flight assignment (affectation vol-équipage)

Flight assignment is the act of naming the crew members who will operate a specific flight, after checking their type rating, recency, rolling flight and duty counters, preceding rest and contractual availability at the exact time of the planned departure. It is the atomic operation of crew planning.

Every roster is, in the end, a set of mutually consistent flight assignments. Each one has to be valid in isolation (is this person qualified, rested, available?) and compatible with everything before and after it in their individual schedule.

That dual test is what makes manual verification so expensive. An assignment that is perfectly legal today can quietly make a crew member unavailable ten days from now, because it pushes a rolling 28-day counter past its limit on that future date. A planner working in a spreadsheet typically discovers this at the moment it bites, which is far too late to fix without breaking several other things.

A constraint-based engine treats each assignment as a variable coupled to all the others. Rather than auditing a finished plan after the fact, it imposes validity while the plan is being built. That is the difference between finding out that a roster is illegal and never being offered an illegal roster in the first place.

Roster optimisation (optimisation des plannings)

Roster optimisation means producing, among all the legally valid crew schedules available, the one that performs best against a chosen objective (cost, unproductive hours, fatigue exposure, fairness) instead of stopping at the first workable answer a human happens to find. It requires the constraints and the objective to be stated explicitly.

The gap between "compliant" and "optimised" is routinely underestimated. A compliant roster satisfies the rules. An optimised roster satisfies the rules and is the best available plan against something the operator can measure. The distance between the two shows up as days of planning effort, reserve hours burnt for no reason, avoidable hotel nights, and last-minute changes that erode crew trust.

Which objective to optimise is a management decision, not a technical one. Minimising direct cost, maximising the stability of the published roster, levelling night duties across the crew community, or holding slack back to absorb disruption all produce visibly different rosters. A credible tool makes that trade-off explicit and adjustable rather than burying it inside an opaque heuristic.

At SkAI Tech, optimisation covers initial production (an average of 12 minutes to generate a roster) as well as robustness: operators running the platform see 20 % fewer schedule deviations.

Crew bidding (bidding, choix de vacations)

Crew bidding is the process through which crew members express preferences about the coming period (trips they want, days off they need, flights they would rather avoid) before the roster is published. The operator then resolves competing requests using published rules, usually seniority or a points system.

Two families dominate. Bidline systems publish pre-built lines of work that crew members rank in order of preference. Preferential bidding systems (PBS) collect granular preferences, such as "no night duty after the 15th" or "long trips rather than short ones", and let an engine build individual rosters that satisfy as many of them as possible, processed in seniority order.

Outside the large scheduled carriers, formal bidding is the exception rather than the rule. Preferences travel by email, by message, or in the corridor, and the planner holds them in their head. That informality has one genuine virtue, responsiveness, and two structural weaknesses: it leaves no auditable record of how fairly requests were treated, and it makes the whole function dependent on one individual's memory and goodwill.

Turning preferences into weighted, traceable soft constraints answers the question that poisons crew relations more than any other: why them and not me.

Roster stability (stabilité du roster)

Roster stability measures how far the schedule crew members actually work diverges from the schedule they were published. A stable roster is one in which few assignments change after publication; instability is paid for in fatigue, disrupted personal lives, last-minute costs and eroded trust in the planning function.

It is an under-used metric precisely because no regulation asks for it. An operator can be perfectly compliant and still reissue its roster three times a week. Crew members, however, judge the planning department almost entirely on this criterion: a roster you cannot believe is a roster you cannot arrange childcare, a medical appointment or a weekend around.

Refusing to change plans does not make a roster stable. What does is building plans that can absorb change: leaving slack on the most exposed trips, positioning reserve where the knock-on risk is highest, and avoiding brittle sequences in which one delay breaks the next day's legality.

This is exactly what the -20 % schedule deviation figure observed at operators using SkAI Tech measures: fewer post-publication changes for the same flying programme, which is a crew retention argument as much as an operational one.

Roster pattern (schéma de rotation)

A roster pattern is a repeating schedule structure, such as five days on followed by four days off, or one week on standby followed by one week free, used as a skeleton for building rosters. It buys predictability and demonstrable fairness at the cost of responsiveness to an uneven flying programme.

Fixed patterns are standard wherever demand is steady or coverage must be permanent: air ambulance bases, offshore operations, long-term ACMI contracts, some corporate flight departments. Their appeal is real on both sides. Crew members know their free days months ahead, and the planner works inside a frame rather than from scratch.

The limit appears as soon as demand becomes irregular. A rigid pattern applied to a variable programme produces under-crewing on peak days and paid non-flying on quiet ones, frequently in the same month. Many operators therefore adopt a hybrid: a guaranteed pattern for days off and protected nights, with optimised assignment inside the duty days.

Expressed as a constraint in an optimisation engine, a roster pattern stops being a straitjacket and becomes one rule among many, respected without giving up the ability to follow the real programme.

Trips and rotations (rotations)

A trip, also called a rotation or a pairing, is a sequence of duties operated by one crew that begins and ends at a home base, covering one or several days including layovers and rest. Trips are the building blocks from which individual rosters are assembled.

A single-day trip returning to base the same evening is a turnaround. A multi-day trip means nights away from base, with their direct costs (hotel, ground transport, per diem) and their own rest rules, which are less demanding away from base than at home base.

Two numbers reveal the quality of a set of trips. The ratio of flying time to paid duty time shows how much unproductive time the structure carries; North American operators often express a related idea as pay-to-credit. The count of nights away from base drives both budget and accumulated fatigue, and it is the figure crew committees look at first.

