This is pointless, because boarding time does not waste any time. When a plane lands, it has to stay in the airport for quite a bit of time in most cases - even if you made boarding more efficient, you would not save any time. Refueling, getting in the right take off slot take longer than boarding the plane.
And the question is - does this method feel better to the passengers getting on? I don't know, but I kinda doubt it.
Obviously many people would enjoy faster boarding times, but the article barely mentions the huge problems that would come from splitting up families.
When a family's 8-year-old has the window seat, they're expected to enter the plane alone and stow their bag in the overhead compartment?
The only time this is mentioned is where it says "The outcome is fairly robust in that it's relatively insensitive to deviations because of couples or families being seated together."
So now we don't just have 18 separate boarding zones, we have some significant percentage of people in 'boarding groups' with each other. Sounds like a mess.
One of the most advantageous parts of the current system is that it works and it's (probably) not the bottleneck for departure time. Any other method would no doubt cause extra problems, at least for a while, with little to no benefit for the airline.
I don't think the time taken for refueling, etc is the limiting factor. There are a couple sequences of events that need to happen (simultaneously) to get the plane ready to leave.
Sequence 1: allow arriving passengers to deplane, then clean the plane, board new passengers
Sequence 2: Refueling, other maintenance
Sequence 3: load checked baggage
The time taken to ready the plane is the maximum of the times for the 3 sequences above. It seems refueling takes less than 30 minutes (http://answers.yahoo.com/question/?qid=20061018151916AAMkss8), and I'd guess the checked luggage can be loaded within 30 minutes as well.
I'm inclined to believe that deplaning, then cleaning, then boarding new passengers takes longer than 30 minutes, and boarding passengers IS indeed a bottleneck (unless there are special circumstances like deicing at the gate).
Boarding passengers takes less than 20 minutes in most cases. Even with very fast airlines like easyjet, the plane sits longer the passengers board.
Also, you know that airports queue planes, right? So a plane can't just hurry up to fill in passengers and take off.
And in any case, why? Let's say you saved enough time to make an extra trip in a single day, then it would be worth it, but this is not going to save it. So you finish your pilots workday 20 minutes earlier, this is no saving. And if you are pushing your planes that hard anyways, there is no time for proper plane maintenance and checkups.
I've done a lot of short trips within Europe by plane, and I've never observed boarding time to be an important factor.
In 3 of the last 4 flights I've taken, boarding was the limiting factor. We left the gate within 5 minutes after the last person boarded, and were airborne shortly after without having to wait on the tarmac. On one of those flights, the plane didn't finish boarding until 10-15 minutes after it was scheduled to take off.
Curiously, the one exception was a United red-eye back from SF to Boston. And I think they may've been using this boarding method - the gate attendants said that the boarding groups were window seats first, then middle, then aisle. There certainly was no rhyme or reason to seating within boarding groups, and there were about 12 boarding groups.
But when did you start boarding? When did the planes arrive? They usually make you start boarding about 20 minutes before they are allowed to take off, so people sit in the plane the minimum amount of time.
Planes usually arrived about 40-45 minutes, sometimes up to an hour before we started boarding.
In 2 of the 3 flights, that might've been the case. But that other one, we were actually late, so I don't think they delayed boarding so we wouldn't have to wait. Actually, we started boarding about 15-20 minutes after we were scheduled to start boarding.
Cleaning seems like the real bottleneck, but assuming they can't get the flight attendants to work any faster, any time saved on boarding shortens the deplane/clean/board cycle.
If the pilot finishes 20 minutes sooner as a stable number you can send out a new pilot and keep going or reduce airspeed and save fuel. If nothing else you can have a large cushion so delays don't propagate though the day. AKA your 10 min late at 10 am and your still 10 min late at 10pm. I suspect most large airplanes fly fairly complex routes so you can consider 30 min a day as 1 / 50th of a virtual airplane if you change which route each plane takes.
> "Also, you know that airports queue planes, right?"
Yeah, but usually a queued plane is waiting on the tarmac, not at the gate, right? Fixing boarding delays may not always lead to earlier takeoffs, but it can definitely get planes pushed out of the gate faster (freeing that gate for other planes).
