Hub Scheduling Comparison

[BoeingBoy]

Much has been said about the "business model" and how it contributes to or detracts from efficient use of resourses. Thought I'd put together something on the hub scheduling, comparing US to two other carriers - AA (the legacy carrier that's been the most aggressive at "depeaking" their hubs, and WN (the grandfather of LCC's).

These charts are derived from the May data from the BTS - the latest available at this time. They show the number of monthly scheduled departures during each one hour block during the day, i.e., 6:00 - 6:59, 7:00 - 7:59, etc.

Jim

 

[whatkindoffreshhell]

Nice work (hope you did get to enjoy some weekend sunshine though).


So tell us your conclusions.

 

[ITRADE]

What is your "Y" axis?????????

 

[MrAeroMan]

What is your "Y" axis?????????

It looks like it's # of flights.

 

[sfb]

What is your "Y" axis?????????

To quote BoeingBoy as well:

"the number of monthly scheduled departures during each one hour block during the day"

So, for example, US had around 950 total departures from CLT during the 0800-0859 hour scheduled for that particular month. Divide by some number between 28 and 31 to get average daily departures.

 

[BoeingBoy]

whatkindoffreshhell,

Sure did, thanks...

The BTS On Time database has all the information. It was just a matter of discarding everything but airline, departure city, and scheduled departure time, then sorting those. Then discarding all but the cities I was interested in.


ITRADE,

Y-axis is number of monthly flights in each 1 hour time block, i.e., 6:00 - 6:59, 7:00 - 7:59, etc. These are based on scheduled departure times, not actual.

X-axis is time with a data point at each 1 hour block, i.e., 1st data point is for departures scheduled from 6:00 to 6:59, second data point is from 7:00 to 7:59, etc. I couldn't figure a way to put an X-axis scale on the chart for it to be readable without the file getting to big.

Jim

 

[BoeingBoy]

MrAeroMan & sfb,

You are both correct. This data is for May, the latest available from BTS, so divide by 31 to get average daily flights.

Jim

 

[BoeingBoy]

Here is a more detailed look at PHL. This uses 15 minute time blocks.

Jim

ps - it finally hit me that I should leave out a lot of the X-axis legend and shorten them.

 

[RowUnderDCA]

Well, there would appear to be a lot of depeaking potential.

 

[BoeingBoy]

And here's a detailed look at CLT. Same as above - 15 minute intervals, all flights for month of May.

Jim

 

[mweiss]

Looks to me like CLT could cut per-unit costs of ground crew and equipment by 50-60% without doing anything but depeaking. That's huge.

Then you can factor in the increased aircraft utilization resulting from less time on the ground.

Then you can increase the total number of hours between the first and last flights of the day.

And that's just for one hub. Jim, how much time do you typically spend on the ground (excluding taxi time) when you're going through CLT?

OBTW, fabulous analysis. It didn't occur to me that you could pull that out of the BTS table until you pointed it out.

 

[BoeingBoy]

For comparison, here's DFW in 15 minute increments....

Jim

 

[BoeingBoy]

mweiss,

I assume you mean how much time the airplane spends on the ground, since we crews often have an airplane change (get off one and get on another) and/or "productivity break". Someone that works the flights from the ground perspective could answer better, but I'd guess the average plane sits on the ground an hour or a little over (for domestic flights - international could be longer).

From the way I understand it, successfully depeaking a hub requires two basic ingredients:

1) Enough "banks" of flights a day. Since the airplanes spend less time on the ground, some connections that are possible in a traditional hub operation will "mis-connect" in a depeaked hub. In other words, with a traditional hub a passenger on the last flight in still has 30 or so minutes to make a connection to the first flight out in that "bank". In a depeaked hub, that passenger might misconnect with the first flights out in that spread out "bank", so you need another flight in the next group to connect them with (preferably one of the first flights).

2) Enough flights a day to each destination. Pretty much the same rational as above (and something of a tip of the hat to those who say you need O&D traffic). If a particular destination has only 2 or 3 flights a day, the possible connecting times get too long if they can't connect with a flight in the same "bank" they arrived in. This can be somewhat alleviated by using other hubs as connecting opportunities - if CLT only has 3 flights a day to a particular destination and PHL has 3 more, the hub schedules have to be coordinated to provide 6 connecting opportunities. I understand that AMR uses DFW & ORD this way.

Jim

 

[mweiss]

I assume you mean how much time the airplane spends on the ground, since we crews often have an airplane change (get off one and get on another) and/or "productivity break".

Actually, I didn't. I knew how long the aircraft sat on the ground, but not how long you sat on the ground. The issue there is that you are still being paid in benefits (though not wages) when you're not moving a metal tube around.

Max out the allowable hours per year, and minimize the time between beginning and ending a trip. That would mean that the pilots would get as much time at home as possible, while still providing as much as possible to the company.

Even if you ignored the improvements in morale that such a change would bring, it also would reduce expenses arising from meals and hotel rooms.

 

[jimntx]

Just FYI, when we depeaked ORD, the average ground time for passengers increased by only 19 minutes. And surveys determined that passengers liked it.
When you are on that last flight in of a bank and you have 30 minutes to get off the plane and get to the first flight out in that bank and the gate is 2 concourses away, it makes for nervous.