In charter and business aviation the concept still applies but loses its repetitive character. Each trip is a one-off, shaped around a specific customer request, often with empty legs at one or both ends. That is precisely why optimisation pays so well in these segments: there is no reusable good pairing to carry over from one month to the next.

Qualification constraint (contrainte de qualification)

A qualification constraint is any rule that restricts which flights a crew member may operate based on their credentials: type rating, recency, special approvals such as category III approaches, ETOPS or difficult aerodromes, medical validity, and pairing restrictions between inexperienced crew members.

This is the most commonly underestimated constraint family in crew planning projects. It is structurally different from FTL: flight time limits are counters, qualifications are set membership. A crew member either holds a qualification or does not; holds it until a date; and can lose it by not flying often enough.

Layered on top are the operator's own pairing rules, usually written into the operations manual: no two crew members under line training together, an experienced commander required into certain aerodromes, restrictions applying for a period after a type rating is issued. These rules shrink the space of valid assignments far more than most operators expect, particularly on small bases where the eligible pool is already thin.

An optimisation engine handles them naturally, provided the qualification records in the source system are accurate. This is a standing integration risk: the quality of the roster produced can never exceed the quality of the qualification data feeding it.

Flight Time Limitations (FTL)

Flight Time Limitations are the rules that cap how much a crew member may fly and work, and how much rest they must be given. They set maximum flight times, maximum duty times, maximum flight duty periods and minimum rest periods, and they are the structural constraint on every crew schedule.

In Europe, the FTL applicable to commercial air transport by aeroplane sit in Subpart FTL of Part-ORO of Regulation (EU) No 965/2012, supplemented by certification specifications. In the United States, the equivalent body of rules is split between 14 CFR Part 117 for Part 121 passenger operations and the separate provisions of Part 135. The two systems differ in structure as much as in numbers, which is why compliance logic cannot simply be ported from one to the other.

Nearly every regime combines three types of limit: cumulative caps over rolling windows, daily caps on the flight duty period that depend on report time and sector count, and minimum rest before each duty.

Two warnings matter for anyone building rosters. First, FTL is a regulatory floor, not a target: collective agreements and company rules almost always add constraints that the roster must satisfy at the same time. Second, a compliant roster is not automatically a safe one: FTL bounds fatigue, it does not model it. That gap is what fatigue risk management exists to close.

Flight Duty Period (FDP)

A flight duty period runs from the moment a crew member reports for duty until the aircraft finally comes to rest after the last sector. It therefore includes briefing, turnarounds and ground waiting time, not just block hours, and it is capped daily according to report time and the number of sectors flown.

Under ORO.FTL the basic maximum daily FDP is read from a two-input table: report time expressed in the time zone the crew member is acclimatised to, and sector count. It reaches 13 hours for a favourable daytime start, steps down to 11 hours for the most penalising night reports, and loses 30 minutes per sector from the third sector onwards, with a floor of 9 hours.

Three mechanisms allow that table to be exceeded, and conflating them is a common and consequential mistake. A planned extension adds up to one hour, at most twice in any seven consecutive days, with strengthened rest around it. In-flight rest with an augmented crew allows the FDP to reach up to 17 hours, depending on how many additional pilots are carried and on the class of rest facility. Commander's discretion allows up to two further hours (three with an augmented crew) for circumstances arising at or after report time, a real-time safety valve that is reported to the authority, never a planning assumption.

ORO.FTL (Part-ORO Subpart FTL)

ORO.FTL is the Flight Time Limitations subpart of Part-ORO in the EU Air Operations Regulation. Applicable since 18 February 2016, it sets the flight, duty and rest limits for crew members in EU commercial air transport by aeroplane, and it replaced the older EU-OPS Subpart Q.

It stems from Regulation (EU) No 965/2012 as amended by Regulation (EU) No 83/2014. The provisions planners quote most often are ORO.FTL.210, which carries the cumulative flight and duty caps, and ORO.FTL.235, which sets minimum rest.

Three scope points matter in practice. Air taxi, emergency medical service and single-pilot commercial operations by aeroplane are carved out of ORO.FTL and remain under Subpart Q of the former EU-OPS and under national rules until EASA adopts specific requirements, so a business jet operator may sit under a materially different regime from a scheduled carrier flying out of the same airport. Each operator implements ORO.FTL through its own flight time specification scheme, approved by its competent authority, which means two fully compliant operators can run different internal rules. And ORO.FTL does not displace employment law: the roster has to satisfy both layers simultaneously.

For operators holding several AOCs across several states, this adds up to several rule sets to plan against in parallel: a configuration problem as much as a compliance one.

Acclimatisation (acclimatation)

Acclimatisation describes the state of a crew member whose body clock is aligned with the local time of the zone they are operating in. While acclimatised, FDP tables are read against that local time; after enough time-zone crossings the crew member counts as unacclimatised and different, generally stricter, values apply.

The determination is mechanical rather than subjective. The rules define time-difference thresholds and elapsed-time conditions that decide whether a crew member is still acclimatised to their departure zone, has become acclimatised to the destination zone, or is in an unknown state. Each state points to a different reading of the maximum FDP table.

For planners this is one of the least intuitive constraints, because it depends not on the duty being assigned but on the complete history of the preceding days. Two crew members standing in the same airport at the same hour can hold different maximum FDPs simply because they arrived from different places by different routes. No spreadsheet tracks that state reliably across a whole crew population.

Acclimatisation only bites on operations crossing several time zones. When it does, though, it affects every assignment, including return sectors and positioning legs, and it is a frequent source of late compliance surprises on ad hoc long-haul charters.

Rest period (temps de repos)

A rest period is a continuous, uninterrupted span during which a crew member is free of all duties before starting a new flight duty period. Under ORO.FTL it is at least 12 hours at home base and at least 10 hours away from base, or as long as the preceding duty if that was longer.