Also, boarding time probably depends a lot on the type of plane (ie, even more so than fueling time does). I fly trans-Atlantic every now and then -- boarding a 747 full of people who don't fly regularly almost always takes more than 30 minutes.
Maybe boarding time isn't a delay for small planes, but I'd bet it is for big planes.
If you fly transatlantic, then those planes require more than an hour on the ground before flight. And planes get queue times allocated as soon as possible and pre-planned, not just when they leave.
But it really doesn't matter. Most flights will finished boarding well before the departure time, and many will arrive late as they always do. The plane can't leave until the departure time, so unless there's a direct cost benefit, why bother?
They load so early because they have to guarantee that everything can happen in time. If you can reduce the time it takes to board, you can board passengers closer to the actual departure time. This allows you to squeeze flights closer together, which saves a ton of money.
Just to clarify, once the plane is wo passengers, its a race against time, people clean the plane at the same time its being refueled, and as soon as it is clean you can board it, and as soon as its loaded you can leave. So any minute means money.
For simplicity, his model assumes a plane with 120 passengers seated in 40 rows, each with a central aisle having three seats to the left and three seats to the right of it.
So this only works with 50% occupancy? 120/(40*(3+3))
Imagine an airplane whose ceiling opened up to reveal a completely empty cabin. These planes could be parked underneath their gates at airports. At the gate itself would be the entire cabin (seats, overhead storage, etc.) as a complete unit which could be accessed by the passengers at any time before departure. At departure time, the floor of the gate opens, the cabin is lowered into the waiting plane and the ceiling of the plane closes.
I think if the boarding area contained a large vacuum chamber, and the plane cabin were heavily pressurized before docking, explosive decompression could be used to remove all passengers, or their remains, from the cabin in a fraction of a second. A similar technique could be used to load the plane.
(Most) planes use the skin as a load bearing structure to reduce weight. Punching holes in it is therefore mostly avoided, and those holes that are required (windows, door) tend to be designed to minimize stress points.
Your proposal would require a massive increase in weight (or reduction in strength/rigidity) that would most likely offset any benefit from improved loading times.
Aviation is all about trade offs. Even when I was doing systems software any benefits from software upgrades were compared to weight/power/cooling/operating costs religiously.
This was addressed in the article. It states "The outcome is fairly robust in the sense that it's relatively insensitive to deviations from it, say, because of couples or families being seated together." During ticket purchase, families could be noted as such and all given the lowest group number amongst them.
Traveling Ryanair or other budget airlines, you get a "sequence number" which is the order in which you've checked in, theoretically this is also the order in which you board (seating is unassigned). But then you get there and they say "families traveling with children board first" and as this is an Irish airlines that's, umm, everyone but me. Which is understandable perhaps if people want to sit together, but hey, didn't I pay for my ticket too? So if you're going to have a free for all then do that, don't sneak in seating policies by the back door!
The good news is that exit rows are desirable, and children and their keepers are prohibited (in the US, anyway) from sitting in them. So they can pre-board all they want, and not take away any of the good seats.
(Of course, people are idiots. The one time I flew Southwest, I was on standby and boarded last -- the only remaining seat was an exit row. I was amazed.)
I recently flew Southwest and sat in an exit row. The flight attendant came by to give her spiel, and the 100 pound young woman sitting next to me expressed doubt that she could lift 50 pounds. I think it took about 5 minutes of pleading(!) with the rest of the completely full plane to find a volunteer to change seats with her.
shush... please don't reveal this well guarded secret... I love always coming late in the plane and still getting the exit row.
But if you start teaching other people that those are nice sits....
Airline recap: entirely focused on operations, always on the brink of collapse. Operations consist largely of optimizations problems, so lots of attention is paid to these problems (peanuts vs. pretzels). For whatever reason customers are not insulated from the volatility (pretzels, bag fees, route duration, availability, etc., in addition to price fluctuation).
Typically you see finely-tuned optimization in old and evolved markets, e.g., cola, where it makes sense, and where customers are insulated from changes in the product or service. Here though it reeks of over-optimization and misapplied attention. Nobody makes a fuss about boarding the bus. When the price of corn goes up I don't have to use a can-opener on my soda.
