The Astrophysicist Who Has a Better Way to Board Airplanes

What is the best way to board airplanes?

He’s not the only person trying to answer the question. The most effective way for people to fill a metal tube is a problem that airlines have been trying to solve for decades.

For companies that squeeze every drop of profit from their operations, shaving two minutes of boarding time per flight can save countless millions of dollars a year, so of course they’re obsessed with smoothing out this form of turbulence. Every march down the aisle of a plane is a parade of inefficiency. But the only thing as difficult for airlines as making planes efficient at 35,000 feet is doing it while they’re on the ground. They’re constantly looking for a less awful way, which is why United Airlines is tweaking its boarding process—again.

United recently implemented a system called “Wilma,” a rough acronym for its new boarding chronology: window, middle, aisle. Except it’s not new. In fact, the airline boarded coach passengers this way until 2017. The “L” doesn’t stand for anything, either. But the really odd part about this supposedly better, faster way of boarding a plane is that it could be even faster and better.

Just ask Steffen, an associate professor of physics at the University of Nevada, Las Vegas, who developed what he says is the optimal boarding strategy and published his findings 15 years ago.

Steffen started thinking about the boarding process only because he couldn’t stop thinking about how annoying it was. You don’t have to be an astrophysicist to understand his frustration, which began with one unpleasant trip to the airport in Seattle. “I had to wait in line in traffic. And then I had to wait in line at security. And then I had to wait in line to check my ticket at the gate. And then I had to wait in line at the jet bridge,” said the 48-year-old scientist. “I thought that last line was probably unnecessary.”

That thought nagged at him for a few years before he decided that he should forget about it or do something about it.


View Full Image

‘There should be a solution to this,’ Jason Steffen said of slow airplane boarding. PHOTO: JOSH HAWKINS/UNLV

“One of the reasons I allowed myself to be entertained by the idea was that it seemed like a tractable problem,” he said. “There should be a solution to this—no matter how weird it might be.”

As it turned out, there was a solution, and it was definitely weird.

He suspected the worst way to board a plane was front to back, but he was surprised to discover the next-worst way was back to front. That had been the standard for decades, but it was quicker to just board randomly. This counterintuitive finding made him even more curious about the best way, so he applied the computing methods he uses for his astrophysics work and coded an optimization algorithm.

That led him to the Steffen method.

Here’s how it works. The first person to board a single-aisle jet like a Boeing 737 is the passenger in the window seat of the last row. Say that’s 30A. The next person would be exactly two rows away in 28A, followed by 26A, 24A and 22A until the window seats in even rows on the right side were full. Next are the window seats in even rows on the left side: 30F, 28F, 26F and so on. Then come window seats in odd rows on the right and left starting from the back. The same patterns apply to middle seats and aisle seats until the last person on board plops into the front row. That’s just one permutation. There are others that would achieve identical results, he says.

The idea behind spacing out passengers in alternating rows is to reduce the probability of traffic jams. If the primary bottleneck of the boarding process is people waiting in the aisle, mostly because of how long it takes for others to load their luggage, Steffen’s fix maximizes the number of passengers stuffing their bags into overhead bins simultaneously. It takes a serial process (one at a time) and makes it parallel (several at a time).

Once his study was published in 2008, he tested the results for a 2011 paper. His laboratory was a Los Angeles soundstage. His subjects were volunteers and Hollywood extras. They boarded a mock Boeing 757 using five techniques—and the Steffen method was easily the fastest.

If only getting a line of people to sit down were that simple.

Steffen’s boarding system may be the most effective, but it’s not the most practical.

It doesn’t account for multiple travelers, families sitting together or what he calls “other effects of human nature.” Airlines don’t have the luxury of ignoring human nature. They are stuck with “a logistical puzzle as well as a psychological test as they try to balance speed, fairness and revenue,” as my colleague Alison Sider recently put it. There are too many confounding variables that mess with his algorithm—like frequent fliers who expect priority boarding, regardless of where they’re sitting. So there is a better chance of the Wi-Fi holding up for an entire flight than the Steffen method being adopted.

But just because they don’t board in the most efficient way doesn’t mean they can’t be more efficient.

Steffen’s academic theories might be too idealistic to overcome the realities of the business. It’s still worth exploring why they work—and how the perfectly optimal strategy can improve methods that adjust for our imperfections.

(Graphic: WSJ)

View Full Image

(Graphic: WSJ)

Steffen himself devised a more pragmatic version with families in mind: even rows on the left and right, odd rows on the left and odd rows on the right. It still beats most boarding strategies, he says. As it happens, so does Wilma, since it spreads out passengers and lets more of them stow their luggage at once. The outside-in system ranked highly in his testing—slower than the Steffen method but faster than procedures not named for astrophysicists.

United disagrees. Before deciding to bring Wilma back, the airline conducted its own experiments with paying customers on real flights, not Hollywood extras. A spokesman said the company tried several boarding strategies, including Steffen’s, and Wilma was the fastest.

But the purest form of Wilma was too radical for United, which is using a modified version of window-middle-aisle. It’s similar, though slightly and noticeably different, like Wilma Flintstone going blonde. The airline is still preboarding groups (families with young children, active-duty military members, passengers who require assistance, top Premier members), and then boarding Group 1 (first class and elite fliers) and Group 2 (lower-status tiers). The plane might be half-full by the time the gate agent calls Group 3 for windows, Group 4 for middles and Group 5 for aisles.

It may be an improvement, but Steffen says Wilma is still inferior to Southwest’s system, in which passengers line up by alphabetical boarding groups and numerical positions before choosing their own seats.(Southwest now employs a swarm theorist to model boarding scenarios.)

But the Steffen method, Wilma and Southwest’s approach have something in common: They show how there is usually a better way of doing something than the way it has always been done.

Why are we doing it this way? Could we be doing it another way? Should we? Those are the questions that drive progress in any business.

I had another question for Steffen: What would he do if I waved a magic wand and gave him the power to make all airlines board the same way? His answer surprised me. He said he would try to figure out why they have not adopted his strategy.

“I’ve never run an airline company, so it would be a bit presumptive for me to mandate that you should do this,” he said. “The main thing that I would do is see what I’m missing. I’m not really an insider, and it’s easy for me to armchair-quarterback what they should be doing when there’s probably a lot more stuff going on that I don’t know about. I would ask for more information. The boarding process is easy to understand, but how it translates into the rest of the system is interesting and ultimately the most important question.”

He sounded like someone who spends most of his time looking at the stars: He wanted to know everything he didn’t know.

“I think it’s a lot more complicated than people realize to run a multibillion-dollar airline,” the astrophysicist said.

Write to Ben Cohen at [email protected]