Jason Steffen (Physics and Astronomy) was a guest on NPR's Marketplace, talking about how to more efficiently board an airplane. He began researching the issue as a graduate student after being frustrated by slow boarding processes and flight delays.
The algorithm he used to find the optimal boarding method is called Markov Chain Monte Carlo. "You set up a scenario, make random changes to it. If the random change is an improvement, then you keep the change, and if the random change is not an improvement, then you keep the change sometimes.”
The optimal method Steffen came up with, and published with colleague Jon Hotchkiss in the Journal of Air Transport Management ("Experimental Test of Airplane Boarding Methods"): Board the plane back-to-front, every other row, alternating sides, first window seats, then middle seats, then aisles.
Steffen said it’s more efficient, because it spreads passengers out, allowing them to find their seats, stow their luggage and get out of the aisle with minimal crowding in the passenger cabin. He also found in his research that random boarding — “to say, ‘OK everyone, airplane’s open, go ahead and get on board’” — would also be faster than the methods airlines use today. That’s because passengers who line up in random order for boarding are not likely to be assigned adjacent seats. So they’re less likely to be jostling with other passengers for room to maneuver and space in the overhead bins, before settling down to fly."