What's wrong with airplanes? Why don't they work like buses? Let's try to figure this out. Investors have long recognized airlines as a valuable service but a terrible business.
1.) We really like to fly but our technology is behind the curve?
2.) Irrationality of consumers has created government controls which muck too much with an already fragile system? (security lines, shoe removal, bomb sniffing, personal searches, over-training of pilots, too-expensive pilots, burdensome safety regulations, burdensome inspections, all of these perpetuated by expectation)
3.) The average passenger does not fly enough, and so he is inefficient?
4.) Dependent on fixed-___location airports? And their regulation?
5.) There's no clear cause and effect, it's simply a manifestation of the system?
6.) ?
7.) Demand for flight is very volatile but the capital which supports it is very expensive and illiquid?
(one red-herring: fuel efficiency, planes appear to be more fuel efficient than driving alone)
(another red-herring: this google hit http://www.brookings.edu/testimony/2005/0928business_morriso... states the obvious but does little to explain what makes the airline industry any different from anybody else dependent on large capital investments. sure they have to buy planes. and cola bottlers have to build plants.)
A sort of related anecdote: one time I was on a plane where there was some sort of medical emergency in the back, so we had to wait to let paramedics get on. Eventually they had the patient stabilized but needed to do more work before he could be moved, so they allowed the plane to be unloaded as long as the aisles could be cleared quickly if they needed to get through. In order to allow this, the crew only let one row at a time to stand up. I wish I would've timed how long it took, but it was definitely much shorter than the time it usually takes to get off when there is no organization.
The problem with these proposals is that the passengers' incentives aren't aligned with the airline's. I sit and read a magazine or surf the internet until everyone else has boarded, which minimizes time spent queueing, and that spent in the plane itself. (Of course, this isn't a sustainable strategy for everyone to adopt.)
It depends. If you have a big carry-on, it is in your best interests to get in the plane as quickly as possible, so you don't have to gate-check your bag. This is why first class and elites get to board before everyone else, even if they'd rather be surfing the intarwebs.
This won't work, because individuals have different incentives than the airline. The individual wants to board early so that there is plenty of space for his/her bag in the overhead bins, and also wants to board at the same time as family/friends. It's only the airline that wants the "efficient" boarding.
Now add this to the fact that zones are almost never enforced: regardless of what's on my boarding pass, I usually just go when they call "Zone 1" or
Zone 2". They've always just let me through without a comment.
The problem with boarding planes is that people completely ignore when they're supposed to board and immediately line up as soon as boarding starts.
I fly all the time and as soon as boarding starts for the last N rows more than half of the passengers line up, in a completely random fashion. It's as if people are unnerved by waiting while other people are boarding, even though everyone has a specific seat.
People would have to be trained into some form of orderly boarding before improving on the theory has a hope of helping.
Luckily, people are easily trained by enforcing the rules. If airlines actually enforced zone rules, people would learn that it's easier to obey them than to chance the hangup they would cause by a failed attempt to cheat them.
I don't think the average person flies frequently enough to really get a feel for the rules. They hear "my flight is boarding" and feel like the flight will leave without them if they don't hang around the gate. Even I get a little worried when I arrive after boarding has begun (and I fly at least once a month).
The demographic that flies frequently enough to learn the rules probably already has elite status, and gets to board first anyway.
(I imagine there is a small number of people that fly frequently, know they need to beat the rush to get their carry-on stowed, but for some reason don't have elite status. Maybe enforcing the rules would train these people -- but there's really no incentive for them to not break the rules, so they will probably try anyway.)
Ah, but the algorithm only optimizes queue'ing up the plane cabin, when in fact we should optimize the entire queue + dequeue.
I only skimmed the PDF, but I think it overlooks one side benefit from loading a plane back-to-front as airlines do today: Your carry-on bag is very likely to end up near you or in front of you, even on a very packed flight.
The benefit is, when getting off the plane, which is always done front-to-back (and is probably optimal or close to it), everyone can grab their carry-on on their way down the aisle. If your carry-on ever ends up BEHIND your seat in the aisle, then generally you must wait until EVERYONE else gets off (it happened to me once, and that's when I suddenly realized why each flight has people who seem ready to go but must wait patiently for the entire aisle to clear).
