Operations
Roster optimisation for regional airlines: rosters that hold up on the day
Optimising a regional airline's rosters means building plans that survive contact with the day: crews spread across bases, rotations that close, reserve positioned where it will actually be used. Between two equally compliant rosters, the one that absorbs disruption saves far more than the one that looks cheapest on paper.
Regional carriers inherit every hard part of crew planning at once: a dense short-sector programme, multiple bases, versatile crews, and maximum exposure to knock-on effects. This guide covers the levers that matter, and why the cheapest roster rarely turns out to be the best one.
What makes regional rostering harder?
Many sectors means many constraints per day
A regional duty day routinely strings together three, four or more sectors. The maximum flight duty period drops by 30 minutes for each sector from the third onwards (see our ORO.FTL guide), so every extra sector tightens the space of feasible rosters. Where a long-haul operation manages a handful of long assignments, a regional carrier manages a large number of small, tightly constrained ones.
Multiple bases fragment the resource
A crew member based in one place is not freely usable elsewhere without positioning (deadheading), which costs money, burns duty time and makes the day more fragile. More bases mean better local coverage but a fragmented pool: each base has to be sized to absorb its own absences.
Rotations have to close
A rotation that fails to return crew to home base creates a debt: a night stop, a positioning sector, or worse, a crew member stranded away from home the night before recovery rest. “Isolated rotations”, the ones that never quite reconnect, are the classic signature of a roster built flight by flight rather than as a whole.
Knock-on delay is the rule
On short sectors the same aircraft and crew carry the day forward: a delay on the first rotation propagates through everything that follows. EUROCONTROL's 2024 CODA data puts reactionary delay at 46% of European delay minutes, or 8.0 minutes per flight (CODA Digest, Annual 2024). Put plainly: close to half of Europe's delay is not caused on the day by anything specific to the flight. It is inherited. Regional carriers sit on the front line: in summer 2025 the average delay per flight for ERA members still stood at 2.39 minutes, more than double the EU target of 0.9 minutes (ERA, Summer 2025: Regional air travel in numbers).
Why isn't the cheapest roster the best one?
A roster can be compliant, cheap on paper, and expensive in operation. Three sources of gap:
- The slack is in the wrong place. A tight sequence where delays are structural (the first morning rotation, a congested hub) breaks at the first incident. Robustness means placing buffer where it is statistically useful; buffer everywhere is just overstaffing.
- Reserve is badly positioned. A reserve crew is only worth its cost if it sits where the disruption will happen, available at the right hour and current on the right type. Plentiful but misplaced reserve pays without covering.
- Roster deviations are paid twice. Every gap between the published roster and what actually flew burns management time, erodes crew trust and generates unbudgeted cost. It is also the indicator that moves fastest once an operator equips: SkAI Tech measures 20% fewer roster deviations at equipped operators.
What does optimisation look like in practice?
- Optimise globally, not flight by flight. An optimisation engine compares complete rosters against one another instead of solving each assignment in isolation. That is what makes rotations close and bases balance; sequential reasoning mechanically strands rotations at the end of the month.
- Make stability an explicit objective. Robustness, cost and fairness should be weighted deliberately rather than left to chance. A serious tool lets the operator set that dial.
- Plan leave alongside production. Granting leave without a coverage view creates the holes later filled with overtime and reserve callouts.
- Measure deviation, not just planned cost. The honest quality measure for a roster is the gap between plan and reality. That is the KPI worth instrumenting in the crew dashboard.
- Be able to re-generate quickly. On a regional programme disruption will come regardless; what matters is replanning cleanly, in minutes rather than half a day (see our guide to crew disruption management).
That is SkAI Tech's approach for scheduled carriers: as an add-on to the existing operations system, connected by API, the platform generates compliant, optimised rosters in 12 minutes on average. Those results are observed across SkAI Tech deployments in 2025-2026, over a scope of more than 70 aircraft and 900 crew members, and vary with fleet size and operating model.
Flying a regional network? See how SkAI Tech works for scheduled airlines.
FAQ
At what fleet size does optimisation pay off?
Aircraft count matters less than combinatorics: sectors per day, number of bases, qualification diversity and disruption frequency. An eight-aircraft regional fleet across three bases is a harder problem than a larger single-base one.
Does optimisation make crew life worse?
It treats crew life as an explicit objective. Distribution of weekends, nights and layovers can be constrained at generation time. An optimiser aimed only at cost would indeed produce hostile rosters, which is why the ability to weight objectives yourself belongs on any selection checklist.
Do we have to replace our operations system?
No. An API-connected add-on reads flight and crew data where it already lives, optimises, and pushes rosters back. Team workflows are untouched (see our add-on vs all-in-one suite comparison).
How do we measure the benefit?
Three indicators are enough to start: time to produce the monthly roster, deviation rate between planned and actual, and the share of reactionary delay in total delay. The first two move within weeks; the third reads over a season.
Sources
- EUROCONTROL (2025, 25 July), All-Causes Delays to Air Transport in Europe — Annual 2024 (CODA Digest): reactionary delay at 46% of delay minutes, 8.0 minutes per flight, Europe/ECAC, 2024.
- European Regions Airline Association (2025), Summer 2025: Regional air travel in numbers: average delay per flight of 2.39 minutes for ERA members, against an EU target of 0.9 minutes.
- EASA, Easy Access Rules for Air Operations, Part-ORO Subpart FTL (ORO.FTL.205: maximum daily FDP reduced by 30 minutes per sector from the third sector).
- SkAI Tech KPIs (first-party data): average generation time and deviation reduction observed across 2025-2026 deployments, varying with fleet size and operating model.
