Boarding sequence trials have been run, in one form or another, for thirty years. Back-to-front, window-middle-aisle, outside-in, random assignment: each has been modelled, each has been trialled, and each has produced gains in simulation that largely evaporate in operation.
The reason is well understood. Real boarding is dominated by two things the sequence cannot control: how long each passenger takes to stow a bag, and how many passengers are travelling in groups who will not board separately regardless of what the sequence says.
Where the gains actually came from
Operators reporting genuine reductions in boarding time over the last two years have made three changes, none of which involve the queue.
Larger bins. Bins sized to take a standard cabin bag on its side rather than flat increase capacity by roughly half. Fewer bags go to the hold at the gate, and gate-checking is one of the largest single sources of boarding delay.
Second door boarding. Where the stand geometry allows it, using a rear airstair or second jetbridge splits the aisle congestion in two. The gain is close to linear and is the largest single lever available.
Bin state awareness. Cabin crew with a live view of which bins are full can direct passengers immediately rather than after a failed attempt. Several operators now do this with a simple tablet view updated by crew rather than any sensing technology.
What the data says about sequences
A large multi-operator study found that switching between the common boarding sequences changed average boarding time by less than the variation between two consecutive flights on the same route with the same sequence. The signal was smaller than the noise.
That does not mean sequence is irrelevant. It means it is a second-order effect that only becomes visible once the first-order problems — bag storage and single-door access — have been addressed.
The remaining constraint
Turnaround time is now frequently limited by cleaning and catering rather than by boarding at all. Operators who have optimised boarding to the point where it is no longer the critical path have discovered that the next constraint is considerably harder to move, because it depends on third-party contractors working to their own service level agreements.
