Several days ago, I wrote about a strange occurrence, whereby two American Airlines plane ended up on the same frequency with the same callsign. That’s problematic, since air traffic controllers identify planes with their callsigns, and they’re intended to be distinct.
Well, as it turns out, exactly the same thing happened on another flight a day earlier, and it sure seems like something should be done to address this! There’s a good reason American uses the same callsign for multiple flights, but there’s no good reason for this confusion with ATC communication.
In this post:
Two AA5083 flights end up on same ATC frequency
Every flight has a callsign that’s used with air traffic control, and for commercial aircraft, that’s generally going to be the generic airline callsign plus the flight number. Callsigns are of course unique to a particular flight, because that’s the way that air traffic control communicates with pilots.
That brings us to what happened on Wednesday, August, 12, 2026, involving flight AA5083. For context, this was a flight operated by PSA Airlines, which is a wholly owned subsidiary of American, operating regional jets under the American Eagle branding. Note that while there’s an AA flight number, the callsign is “Bluestreak” rather than “American” (since PSA has its own air operator certificate).
AA5083 is the flight number that American uses between Philadelphia (PHL) and Providence (PVD), and it’s actually a flight number used for both directions of travel. Typically, the same CRJ-700 would operate the flight in both directions, but that’s not always going to be the case, especially if there’s a delay.
That brings us to the flight on August 12, where the inbound flight to Providence was over an hour behind schedule, while the outbound flight to Philadelphia departed on schedule (early, actually).
Below you can see the VASAviation video about how this played out, including a visualization of the path of both planes, as well as the air traffic control audio. As you can see, it’s not just that two planes were in the air at the same time with the same callsign, but they basically flew right past one another.
All parties did a great job handling this situation. The air traffic controller advised the pilots of the duplicate callsign, and made it clear which plane was “outbound” and “inbound.”
Why do some airlines use the same flight numbers for multiple flights?
I think one logical question is why some airlines use the same flight number for multiple flights, including those to and from a destination.
It’s true that this is mostly an issue in the United States, but that’s for a simple reason — a few airlines in the United States are absolutely massive.
Flight numbers are generally limited to four digits (it’s largely an industry equipment limitation, and it’s something that the industry doesn’t consider to be worth addressing), and an airline like American can operate 6,000-7,000 flights per day. The catch is that not all of those flights numbers can be used.
Generally, flight numbers 1 through 2949 are used for mainline flights, 2950 through 6099 are used for regional flights, 9000 through 9999 are used for ferrying and other repositioning flights, and that doesn’t even begin to address codeshare flights. So that explains why we have this issue.
However, what’s supposed to happen is that if there are multiple flights in the air with the same flight number, you’d add a letter to the end of the callsign for that second flight in the air, to differentiate the two.
On my last post about a similar incident, I posed the question of who is responsible for this mistake, and what went wrong. We got a really insightful answer from steve64, which I figure is worth sharing here (note that this is in reference to the previous incident, but the same principle applies here):
I used to work DFW Ops for legacy AA many years ago.
The Dispatcher for the PHX-ORD should’ve “stubbed” that flight. In the Reservation system, nothing changes. But in the Flight Planning System (and thus the Flight Plan sent to ATC) the flight number would be altered.
A 3 digit flight number would simply have a “P” (pronounced “Papa”) suffix added. The system (both AA’s and ATC/FAA) can’t handle more than 4 characters. So on a 4 digit number, the 1st digit gets dropped. 2482 stubbed would be 482P. Double/triple subs can happen. The next letters used are (I think) Romeo & Sierra.
The concept for a “stub” was originally for a flight passing thru a hub, back in the days when flights were actually scheduled “thru.” The inbound would be late but the hub has a spare plane so the outbound could depart on-time as a stub. Now that flight number philosophy has changed (the number doesn’t go “thru” a hub anymore, it’s ‘”out & back,” on return, the plane will go thru the hub but changes flight numbers), stub really apply more at a spoke than the hub.