With the proposed algorithm in the PDF, if you are in an odd-numbered aisle seat, you might just have to go all the way to the back to find spare carry-on space, which is a great disaster for the deplaning part. Why is there space in the back? Because when people book their seats, my personal observation shows they tend to pick the back last... I guess they don't want to be near the bathrooms.
The suggested method isn't going to work for families and couples with adjacent seats.
If the primary impediment to rapid boarding is cargo stowage, it seems like it would make sense to focus efforts on developing an alternative carry-on cargo stowage system.
I took a coupla air trips recently, and in all four cases, the boarding took less than 30 minutes. Reducing that to 5 minutes seems essentially impossible, since some families would take more than 5 minutes by themselves.
How about doubling the number of entrances/exits on a plane: have one on the front that boards the first half of the plane and one on the back that boards the rear half of the plane? I'm sure there would be some way to "thread" passengers once they all go past ticketcheck, perhaps with a re-engineered variant of something like: Rows 1-20 to the left and Rows 21-40 to the right. Only potential families sitting in rows 20-21 would have to be separated at boarding time, but they'd meet right back up in the middle of the airplane.
Southwest has the lowest turnaround times in the industry, and they have essentially no load order (the A-B-C groups just determine who gets to pick seats first).
"it turns out that the best method (one of several more or less equivalent methods) calls for passengers in even-numbered window seats near the back of the plane to board first", followed by even middle, even front, and odd rows int same order.
I think I would be very annoyed if I arrived first and was the last to board, but the article says that its method is 6x faster than the current method.
This is a perfect example of where a computer fails to interpret the intricacies of real life.
For reasons such as groups sitting together, too many divisions, etc.
Planes have rear and front doors. Why don't we start by just splitting people into front and back and filling in from the middle - assuming this really is an issue which needs to be solved.
When I saw "load" in the title I thought cargo not people.
I think a great way to unload cargo is like a bomber, have the cargo in pods and just drop it into a special cargo zone and then let a truck transport it from there. Meanwhile the plane goes and drops more cargo and then back to the airport to reload.
I wonder about families boarding, though. It seems like the proposed solution would fall apart fairly often. Families often sit together and want to board at the same time.
(also, thank you so much for submitting the print version)
The biggest problem with this proposal is that the airline industry's only interest is finding ways to cut costs. (And that's because price is pretty much the only thing we consider when we buy tickets.)
In theory I guess it sounds good, but I don't see it working at all. This would create 18 seperate groups of passengers, and unless they're already in line in the right order, there's no way this would be faster. When they call your zone to board, it usually takes a few minutes for every person in taht zone just to get in line. Multiply that by 18 and there's no fucking way this saves any time.
I could see this working very well for Southwest, though, for instance, as they already have everyone pre-line up not only by lettered zones, but also in groups of five within those zones (eg "B 15-20") at most every Southwest gate they service.
Instead of milling around the gate in the 15 minutes before boarding, they get everyone where they're supposed to go, and then funnel them through the door single file, in order. As the article references, though, it breaks down into a clogged jam-up of humanity trying to stuff their carry-ons into the compartment the moment you get to the plane.
I'm not sure this would work very well with Southwest because they don't assign seat numbers to passengers (at least in my experience). Most of the time, the people at the front of the group choose seats toward the front of the plane and hold up everyone else while they're stowing their bag.
I've been on an airline where they line up individuals in a specific order (there's number markers along the line at the gate to make it easy). I don't know what the algorithm was, but it seemed fairly efficient.
At every airport I've ever been they are smart enough to call the next zone before the previous zone's line is half way done. And even when the line dwindles, the next zone people are all huddled around and ready to get on.
Perhaps the 3 subsections they present is too many; imagine it with 2 subsections resulting in 12 groups. Using an organized line-up system like Southwest, you could reduce the ready-time. First, groups 1 and 2 would line up. Once the group 1 line is clear, group 3 would take its place, then group 2 would load. This would also reduce the number of people per lineup, something that in small terminals still plagues the Southwest system (with their 6 effective groups).
And the question is - does this method feel better to the passengers getting on? I don't know, but I kinda doubt it.