Encouraged by a recent and particularly inefficient experience with air travel, I note a relevant, intriguing academic article about airplane boarding schemes, published in the arXiv by Jason Steffen, a researcher at the famed Fermilab.    

Via computer modeling, Steffen has discovered that the current process by which airlines load passengers onto aircraft is, in fact, extremely inefficient. The popular method of loading passengers into the plane starting from the back, working toward the front, is not an optimally efficient loading scheme. Although it appears to be the exact opposite of the intuitive worst option– loading passengers front to back– it surprisingly is only the second worst to this method.   

But why? Steffen tells us the biggest hinderance to boarding efficiency is carry-on luggage stowaway. How many times have we all stood in that narrow isle only to wait for the woman in 15A to get her kids seated while simultaneously dodging the man from 14D’s wild gestures in his attempt to close the overhead compartment? Steffen concludes the optimal solution is to board the plane in a manner conducive to multiple, simultaneous luggage-stowing operations. Ten passengers at time, in alternating rows, then opposite aisles, produces “nearly a five-fold reduction in the time that it takes over the worst case.”  

The network effect here is two-fold. First, in a sense, we notice the Braess’ Paradox comes into effect. If we consider loading people on from the back to front aisle by aisle, this is similar to a situation where many paths are open at “low cost” to a destination. However, as we’ve seen from class, too much access at a low cost actually can reduce efficiency overall due to too much traffic. In a similar sense, too many people causes crowding– traffic– and incurs an expense on all passengers boarding according to the current popular boarding scheme.  

Secondly, why is it that airlines have not adopted Steffen’s solution? Sure, the article was published only half a month ago, but it certainly demonstrates, at least in-silico, that the current passenger boarding model is highly flawed. In an age of airline cost-cutting, financial reasons demand that this virtually free performance booster can be and should be implemented immediately. However, we recall from class that cascade behavior reinforces status-quos. Indeed, despite the superiority of Steffen’s boarding scheme, it will likely require multiple airlines to make the switch simultaneously– simply due to its current relative unpopularity. Publishing an excellent idea in an open academic archive, unfortunately, does not necessarily lead to appropriate action.   

http://aps.arxiv.org/PS_cache/arxiv/pdf/0802/0802.0733v2.pdf

Steffen, Jason H. “Optimal boarding method for airline passengers.” arXiv:0802.0733, 19 Mar 2008.