There are two ways to find out an inspection has come due. On the calendar, weeks ahead, with the hangar booked and the part in stock — or on the ramp, with the customer waiting and the aircraft stuck. The difference between the two is not luck: it is knowing the date of each due item, and not only the number.
Because this is the detail no spreadsheet solves by itself: a due limit in hours has no date. "40 hours left to the inspection" does not say whether the downtime is next month or a year from now — it depends on how much that aircraft flies. And it is the date, not the hours balance, that books the hangar, buys the part and schedules the crew.
Alis now answers that question across the entire fleet: each due item projected to a date by each aircraft's real pace, and those that fall together grouped into a suggested downtime. It is the new downtime forecast feature — and this post explains exactly how it works, because a forecast that does not explain where it came from is no use for deciding maintenance.
Where the pace comes from
The projection uses the data the operation already produces, in two distinct situations:
Aircraft with a logbook on the platform. The pace is the sum of hours flown in the last 90 days divided by 90. A deliberately short window: the pace of operation changes — an aircraft that changed mission, entered peak season or sat idle — and an average from years ago does not describe today's operation. The same goes for landings and cycles, each with its own series.
Maintenance-shop portfolio aircraft, with no logbook here. The shop does not have the customer's flight-by-flight record — but it has something almost as good: the hour meter readings that each work order records. Two consecutive readings become an "hours per day" pair; the series of pairs becomes the aircraft's pace.
Why the median, not the average
Here is the most important technical decision in the feature — and the reason this post does not use the word "magic."
A real maintenance record has absurd readings. A swapped hour meter, an extra digit in data entry, a record migrated from an old system. In an average, a single rotten pair — say, a reading that suggests 400 hours flown in one week — pulls the entire projection into the unreal. In the median, the absurd pair sits at the end of the line and decides nothing: the pace that counts is the one in the middle of the series, the aircraft's typical behavior.
It is learning from the operation's own data — which is what matters in what the market calls predictive maintenance. No black box: the forecast shows where the pace came from (logbook or work order readings), how many readings support it and which was the last. The decision still belongs to whoever knows the aircraft — with a defensible date in hand instead of a guess.
Each item on the right clock
An honest projection also respects the mechanics of maintenance tracking:
- An engine item is measured by the engine — the component's TSN is that of its own clock, not the airframe's. A remanufactured engine installed on an old airframe comes due by the engine's life;
- A calendar item already has a date — fire extinguisher, ELT, certificates: they enter the calculation directly, without projection;
- Whichever comes first, comes due first — an item with a limit in hours and in calendar time is projected on both clocks, and the nearest date rules;
- An overdue item always enters — even when no pace can be calculated. Absence of data cannot hide an open item.
The part that saves money: grouping
Projecting dates is half the value. The other half is what the system does with them: items on the same aircraft that come due less than 45 days from each other become a single suggested downtime.
The logic is that of whoever pays the bill: open the hangar once. If the 100-hour inspection falls in October and the extinguisher test in November, moving the extinguisher up to the same window costs much less than two periods of unavailability, two repositionings and two crew mobilizations. The suggested downtime is born with the list of items in the window — and becomes a real downtime with one tap, with the window, location and items pre-filled.
Anyone who follows the cost of the grounded aircraft knows: unplanned unavailability is the most expensive item in the maintenance spreadsheet. Grouping due items is the cheapest way to buy it back.
Where this appears in Alis
The forecast is on the Maintenance downtime screen, in the "Fleet forecast" section — for the operator's own fleet and for the shop's portfolio, in the same place. Each suggestion shows the window, the items that make it up, how many are already overdue and the pace that supported the projection. From there, the suggested downtime becomes a planned downtime, tracking the quote and work order when a maintenance shop is involved.
And like the rest of the platform, the projection is fed by the operation: the flight logged in the logbook moves the hours, the hours move the projection, and the projection moves the downtime — without retyping anything anywhere.
Want to see your fleet's forecast? Talk to Alis — if your operation already logs flights on the platform, the forecast is already calculated.
This article is informational and does not replace current regulations, the manufacturer's manuals, the approved maintenance program or the technical responsibility applicable to the maintenance of your aircraft. The due-date projection is planning support — the limits that count are those in the aircraft's technical documentation.