The "at least as long as the preceding duty" clause is the one that most often catches manual planning out. A long duty day mechanically forces a long rest, which pushes the next report time later, which can invalidate a trip built earlier in the month. Away from base, rest must additionally provide a genuine eight-hour sleep opportunity once travel and physiological needs are accounted for, which quietly disqualifies some hotel and transfer arrangements that look compliant on paper.

On top of pre-duty rest sits recurrent extended recovery rest: at least 36 consecutive hours including two local nights, with never more than 168 hours between the end of one and the start of the next. In practice this rolling weekend is the skeleton of any monthly roster. It is placed before the first flight, not fitted around the flying once the programme is built.

Standby duty (réserve, astreinte)

Standby duty is a period during which a crew member must remain contactable and available for a duty at short notice, with no flight yet allocated. Depending on its form (airport, home or hotel standby) it is credited differently towards duty time and can reduce the maximum FDP of any duty subsequently assigned.

The fundamental distinction separates airport standby, which counts in full as duty time, from other forms of standby, which are subject to a maximum duration and to a threshold beyond which standby time eats into the maximum FDP of the assigned duty. It is among the most frequently misapplied parts of the regulation, largely because standby that never turns into a flight leaves no visible operational trace to audit.

The topic is central in charter, business aviation and air ambulance operations, where standby is the primary mode of working rather than a marginal safety net. At an air ambulance operator the roster places coverage first and lets missions land inside it; the cost of availability is the dominant cost of the business.

Placement decides whether standby earns its keep. In the right place it absorbs disruption without touching the published roster; in the wrong place it is paid in full while covering none of the slots that were actually at risk.

Home base (base d'affectation)

Home base is the location assigned by the operator to a crew member, where they normally start and end their duty periods and where the operator is not responsible for their accommodation. It determines which rest values apply, and in most jurisdictions it drives tax residence and social security affiliation.

It is a regulatory designation, not a home address. A crew member may live hundreds of kilometres from their home base; commuting is their own time and their own cost, which is why base assignment and base closures are among the most contested industrial topics in the industry. For the operator, base structure drives very concrete costs: longer minimum rest at home base, hotels and per diem away from it, and positioning flights to get people where they need to be.

Opening a secondary base is therefore a structural and rarely reversible decision. It shortens trips and cuts layover costs, but it fragments the available crew pool: the same fleet spread over three bases offers materially less assignment flexibility than one concentrated pool.

Simulating what a base does to roster cost and robustness before committing to it is one of the highest-return uses of an optimisation engine, and one of the few that pays for itself before the software is even in daily use.

Air Operator Certificate (AOC)

An Air Operator Certificate is the approval issued by a civil aviation authority that permits an operator to conduct commercial air transport. It certifies that the organisation, the personnel, the procedures and the resources needed to operate safely are in place for the specific aircraft and operations it lists.

An AOC never stands alone: it comes with operations specifications setting out precisely which aircraft types, geographic areas, types of operation and special approvals are held. That envelope defines what an operator may legally do, and therefore what its rosters may contain.

For crew planning, the AOC matters in two distinct ways. It determines the applicable FTL regime, including the individual flight time specification scheme approved by the authority. And in leasing arrangements it identifies who carries operational responsibility: under a wet lease or ACMI contract, flights are operated under the lessor's AOC and the lessor's rules, even when they carry the lessee's flight number and livery.

Groups holding several AOCs in several states (a common structure in European ACMI and regional aviation) run several parallel compliance environments. Planning tools that assume a single rule set are quickly outgrown in that configuration.

See also: ORO.FTL, ACMI, Wet lease, Part 135

Part 117 (14 CFR Part 117)

14 CFR Part 117 is the United States regulation governing flight time, duty and rest for flight crew members in Part 121 passenger operations; all-cargo operations remain under Subparts Q, R and S of Part 121 unless the operator elects to come under Part 117. In force since 14 January 2014, it replaced a framework built almost entirely on flight time with one explicitly grounded in fatigue science.

Its logic resembles ORO.FTL without matching it. Maximum flight duty period also depends on report time and number of flight segments, within a range of 9 to 14 hours, reduced by 30 minutes for an unacclimatised crew member. Minimum rest is 10 hours including an 8-hour uninterrupted sleep opportunity. Part 117 further requires 30 consecutive duty-free hours within the 168 hours preceding any duty or reserve period, plus cumulative caps of 100 flight hours in any 672 hours and 1,000 hours in any 365 days, alongside flight duty period caps of 60 hours in any 168 hours and 190 hours in any 672 hours.

Two differences matter when comparing regimes. Part 117 places fitness for duty jointly on the operator and the crew member: a pilot must decline a duty if too fatigued to perform it safely, regardless of whether the numeric limits are met. And its scope is confined to Part 121 passenger operations: Part 135 operations run under a separate and, on several points, more permissive body of rules, which is the source of a long-running harmonisation debate in the United States.

For international operators, the practical consequence is that the same trip can be legal on one side of the Atlantic and not on the other, which makes rule sets a configuration item rather than a constant.

Part 135 (14 CFR Part 135)

14 CFR Part 135 is the United States regulation covering on-demand and commuter air carrier operations: charter, air taxi, air ambulance and a large share of business aviation flown for compensation. It carries its own flight time and rest regime, distinct from the Part 117 rules that apply to airlines.

For unscheduled operations the limits are expressed as flight time caps per calendar quarter, per two consecutive quarters and per calendar year — 500, 800 and 1,400 hours respectively, under §§ 135.267 and 135.269 — alongside a capped duty day and a minimum rest requirement of 10 hours before and after it. Scheduled operations under Part 135 fall under separate provisions with their own ceilings.

