You may remember that late in 2013, the new management team at American decided the time had come to rebank its big connecting hubs in Chicago, Dallas, and Miami. Miami went first and things went pretty well. But the big test was at the end of March when the more complex changes in Chicago and Dallas went into effect.
I spoke with both Ilhan Ince, Managing Director of Operations Planning and Performance, and Kerry Philipovitch, SVP of Customer Experience, about how the change was executed. Then I looked at the data to see that sure enough, the operation is running better than before.
More than a decade ago, American got the idea that unbanking was a good plan. A traditional hub had a bunch of planes landing, then a little ground time, and a bunch of planes leaving soon after. This was great for travelers because it meant shorter connecting times, but it also meant the airline needed to staff and have gates for the peaks that would go unused during the downtime. So American came up with a cost saving plan to smooth out the hub operation and spread arrivals and departures more evenly. It saved money and improved utilization of assets, but it made for a worse schedule for customers.
The US Airways management team had learned long ago that the revenue benefits of a true banked hub outweighed the cost savings of a rolling hub, and they brought that strategy to the new American. Miami went first, and things went well, but Chicago and Dallas were more complex. I was curious if a banked operation would cause big issues.
As the dominant airline at DFW by far, American had fewer constraints here than in Chicago. From back in the day when American had banks here, the team realized it had the facilities it needed. The biggest issue was runway and airspace constraints.
With the existing schedule, DFW had been operating pretty simply. It generally used the two parallel runways on each side of the terminal complex with west operations on the west side and east operations on the east side. The other two runways went largely unused. In a banked schedule, this had to change. So American worked with air traffic control to implement a system that would use every runway.
In this new schedule, American had some east-heavy banks. The airline realized that if it kept operating as it had, there would be queues of 20+ airplanes waiting to go east with no issues going west. So the team put together a plan that would mix the runways to manage capacity better. Combining this with a lot of hard work by the hub team to make sure that the operation would run right, and DFW was ready.
The change happened on March 29. Here’s a look at March numbers until the changeover and then a few weeks after. Note that the completion and D0 numbers are for departures from DFW while the A14 numbers are for arrivals into DFW.
As you can see, on-time performance has improved across the board. In general, departing aircraft have longer scheduled flight times because they can take a little longer to get out in a true banked hub, but airplanes are still pushing back on time more often, regardless of scheduled duration. What is also significant is that airplanes now tend to sit in the destination a bit longer so they can be timed to come back at the perfect time for the hub. That may cost a little more, but it means there’s more time to recover when airplanes are delayed.
These numbers would actually be even better if not for the nasty thunderstorms that snarled operations in Dallas in April on a few days. So far, so good. But what about Chicago? Now THAT was a challenge.
As we all know, O’Hare is not exactly an empty airport. It has long had capacity issues though recent runways additions have alleviated that. Still, going to a straight banked schedule had the potential to really cause problems. So American started by building a schedule to do exactly that. It was underperforming in Chicago anyway, so it ran simulations to create what it thought would be best for improving revenue. Then it worked backwards to figure out what could be accommodated at the facility.
Then something funny happened. While American was hard at work modeling based on previous schedules, United decided to do the same thing. And it not only made the decision but it published the schedule right away, throwing American’s models off.
American knew this was going to be ugly if it went as planned. United alone had some banks where it scheduled 45 departures in 15 minutes. Clearly this was going to result in a mess at certain times during the day.
Of course, American couldn’t go to United to coordinate schedules since that’s highly illegal. Instead, it went straight to the FAA and explained the situation. FAA then decided to work on the issue to make sure that Chicago wouldn’t end up in total gridlock.
The end result is that both American and United independently, but with the help of the FAA, decided to make changes to their schedules to ensure the airport would keep operating. United, for its part, smoothed out when it would have flights departing and arriving a bit. And both American and United shifted times on some of their banks. Because of all this, American actually hasn’t fully implemented its new schedule yet. It ran out of time to get it out there. By June, it should be fully implemented, but it’s already substantially in place.
So how are things going?
Once again, it’s green across the board. The airline couldn’t put into place its ideal schedule but it should be better for travelers than it was before by far. And American will make more money with this schedule. If American can continue to run flights on time more often, then everyone wins.
We can take a look at this again a year from now, I suppose. But so far, it’s looking good.