I think that “stubs are now at the spoke instead of the hub” concept is what tripped the Dispatcher. While PHX is a hub, operationally for flight 2482, ORD is the hub and PHX is the spoke. If the original reason for the plane swap wasn’t due to the inbound being late AND the Dispatcher was thinking of PHX as a hub, the he/she wouldn’t be tracking the inbound plane thus not realizing it had lost time.
Having said that, I can’t believe neither AA nor FAA’s system wew able to catch this scenario developing.
Bottom line
Two American Airlines regional jets with the exact same callsign ended up passing one another in the sky. The flight number is used for both directions of travel between Philadelphia and Providence, but in this case, the outbound flight was delayed, while the return flight operated on time, obviously with a different aircraft.
The reason for the duplicate use of flight numbers is the limitation of flight numbers being at most four digits, and the big US carriers having many thousands of flights per day, and also needing flight numbers for codeshares, etc. That’s not an issue in and of itself, but instead, the problem is that the dispatchers seemingly didn’t update flight plans with air traffic control in light of this. The fact that this happened twice in two days isn’t good.
I imagine we’ll see American make some sort of change in its system to prevent this in the future.
What do you make of this duplicate flight number situation?
What would be even more interesting if ATC or even the actual aircraft got sent the wrong flight plan....reminds me of when I signed on to AA web site one day and noticed the usual couple of flights being touted...something like ORD-SAN for $400 and PHL-MIA for $333. The more interesting one was DFW-DFW for $299!! I bet DAL could bet that price!!!
I didn’t think this was possible. At my airline, we use “radio callsigns” when a flight inbound to a hub is late, the continuous of the flight number uses a different callsign to avoid this problem of potential confusion with catastrophic consequences possible. I didn’t even think ATC would even allow this in their system.
What surprises me most about this is that this happened under the same operator within the same airspace.
Within Eurocontrol, you cannot have identical callsigns within the same operating window. Such flight plans get rejected and you must add a suffix (as per the extract shared in the article). This happens often for re-departures or long delays due to weather disruption, e.g., departure on the next day. Usually such flight plans would have "D" added...
What surprises me most about this is that this happened under the same operator within the same airspace.
Within Eurocontrol, you cannot have identical callsigns within the same operating window. Such flight plans get rejected and you must add a suffix (as per the extract shared in the article). This happens often for re-departures or long delays due to weather disruption, e.g., departure on the next day. Usually such flight plans would have "D" added as a suffix. In addition, Eurocontrol Network Manager (NM) facilitates checks for callsign confusion. If those are detected, typically, the operator would re-file the flight plan. At least in the airlines I've worked for, we would always run internal checks for callsign confusion risk as well.
Surprised that his could happen within the US airspace....
This is an important distinction. The Eurocontrol's system checks not only for identical callsigns but also similar ones. I don't understand how FAA can not only not have that, but even permit duplicity.
Also, under Eurocontrol it's not always a suffix. The callsign may be completely unrelated to the flight number if needed to avoid confusion - for example AF7435 flies as AFR33NA (AIRFRANS THREE THREE NOVEMBER ALFA) today. And it took me less than...
This is an important distinction. The Eurocontrol's system checks not only for identical callsigns but also similar ones. I don't understand how FAA can not only not have that, but even permit duplicity.
Also, under Eurocontrol it's not always a suffix. The callsign may be completely unrelated to the flight number if needed to avoid confusion - for example AF7435 flies as AFR33NA (AIRFRANS THREE THREE NOVEMBER ALFA) today. And it took me less than 30 seconds to find this example on FR24. There really isn't any reason why callsign has to be related to the fly number, that's just a convention.
American uses Equipment Coordinators at their hubs and they initiate aircraft swaps to try and keep things on time. The EQPT Coordinator should have told the dispatcher, responsible for the flight, to stub the 2nd leg of the flight number and then the EQPT coordinator would assign a new aircraft to the flight leg. Then, depending on the time, the dispatcher might have to contact ATC and cancel the existing flight plan and then refile...