The planning challenge in this segment lies less in the numbers than in the tempo. Missions are confirmed late, change often, and each base holds few crew members. Compliance therefore cannot be a monthly exercise performed when the roster is published; it has to be re-established every time the programme moves, which is several times a day at an active operator.

SkAI Tech supports Part 91 and Part 135 rules natively alongside ORO.FTL, with the operator's own internal policies layered on top of the regulatory baseline.

Part 91 (14 CFR Part 91)

14 CFR Part 91 sets the general operating and flight rules applicable to all civil aircraft in the United States. It is the regime under which corporate flight departments fly their own aircraft for their own business purposes, without carrying third parties for compensation.

Its distinguishing feature for crew planning is the absence of general flight, duty and rest limits comparable to Part 117 or Part 135: the only provisions of that kind sit in Subpart K. The responsibility simply shifts to the operator, which must define and enforce its own fatigue rules, commonly modelled on Part 135, on insurer requirements or on IS-BAO practice. Subpart K, covering fractional ownership programmes, does impose explicit duty and rest limits: 10 hours of minimum rest, a duty period capped at 14 hours, and the same 500, 800 and 1,400 flight hours as Part 135.

In practice the better-run flight departments hold themselves to strict internal limits, because their crews are few, their missions unpredictable and their reputational exposure high. They then meet exactly the same planning difficulties as commercial operators, without the regulatory framework that would force the trade-offs into the open.

This is a textbook case for a configurable planning engine: the rules to enforce are the organisation's own, but they need to be applied with the same rigour as a regulatory limit, and evidenced with the same discipline when the insurer or the board asks.

Type rating (qualification de type)

A type rating is the endorsement on a pilot licence that authorises its holder to operate a specific aircraft type. It is earned through dedicated training and a skill test, and kept valid through periodic checks. Without it, no pilot may be assigned to a flight on that type, whatever their total experience.

Type ratings are required for the more complex aircraft: above a defined maximum take-off mass threshold, and for turbojet aeroplanes, with lighter aircraft covered by class ratings. Closely related variants may be grouped, allowing pilots to operate several models after reduced difference training.

Far from an administrative detail, type ratings are a first-order economic variable. Training a pilot on a new type is expensive and takes them out of the line for weeks; conversely, a mixed fleet fragments the pool of assignable crew. Each additional type reduces planning flexibility by more than the aircraft count alone would suggest: a two-type fleet of ten aircraft is materially harder to crew than a single-type fleet of the same size.

In an optimisation model the type rating is the most elementary constraint of all. In integration it is one of the most error-prone data items: validity dates, upcoming checks and grouped ratings must be read from the source system, never re-keyed.

Recency and currency (maintien des compétences)

Recency, also called currency, is the requirement to have performed certain activities recently in order to keep exercising the associated privileges: a number of take-offs and landings within a rolling window, specific night experience, and periodic proficiency and line checks. Lapsed recency suspends the privilege until it is restored.

The best-known requirement is three take-offs and landings in the preceding 90 days to carry passengers, with a distinct provision for night operations whose content differs between regimes. Commercial operators add their own mandatory checks (simulator proficiency checks and line checks at defined intervals), which are duties in their own right and must be rostered like any other.

What makes recency interesting to plan is that it works in both directions. It forbids assigning a crew member who has lapsed, and it obliges the operator to give people enough flying to stay current. A pilot parked on long-term reserve, on sick leave or in a ground role loses validity, and restoring it costs simulator slots, instructor time and calendar.

Rosters that ignore recency produce a cliff effect: several crew members fall out of currency in the same week, invariably at the worst possible moment. Treating recency as a first-class planning constraint, on a par with FTL, prevents that silent accumulation.

Block hours (heures de bloc)

Block hours measure the time from the moment the aircraft moves off its parking position for departure to the moment it comes to rest at the arrival stand. They are the unit of account for regulatory flight time, for ACMI billing and for a large share of operating cost and crew pay.

Measuring from block to block rather than wheels-up to wheels-down is deliberate: it captures taxi time, which is substantial at congested airports, and it is objectively recorded. Block time feeds each crew member's flight time counters, scheduled maintenance intervals, and remuneration under many collective agreements.

In ACMI contracts, block hours are the contractual unit. The lessee pays an hourly rate, usually with a guaranteed monthly minimum. The lessor's entire economics then rest on producing the contracted block hours with as few crew as possible, across several contracts at once. A crew member in the wrong place is a block hour that cannot be invoiced. This is what makes crew planning a direct margin lever in wet lease operations rather than a back-office function.

Block hours should not be confused with the flight duty period, which is longer and includes briefing and ground time, nor with duty time, which covers any task performed for the operator.

Duty time (temps de service)

Duty time covers any period during which a crew member performs a task for the operator: flying, positioning, training, simulator sessions, checks, administrative work and standby according to its form. Under ORO.FTL it is capped at 60, 110 and 190 hours over 7, 14 and 28 consecutive days respectively.

Those duty caps sit alongside flight time caps: 100 hours in any 28 days, 900 hours per calendar year, 1,000 hours in any 12 consecutive calendar months. The two families are independent, which means a crew member can be blocked by duty counters while their flight counters are nowhere near their limits. That happens routinely at operators with heavy standby, training or ground-duty loads, and it surprises planners who track only flying.

The rolling nature of these windows is the core difficulty. Apart from the calendar-year cap, nothing resets on a fixed date: every new assignment shifts the window and recomputes the remaining allowance at every future date. An assignment that is valid when made can render someone unavailable a fortnight later, with no warning in between.

