We are in the age of data, but those who handle it don’t always seem to want to heed what that data reveals. It happens, for example, with boarding the plane before take-off, a tedious process that usually condemns us to long waits, as well as to some tension and stress.
All this could be minimised if airlines paid more attention to the data. Some do, but science is very clear that the boarding method often used is not optimal. Hopefully those who manage that process will read us.
Shipping Methods: There Are Them, There Are Them
A Boeing study long ago revealed that the boarding of an airplane was 50% slower in 1998 than in 1970. We will have made progress on many things, but the rate of passengers entering the plane by the minute has worsened dramatically in recent years.
We already mentioned in the past that there are several shipping methods that have been used over the years. The best known and most popular are the following:
From back to front: if the shipment is from the door closest to the nose, it is normal to find a process that tries to go from the last rows, the closest to the tail, to the first. The aim is to try to avoid crowds in the first rows with people trying to access the last ones.
From outside to inside (“Wilma method”): first the windows, then the intermediate seats and finally the aisle seats.
By seat numbering: a combination of the two above that takes both factors into account.
Random: it does not take into account any of these possible boarding priority parameters and instead chooses rows and seats entirely at random.
To evaluate the differences in these different boarding modes there are studies like Menkes van den Briel’s, the problem is that although in the first few minutes improvements can be noticed between them, they all converge to a very similar end time.
Much more detailed is the fantastic study (PDF) of Jason Steffen, an astrophysicist from the Fermi National Laboratory in Illinois who even devised his own boarding method through his results.
The so-called “Steffen method” organizes boarding in alternate rows with the window seats first moving backwards and forwards. Thus, 12A would be followed by 10A, 8A, etc., and then we would return to 9A, 7A, 5A, and so on, and then fill the middle seats and the aisle seats below.
That theory was singular, but would it really be fulfilled? That’s what a television producer named Jon Hotchkiss proposed to Steffen. To get the idea off the ground, he created a kind of model of a Boeing 757 and had the help of 72 volunteers who acted as virtual passengers.
The result, as shown in the table above, was surprising: the Wilma method was very good, but his method confirmed his theory and was even better at making a shipment in the shortest possible time.
These experiments did not stop there: the popular TV series “Myth Hunters” studied precisely this question in episode 222.
Those responsible for the program simulated the boarding of an airplane with 173 passengers and even took into account realistic situations and indicated to 5% of the passengers that they were “problematic”: to sit in the seat that was not, to travel with children or to lose time by putting their luggage in very quietly and obstructing the passage of other passengers.
In this process there were several curious results, mainly because the boarding time was evaluated as well as the satisfaction of the passengers who boarded. The quickest (except priority, that each passenger choose their place) was the one that generated the worst satisfaction.
The most complete was probably the singular inverse pyramid (priotaries in the first place, after which a “cascade” of zones was used that was dispersed from the rear window seats), above Wilma and Wilma with blocks: it took a little longer than these methods, but it was the one that caused more satisfaction.
There are other equally interesting variants, and one of the most striking is the so-called “magic carpet” which even has its own website and promises the best results. Referring to the program “Myth Hunters” in which the best was the inverse pyramid with about 15 minutes of boarding time, the magic carpet would achieve shipments of 13 minutes that could even be lower.
Studies in this field are surprisingly numerous: we have the Michael Schultz and Hartmut Fricke (2011), plus an update from the same authors and a third contributor, Thomas Kunze (2013).
Even more recent is the work of Shafagh Jafer and Wei Mi (2017) which concludes that the most appropriate method is the Wilma, but we can also run our own simulations with alternatives such as Simul8, whose responsible coincided with the good behavior of the Wilma method.