American uses Equipment Coordinators at their hubs and they initiate aircraft swaps to try and keep things on time. The EQPT Coordinator should have told the dispatcher, responsible for the flight, to stub the 2nd leg of the flight number and then the EQPT coordinator would assign a new aircraft to the flight leg. Then, depending on the time, the dispatcher might have to contact ATC and cancel the existing flight plan and then refile with the new flight number. American’s old flight planning system sent alerts about overlapping times and duplicate flight plans in the system, but they have moved to a new system that may not run the same checks or send the same alerts. The old system is still there but has been pushed into the background. Many of their newer dispatchers might not be all that familiar with the older system. I’m sure American will find and fix the issue.
Why isn't the system extended to add another digit? We clearly have too many flights for only 4 digits...
We actually don’t the flight number issue is related to airlines forced taxonomy of light numbers. Using four digit numbers created the opportunity for 9,999 unique flight numbers. On any given day, AA operates, at most 7,000 flights.
Issue is numbers are delineated by things like mainline, regional, charter, etc. While perhaps convenient for the airlines, it’s beginning to create a safety risk and should be looked at to see if changes are necessary.
The happens more often than you would think. ATC just adds a letter to the callsign to differentiate the two. AAL241 and AAL241A.
Explainable only if a dim wankpuffin assigned the same number to multiple flights .
The FAA needs to address this issue. Far too many times has the system waited for a catastrophe to happen before taking action. Two similar events in one week is flashing a huge warning sign that best not be ignored.
“The FAA needs to address this issue.” LMAO. They’re too busy hunting down DEI bogeymen. Doesn’t help their boss, Chump, is anti reason and logic. He only cares about stealing…and PDF activities.
Delta would never let this happen, we are above such nonsense
Amending a previous comment, now seeing mainline up to at least 3334:
https://www.flightaware.com/live/flight/AAL3334
I thought AA changed their numbering bands so mainline goes up to 3139 now?
Per Gary: https://viewfromthewing.com/american-airlines-has-too-many-flights-for-its-flight-numbers-so-theyre-making-a-change-roundup/
Twice in a week, flirting with disaster.
Why is this possible? Even if they really really have to use the same flight number, why don't they use pseudo callsigns like we do in Europe where for example AF1181 I'm on right now flies as AIRFRANS 54 ECHO FOXTROT (AFR54EF), and this is done everytime there's a threat of similar callsigns, let alone identical. It's such an easy solution for such a pointless problem...
Who's on first, What's on second, and I Don't Know's on third.
Oh, you didn't catch American's latest rebrand, now it's 'Steve Urkel' Airlines... "Did I do that?"
since PSA runs its own operations and presumably has its own flight tracking systems, this likely says that they don't have the safeguards to prevent a flight from departing while another under the same flight number is still in the air.
Ah, so this is actually one of those 'it's PSA, not American' things, huh? (Gotta say, Republic's "Brickyard" call-sign is still one of the 'best' names out there, other than Speedbird and Shamrock.)
This happened to AA mainline this month as well.
Presumably PSA, being a wholly-owned subsidiary, is running the same dispatch processes that AA mainline is.
American has a problem.
This feels like another example of where the pursuit of corporate efficiency leads to inefficiency.
I think we all know on which airline this would not happen
To be fair, most airlines have rules in place that don't allow for the same flight number to operate on the same day unless it's the same aircraft flying it, to avoid this happening.
To be fair, I'd imagine in both recent scenarios, the SCHEDULE was to have the same aircraft operate both flight legs. But OPERATIONAL ADJUSTMENTS for these flights lead to the outbound being a different aircraft from the inbound.
Airlines have always had the same flight number for multiple segments.
Well, it only happens in the USA, not in any other country. Other countries with dense traffic all use alphanumeric callsigns.