This is why compliance checking belongs inside the generation engine rather than in a review step performed on a finished plan.

Augmented crew (équipage renforcé)

An augmented crew is a flight crew carrying at least one pilot beyond the minimum required composition, so that each pilot can take in-flight rest in a dedicated facility. Carrying additional crew is what allows the flight duty period to be extended well beyond the basic maximum.

This is the long-haul regime. Under ORO.FTL the FDP can reach up to 17 hours depending on how many additional pilots are carried and on the class of rest facility available: a separate bunk, a reclining seat in a screened cabin area, or an ordinary passenger seat, each class earning a different allowance. The logic is explicit: the quality of sleep obtainable on board governs how much extra duty is permitted.

Augmentation converts a compliance problem into an economic one. An extra pilot costs a salary, a seat and sometimes an extra hotel night, but it avoids a technical stop, a crew change down route, or a trip that simply cannot be closed. Which option wins depends on the network, the fleet and actual crew availability, a question to settle by simulation rather than by instinct.

Note that commander's discretion is also extended with an augmented crew, from two hours to three. It remains a tool for the day of operations, not something to plan on.

Reserve crew (équipage de réserve)

Reserve crew are crew members rostered to be available rather than to operate a specific flight, covering sickness, training overruns and operational disruption. Reserve levels are the main lever an operator has to trade the robustness of its schedule against its crew cost.

Sizing reserve is a permanent balancing act. Too little, and a single sick call triggers cancellations or calls to crew on their days off; too much, and the operator pays people not to fly. There is no universal percentage: the right level depends on observed absence rates, on how brittle the flying programme is, and on the real cost of a cancellation, which varies enormously between a scheduled regional flight, an ACMI commitment and a missed air ambulance mission.

Placement matters at least as much as volume. Reserve held at a base with nothing to cover that day, or held by someone not rated on the aircraft in question, is not reserve at all. The meaningful metric is effective coverage, not the headcount on reserve: what proportion of plausible failures can be absorbed without changing the published roster.

Optimisation engines make that coverage testable in advance by simulation, instead of leaving it to be discovered on the day.

Window of Circadian Low (WOCL)

The Window of Circadian Low is the part of the biological night when human performance is at its lowest. European rules place it between 02:00 and 05:59 in the time zone the crew member is acclimatised to, and reduce the maximum flight duty period of any duty that encroaches on it.

The boundaries come from physiology rather than administrative convenience. Core body temperature, alertness and information-processing capacity bottom out in the middle of the night regardless of how long a person has already been awake. A crew reporting at 04:00 after a full rest period is already, cognitively, in a degraded state before the first sector.

In practice the WOCL is what penalises very early starts and late-night arrivals. An FDP that encroaches on it is shortened, and the size of the reduction depends on how much of the window is encroached. It is also why the same sequence of flights can be perfectly legal in the afternoon and illegal at dawn, an asymmetry that surprises commercial teams negotiating slot times.

For air ambulance operators and express freight carriers, whose activity concentrates precisely inside this window, the WOCL is the normal operating regime, not an exception to manage. Distributing night exposure evenly across the crew community then becomes the single most effective lever against chronic fatigue.

Fatigue Risk Management System (FRMS)

A Fatigue Risk Management System is a data-driven system, built on scientific principles and operational experience, through which an operator continuously monitors and manages fatigue-related safety risks, ensuring crew members perform at an adequate level of alertness. It complements prescriptive flight time limitations.

The difference in nature from FTL is fundamental. FTL is prescriptive: it draws the same boundaries for every operator. An FRMS is a management system: it measures the fatigue actually generated by one operator's schedules, on its routes, with its crews, and adjusts decisions accordingly. In most jurisdictions the two coexist, with the FRMS supplementing rather than replacing the prescriptive limits.

An FRMS rests on four components: a fatigue policy and governance, data collection processes (crew reports, actigraphy, modelling), a risk assessment and mitigation process, and promotion and training activity. Where an authority approves an FRMS, it can allow controlled variation from certain prescriptive limits within a defined scope.

The practical obstacle is almost always data. An FRMS is only as good as its picture of what crews actually worked, not what they were published to work. That is a recurring argument for planning systems that keep the full history of assignments and of every change made to them, rather than overwriting the plan each time it moves.

Biomathematical fatigue model (modèle biomathématique de fatigue)

A biomathematical fatigue model estimates the predicted alertness of a crew member at any point in a schedule, from the duty pattern and assumptions about sleep obtained. It allows two equally legal rosters to be compared objectively in terms of the fatigue risk they generate.

These models come from chronobiology research and typically combine a circadian process, a homeostatic sleep pressure component that builds with time awake, and sleep inertia after waking. They output alertness or risk scores to be read against thresholds the operator has chosen. SAFTE-FAST and FAID are among the most widely used in aviation.

Their value to planning is direct: they show what compliance hides. Two rosters can both sit comfortably inside FTL and still produce very different fatigue profiles depending on where the night duties fall, how trips are ordered, and how realistic the sleep opportunities are. The model turns that difference into a number that can be argued about in a management meeting rather than a feeling that a pattern is "heavy".

Their limits deserve equal billing. A model predicts for a population rather than an individual, and it rests on estimates of sleep, not measurements; a favourable score is therefore no guarantee of safety. It informs a trade-off without settling it.

Crew fatigue (fatigue de l'équipage)

Crew fatigue is a physiological state of reduced capability to perform, caused by sleep loss, extended wakefulness, circadian phase or workload. It degrades alertness, reaction time and decision-making, and it is recognised across the industry as a systemic safety hazard rather than an individual failing.

