As the great Glenn Frey once said… the heat is on. Yeah, yeah, it’s summer and it’s hot, and yes, the pressure is rising on airlines in general, but that’s not what I’m talking about. I’m talking about American’s new tool called HEAT, which is short for Hub Efficiency Analytics Tool. Since mid-April, American has been deploying this to help reduce delays and cancellations, and I spoke with Steve Olson, Managing Director of IOC Operations for the airline to learn more.
It’s funny, because when the Cranky Concierge team descended on Dallas/Fort Worth in late April, Steve gave us all a tour of the intergration operations center (IOC) and told us about this new HEAT tool they had been developing for 3 years. At the time, I had to keep it off the record since it had only been used a couple of times at that point. Now with more experience under their belts, the airline is ready to talk.
The idea is fairly simple. If there is weather heading toward a hub, American can systemically pick apart its operation, spreading operations out and delaying as needed, to keep the hub structure functioning with minimal cancellations and strandings. That may sound like a simple idea, but it is unsurprisingly complex to actually implement.
HEAT starts from the IOC level. A team of “highly-trained” coordinators partner with all the different groups in the IOC… air traffic control, meterologists, etc to determine if HEAT will be a viable option. Key decision points include:
- Is there a high confidence of the weather issue actually happening? If not, then you don’t want to move your entire operation around. This has to be implemented further out, so a pop-up storm like you’ll find often in Miami won’t get the HEAT.
- If the weather is coming, is there room to move the bank? The idea is to take advantage of the valleys that do exist between banks at the hub. If they can move some flights into those valleys, they’ll be able to accommodate those airplanes at the airport without disrupting more of the schedule.
- Is more weather expected later? Pushing back a bank might push other banks right into weather that’s predicted later on in the day, so that wouldn’t be a great plan. There has to be a clear weather window.
If they have all these components, the IOC will then enter specifically what the operational capacity rate will be. This isn’t just an overall number, but it will look at both arrival and departure rates. Once that has been set, then, well, then the HEAT is really on.
They system looks at a variety of different factors to come up with the right solution. It starts with the all-important crew rest and availability so that it doesn’t delay a flight that will then cause the crew to time-out. Then it looks at passenger connections to try to preserve as many as it possibly can. Since this system can literally delay an entire bank or parts of it, it will take into account those connections as it chooses which flights to switch and to when.
American coordinates this closely with the FAA, both at the command center and airport levels. In fact, American says this can eliminate the need for ground delay programs in bad weather at some airports where it dominates, because American can set operations to match available arrival rates. That’s great for everyone using those airports and for the FAA.
Once this has all been figured out, the plan goes into action. The first time this was deployed was on April 12 when late afternoon thunderstorms were predicted to roll through Dallas/Fort Worth. After this was done, they scanned the archives to find other days that had had similar weather, a similar-sized operation, similar forecast accuracy, and even general airspace weather similarity. They found two days that matched well, and what did it show?
The use of HEAT reduced cancellations by 80 percent while the number of diversions dropped by 60 percent. While Steve didn’t have passenger numbers he could share, he said that reservations and airport teams have said it is a “vastly better” customer experience. If anyone reading this has experience working an event with and without HEAT, let’s hear about it in the comments.
So far, HEAT has been used 15 times since mid-April when it rolled out, and the results have been promising. The day I spoke with Steve, he said it was being used in 3 hubs that day alone. What I’m curious about is just how often it will be used… and where.
This was designed for American’s largest hubs in Dallas/Fort Worth and Charlotte. Those are hubs that tend to get severe weather and may have a higher predictability than other hubs.
They were using it in Miami the day we spoke, because it was a rare day of predictability. Miami is really a mixed bag. Tropical Storm Alex was approaching so they could use it, but Miami so often has pop-up thunderstorms that it’s hard to predict the impact. (Phoenix during the summer monsoon season would be similar.) On the other hand, Miami has so many international operations that ground delay programs are less effective than at a domestic hub, so it doesn’t take much for this system to provide benefits there.
Further north, there would be more promise in a place like Philly, especially in the winter. The big problem there is that the airspace in the northeast is so crowded that American can’t control the situation as well as it might like on its own. But this would allow American to, for example, set a departure rate in Philly based on the capacity of the de-icing equipment. At the very least it could reduce long sits on airplanes.
Then there are places like Chicago and New York which are unlikely to get much use out of this. American is far from dominant in either place and both have very crowded runways and airspace. American can’t implement this in a vacuum like that. And of course slot-controlled airports like JFK and LaGuardia or Washington/National make that really hard to solve as well.
Despite the limitations, the idea of HEAT is a good one if used properly. Of course, there’s always the risk that American could up and move a bank and then find the weather doesn’t materialize. That’s always a risk but it’s still far better than just having mass cancellations and missed connection.