It must be distinguished from ordinary tiredness. Fatigue in the regulatory sense is measurable, has documented effects on performance, and is driven mainly by structural factors: the roster decides what time someone gets up, how long they stay awake, and how good a sleep opportunity they get, far more than personal habits do.

Two forms coexist. Acute fatigue follows a long duty or a short night and is recovered by adequate rest. Cumulative fatigue builds when successive rest periods no longer clear the debt accrued over weeks; it is the more dangerous of the two, because those affected are poor judges of their own impairment and because it leaves no trace in regulatory counters.

The most effective planning lever is rarely to reduce hours. It is to arrange them better: limiting consecutive night duties, avoiding rapid alternation between early starts and night duties, protecting genuine local nights inside recovery rest, and sharing the most punishing slots evenly across the crew community.

ACMI

ACMI (Aircraft, Crew, Maintenance, Insurance) is a leasing arrangement in which the lessor supplies the aircraft together with its full crew, its maintenance and its insurance, operating under its own Air Operator Certificate, in exchange for a block-hour rate usually backed by a guaranteed monthly minimum.

The acronym lists exactly what the lessor provides. Everything else stays with the lessee: fuel, airport and navigation charges, ground handling, passenger taxes, and commercial distribution. The flights carry the lessee's flight number, but they are operated under the lessor's AOC, the point that legally defines the arrangement and determines which authority and which FTL rules apply to the crew on board.

For the lessor, crew planning sits at the core of the business model rather than in a support role. Margin depends on producing the contracted block hours with the smallest viable crew complement, across several contracts at once, frequently in different countries and under different customer requirements. A crew member in the wrong place is an hour that cannot be invoiced, and a missed contractual minimum is a penalty.

It is the segment where optimisation shows up in the income statement most directly, and where rosters must survive contracts starting, moving or ending mid-season.

See also: Wet lease, Dry lease, Block hours, ACMI operators

Wet lease (location avec équipage)

A wet lease is the lease of an aircraft together with its crew, maintenance and insurance, operated under the lessor's AOC. It is the market's generic name for the ACMI arrangement: the lessee buys complete, immediately operable capacity without recruiting, training or qualifying a single crew member.

What the lessee is really buying is the scarcest resource in the industry: qualified crew available now. Wet leases cover seasonal peaks, delayed aircraft deliveries, extended technical groundings, trial routes, and temporary crew shortfalls, including shortfalls caused by planning that has stopped working.

In the European Union, wet leasing aircraft registered in third countries is governed by Regulation (EC) No 1008/2008, which requires equivalent safety standards, a justified need such as a seasonal peak or exceptional requirement, and limits the duration of the arrangement.

For planning, a wet lease creates an unusual situation: two crew populations operate the same commercial programme under two regulatory regimes and two sets of employment terms. Handling both in one system with distinct rule sets avoids assignment errors and the duplicate data entry that otherwise appears within days of the contract starting.

Dry lease (location coque nue)

A dry lease is the lease of the aircraft alone: no crew, no maintenance, no operational insurance. The aircraft is operated under the lessee's AOC, with the lessee providing crew, maintaining the aircraft and carrying operational responsibility. It is a fleet financing arrangement rather than a capacity arrangement.

Typical terms run to years, against weeks or seasons for a wet lease. The large aircraft lessors work almost exclusively this way: they own aircraft and lease them to airlines that know how to fly. Rent is monthly and independent of utilisation, which places the whole utilisation risk on the lessee.

The crew planning consequences are the exact mirror of a wet lease. A dry lease brings no crew: it grows the fleet without growing the workforce. Training, type ratings and recruitment therefore have to be planned well before delivery, and lead times for type rating courses and simulator slots are the binding constraint far more often than hiring is. An aircraft delivered without qualified crew is an asset generating cost and no revenue.

This is one of the situations where modelling crew capacity several months out pays for itself outright: it determines the date the aircraft can realistically enter service.

See also: Wet lease, ACMI, Type rating, Pilot shortage

Damp lease

A damp lease is a variant of the wet lease in which the lessor provides the aircraft and the flight crew while the lessee supplies its own cabin crew. Operations remain under the lessor's AOC. The term is market usage rather than a regulatory category, so the contract wording governs.

The motive is nearly always commercial. An airline leasing in capacity wants to keep its onboard product (service language, catering, uniform, retail) because that is where passengers perceive the brand. A damp lease lets it buy flying capacity without buying someone else's customer experience.

It does create a hybrid planning situation that deserves to be spelled out in the contract: two employers, potentially two different duty regimes, and a coordination requirement on every rotation so that flight crew and cabin crew are available at the same place and time. Every mismatch is a cancelled flight, and responsibility for the mismatch is exactly the kind of thing contracts fail to anticipate.

The absence of a regulatory definition implies one working rule: never reason from the name of the deal, only from what it specifies. Whose AOC, who provides the flight crew, who provides the cabin crew, who pays for operations.

Deadheading (mise en place)

Deadheading is the transport of a crew member as a passenger, at the operator's request, to reach the place where they will start a duty or to return to base afterwards. It counts as duty time and is paid, but it produces no flight hours and no revenue.

It is a doubly punitive cost. It consumes budget (a ticket on another carrier, or a seat blocked on the operator's own flight) and it consumes duty time, which reduces the crew member's future flying capacity. A roster full of deadheads is a roster wasting its most constrained resource.

That does not make deadheading an anomaly to eliminate. It is sometimes the only way to recover a crew stranded away from base, to cover a flight at an under-crewed station, or to repair an irregularity. Trying to remove positioning altogether is the wrong goal; the useful question is which positioning legs create more value than they cost.

Deadhead volume is also an excellent proxy for the quality of a pairing structure. For an unchanged flying programme, fewer deadheads means better-built trips, not a degraded service. It is one of the easiest savings to quantify when building the business case for an optimisation tool, because the tickets are already in the accounting system.

Positioning flight (vol de positionnement, ferry flight)

A positioning flight is a flight operated without revenue passengers or cargo for the sole purpose of moving the aircraft to where it is needed, or back to base. It burns fuel, airframe hours, duty time and regulatory flight time counters while generating no revenue.

It should not be confused with deadheading, which moves crew as passengers. A positioning flight moves the aircraft with its crew at the controls. Both stem from the same cause, a resource in the wrong place, but their costs and their regulatory consequences differ.

In charter and business aviation, empty legs are a structural share of activity rather than an occasional inefficiency: a client books one way, and the aircraft must reposition for the next mission or return home. An operator's profitability rests substantially on chaining requests so as to minimise those segments, and on selling the empty legs that cannot be avoided.

From the crew side, a positioning flight counts as a sector: it reduces the maximum FDP available for the rest of the day exactly as a revenue sector does. This consequence is routinely overlooked when accepting a last-minute mission, and it is a common source of same-day legality problems in on-demand operations.

Medevac (air ambulance, évacuation sanitaire)

Medevac is the air transport of patients who require medical monitoring or treatment in flight, with medical equipment and a clinical team on board in addition to the flight crew. Missions are unpredictable by nature, frequently nocturnal, and governed by contractual activation and launch times.

It is the segment where crew planning least resembles the rest of aviation. There is no flying programme to roster against: the roster places availability, in the form of on-base duty periods or home standby, and missions land inside it. A gap in coverage means a breach of contract with a health authority or an insurer, not a mere commercial inconvenience.

Three constraints compound. A large share of missions launches at night, squarely inside the Window of Circadian Low, which makes fair distribution of night exposure the central fatigue question. Every mission needs a flight crew and a medical team simultaneously, with two qualification frameworks and often two employers. And base establishments are small, so a single uncovered absence closes the base outright.

On the regulatory side, aeroplanes and helicopters part company. Emergency medical service operations by aeroplane are carved out of ORO.FTL and fall under Subpart Q of the former EU-OPS and under national rules. Commercial air transport by helicopter is governed instead by the national law of the member state where the operator has its principal place of business. In both cases the applicable regime must be confirmed with the competent authority rather than assumed.

HEMS (Helicopter Emergency Medical Service)

HEMS refers to helicopter operations conducted to facilitate emergency medical assistance where immediate and rapid transportation is essential: response to an accident scene, inter-hospital transfer, or the carriage of medical teams, organs or supplies. It is a specialised category with its own operational requirements.

HEMS operations carry dedicated rules on crew composition and training, on the selection and use of operating sites, on operating minima and on aircraft equipment. The HEMS technical crew member holds a role of their own, distinct from cabin crew in passenger transport, and their availability is as binding a constraint as the pilot's.

Commercial air transport by helicopter, HEMS included, does not fall under ORO.FTL: its flight time limits are set by the national law of the member state where the operator has its principal place of business. Two HEMS operators in two European countries can therefore plan against noticeably different rules.

From a planning standpoint, HEMS pushes every medevac characteristic to its extreme: very small bases, often two to four crews, permanent coverage, launch times measured in minutes, and missions triggered without notice. The margin is so thin that one unforeseen absence can suspend the coverage of an entire region. A HEMS roster is therefore judged on robustness before cost: how many simultaneous failures it can absorb before a coverage gap opens.

See also: Medevac, Standby duty, Roster pattern, Reserve crew

Disruption management (gestion des irrégularités)

Disruption management is the set of real-time decisions taken to restore a flying programme disturbed by weather, technical failure, airspace closure or crew unavailability. It means rebuilding a feasible and compliant plan from the situation as it actually is, under severe time pressure.

The scale of the problem is industry-wide. According to EUROCONTROL's CODA Digest 2024, reactionary delay, meaning delay inherited from a previous rotation rather than caused locally, remained the largest contributor to average delay per flight, at 46 % of delay minutes, or 8.0 minutes per flight. Most delay, in other words, is propagated, not created.

Three resources must be restored simultaneously and in the right order: aircraft, crew and passengers. Crew recovery is almost always the binding one, because it is bounded by rules that do not negotiate. A crew member who has reached a duty limit cannot fly, whatever the commercial stakes and however willing they are.

Which is why the best disruption management begins at roster construction. A plan built with slack in the right places and reserve positioned where knock-on risk is highest absorbs quietly what a tightly packed plan converts into a cascade of cancellations.

Crew recovery (récupération équipage)

Crew recovery is the part of disruption management concerned with reassigning crew after a disruption so that remaining flights can operate without breaching flight, duty or rest limits. It is usually the limiting factor in restoring a programme, ahead of aircraft and ground resources.

The mechanism is familiar to every operations controller. A delay extends the current duty period, pushes a crew to its limit, and leaves the next flight without a crew even though the aircraft is serviceable and standing right there. The problem then propagates, because whoever is called in as a replacement is no longer available for their own duties the following day.

The available levers are few and all costly: call out reserve, contact crew on rest where permitted, position a crew by another flight, deliberately delay a departure to preserve the legality of the rest of the day, or cancel, choosing the cancellation that propagates least damage. A tool's job is to show within seconds which options are actually legal; the choice itself stays with the controller.

Commander's discretion is not a recovery lever that can be planned. It applies to unforeseen circumstances arising at or after report time and is reported to the authority.

Operations Control Centre (OCC)

An Operations Control Centre is the unit that runs an operator's flying in real time: following flights, managing disruption, and coordinating aircraft, crew and ground handling. It is where the decisions to delay, divert, reassign or cancel are taken, and where their consequences are absorbed.

Its shape varies enormously with the size of the operator. A scheduled airline runs a staffed centre around the clock, with separate desks for flight watch, crew control, maintenance control and customer recovery. At a charter or air ambulance operator the same function is held by one or two people who simultaneously handle flight following, planning and customer contact, often on a phone at three in the morning.

What they have in common is the mismatch between the speed at which decisions are required and the volume of information that must be cross-checked. A controller with ten minutes to decide whether to delay a departure or swap a crew cannot open a spreadsheet of FTL counters. What they need is the list of crew who can legally take the duty, current as of now, with the consequence of each option visible.

That is why the planning system and the operational picture must share one source of truth. An API integration with the existing operations system provides that; parallel data entry never does.

Crew performance KPIs (KPI de performance équipage)

Crew performance KPIs are the measured indicators used to steer the crew planning function: crew productivity, utilisation, unproductive time, deviations after publication, reserve usage, overtime, training compliance and publication lead time. Which ones an operator tracks determines the behaviour it gets.

Choosing indicators is a policy decision. Steering on cost per block hour alone pushes mechanically towards tightly packed rosters, whose price is paid elsewhere: in fatigue, attrition and irregularity. A balanced dashboard carries at least three families: economic efficiency, operational robustness, and crew quality of life.

A few indicators prove unusually informative in practice. The gap between published and operated schedule measures how credible the roster is. Publication lead time drives crew satisfaction more than almost anything else within the planner's control. The distribution of nights and weekends across the crew community turns fairness from a debate into a measurement. And roster production time decides whether the operator can afford to replan at all when the programme changes.

That last dimension is where the figures observed at operators running SkAI Tech sit: an average generation time of 12 minutes and 20 % fewer schedule deviations, across a scope of more than 70 aircraft and 900 crew members. These results are observed across SkAI Tech deployments in 2025-2026 and vary with fleet size and operating model.

Pilot shortage (pénurie de pilotes)

Pilot shortage describes the gap between the supply of qualified, available pilots and operators' demand for them. It shows up less as unfilled vacancies than as concrete friction: hard-to-fill type-rated positions, long training lead times, wage escalation and schedules with no resilience left.

The most cited projections concern North America. Oliver Wyman's January 2023 analysis placed the widest supply-demand spread in 2026, at around 24,000 pilots, easing to roughly 17,000 by 2032; its October 2023 revision lowers the projected 2032 shortfall to about 13,300 pilots. That October 2023 revision also documents the tension through pay: captains' wages up 46 % at US mainline carriers since 2020, and up 86 % at US regionals. These are projections and North American data, sensitive to assumptions about traffic growth and retirements: they should always be quoted with their source and vintage, never as settled fact. The gap between two successive editions makes the point on its own.

In Europe, and particularly in niche segments (business aviation, air ambulance, ACMI), the tension is less about the total number of pilots than about the availability of specific profiles: rated on the right type, based in the right place, current on their checks. A market can be comfortable in aggregate and acutely short of the twelve people a given operator actually needs.

The planning implication is direct. When the resource is scarce, every hour lost to unnecessary positioning or misplaced reserve costs more, and every unanticipated resignation does more damage. Optimisation becomes a capacity lever, not merely a cost lever.

Flight management system (FMS, système de gestion des opérations)

The abbreviation FMS covers two different things. In avionics, the flight management system is the onboard computer that handles navigation and flight profile. In operations language it also refers to the flight and crew management software that holds an operator's programme on the ground.

It is the second meaning that concerns crew planning. The operations system is the source of truth for flights, aircraft, crew members and their qualifications. It receives requests, produces operational documentation and feeds invoicing, and it is where the operator's processes already live.

SkAI Tech does not replace that system. The platform connects to it through an API as an add-on: it reads flights, crew and constraints, computes an optimised and compliant roster, and writes the resulting assignments back into the system of record. The operator keeps its processes, its documents and its billing tool, and gains the optimisation capability it was missing.

This positioning avoids the classic failure mode of all-in-one suite replacement projects: a long, risky and expensive migration undertaken to obtain one genuinely new function, with everything else rebuilt only to reach parity.

Sources

Regulation (EU) No 965/2012 and Regulation (EU) No 83/2014: Part-ORO Subpart FTL, together with the scope and derogations of Article 8 (air taxi, EMS and single-pilot operations by aeroplane; commercial air transport by helicopter).
EASA, Easy Access Rules for Air Operations — Part-ORO Subpart FTL and CS FTL.1: cumulative caps, FDP tables, extensions, in-flight rest with an augmented crew (CS FTL.1.205) and rest requirements.
Regulation (EEC) No 3922/91, Annex III Subpart Q, and the national rules adopted under its Article 8(4) — in France, those overseen by the DGAC.
14 CFR Part 117 (§§ 117.1, 117.13, 117.23, 117.25), Part 135 (§§ 135.261 to 135.271), Part 91 Subpart K (§§ 91.1057 to 91.1062), and Part 61 (§§ 61.31 and 61.57) for type ratings and recency.
Regulation (EC) No 1008/2008 for the EU framework governing the wet leasing of third-country registered aircraft.
Oliver Wyman (January 2023), Industry Efforts Are Easing Pilot Shortage Severity, and (October 2023) North American Pilot Shortage Shows 23% Smaller Gap In 2032.
EUROCONTROL (2025), All-Causes Delays to Air Transport in Europe — Annual 2024 (CODA Digest), Europe/ECAC scope, year 2024.

All resources