Forgot your password?
typodupeerror
Transportation News

Airbus Issues Major A320 Recall, Threatening Global Flight Disruption (reuters.com) 58

Europe's Airbus said on Friday it was ordering immediate repairs to 6,000 of its widely used A320 family of jets in a sweeping recall affecting more than half the global fleet, threatening upheaval during the busiest travel weekend of the year in the United States and disruption worldwide. From a report: The setback appears to be among the largest recalls affecting Airbus in its 55-year history and comes weeks after the A320 overtook the Boeing 737 as the most-delivered model. At the time Airbus issued its bulletin to the plane's more than 350 operators, some 3,000 A320-family jets were in the air.

The fix mainly involves reverting to earlier software and is relatively simple, but must be carried out before the planes can fly again, other than repositioning to repair centres, according to the bulletin to airlines seen by Reuters. Airlines from the United States to South America, Europe, India and New Zealand warned the repairs could potentially cause flight delays or cancellations.

This discussion has been archived. No new comments can be posted.

Airbus Issues Major A320 Recall, Threatening Global Flight Disruption

Comments Filter:
  • by yuvcifjt ( 4161545 ) on Saturday November 29, 2025 @01:26AM (#65824133)

    ... for calling themselves out when they think there may be a major issue with their airplanes.

    Unlike another major manufacturer that hides, dithers, and delays.

    • The fix mainly involves reverting to earlier software

      Is it anything to do with undoing an AI produced PR? Or how about some more information about how a newer version has problems due to some "fix" ? How did that come about? What change was it, what drove the change, and why did an update end up being problematic instead of fixing issues? Concerned potential air travelers deserve answers. Maybe we need an FAA that instead of being in bed with air travel, acts more like the ATF, goes after the industry hard.

      • by Cochonou ( 576531 ) on Saturday November 29, 2025 @03:29AM (#65824233) Homepage
        Yes, it's quite surprising. What we know from the EAD [usersmatth...0268-e1pdf] :
        - The risk is related to solar storms (so more accurately atmospheric neutrons).
        - Most of the airplanes will undergo a software rollback to a previous version.
        - Some of the airplanes will need a hardware retrofit... but this is just because the software upload cannot be performed as easily on the affected equipment.

        So where does this lead us to ? An error in the EDAC/ECC code protecting the memory from neutrons-induced bit flips ? From a hardware perspective it is the most likely explanation, but it would be suprising to have such a bug introduced in a software update since this would be quite a low-level function which would be unlikely to be updated. A bug in the error handling code when such an error happens ? This could be more likely, since error handling structures can be shared between several components. Especially if the error in question is very rare (e.g. double error in the same word) and is not correctly tested during regression testing.
        • by Cochonou ( 576531 ) on Saturday November 29, 2025 @03:30AM (#65824237) Homepage
          I botched up the link, here is the correct one: EAD [europa.eu]
          • Re: (Score:2, Interesting)

            by saloomy ( 2817221 )
            Airplanes are connected to the internet and private networks for their airlines This is how they get diagnostics data on the ground. There should be an automatic update feature, so I imagine all they have to do is increment the version number of the old code to more advance of the new code, and presto, fixed planes. I wonder if their software releases are staggered to prevent faulty updates from causing huge issues, and do a 10% new, 90% last version for testing and validation.
            • by AmiMoJo ( 196126 ) on Saturday November 29, 2025 @05:26AM (#65824313) Homepage Journal

              I would be surprised if software updating an aircraft is that simple. It probably needs to be controlled and tested after the update, with records kept by maintenance staff, and notifications sent to pilots.

            • They aren't complete idiots, if the solution was that simple they would do it that way.

            • News I get is some big airlines performed all updates in an evening and are back to normal, but a small fraction of planes will take much longer time. I assume these are the oldest planes and they need specialised hardware for the update (e.g. floppy disks, emulators).

              • News I get is some big airlines performed all updates in an evening and are back to normal, but a small fraction of planes will take much longer time. I assume these are the oldest planes and they need specialised hardware for the update (e.g. floppy disks, emulators).

                Schiphol handles a lot of those Airbus planes every day. From the moment I heard about this immediate requirement at roughly midnight UTC Friday/Saturday I followed all the flight departures for Saturday and only about 4 flights were cancelled.

                I didn't bother to check what aircraft they were because 4 cancelled flights seems about normal on any given day for all I know. I checked several times throughout Saturday. Schiphol has many, many flights every day.

            • Being able to automatically update the software that actually moves the flight control surfaces over the internet is pretty much top on the terrorist wish list. I wouldn't fly on a plane where that was possible, and I don't think there are many pilots that would be ok with this.

              Flight control surface software should be air gapped from any kind of remote network, even private ones.

          • by Viol8 ( 599362 )

            From the directive:

            "This condition, if not corrected, could lead in the worst-case scenario to an uncommanded elevator
            movement that may result in exceeding the aircraftâ(TM)s structural capability."

            Thats a bit worrying. Surely the fuselage should be able to cope with any possible elevator input or am I being naive?

            • A little naive.

              Consider the angle of elevator deflection needed to lift the nose at takeoff(~140MPH) on a runway 8,000 feet above sea level. I have no idea what the number actually is, but let's assume it's 30 degrees. That's fine at that speed. But, a 30 degree deflection at 500MPH would create such a dramatic and sudden pitch that the G forces and wind pressure would shear the wings off. Many regard the wings coming off mid flight as being sub-optimal.

              Now, the software is designed to prevent that extreme

              • Re: (Score:2, Interesting)

                by ghinckley68 ( 590599 )

                you have no idea what you talking about.

                "Consider the angle of elevator deflection needed to lift the nose at takeoff(~140MPH) on a runway 8,000 feet above sea level. "

                there are only hand full of runways at or above 8000ft
                the defection need to get the aircraft to rotate on it main trucks is about 2-5deg. a 30deg deflection at 140kts on the runway would just slam the tail into the ground its called a tailstrike. they are design to handle this and are even tested to survive it. One test a airplane has to pass

                • I don't understand anything of that, so what is your proposed scenario for "uncommanded elevator movement that may result in exceeding the aircraft's structural capability."?

                • by caveat ( 26803 ) on Saturday November 29, 2025 @12:40PM (#65824765)

                  there is no control on an aircraft that if it goes to max deflection capable of breaking an airframe. The worst control surface going hard over is the ruder. it will make the plane yaw to one side them pitch over.

                  See American Flight 587 [wikipedia.org]; also Maneuvering Speed [wikipedia.org].

                • there are only hand full of runways at or above 8000ft

                  Thanks for backing me up and confirming that they exist.

                  the defection need to get the aircraft to rotate on it main trucks is about 2-5deg.

                  Thanks for letting me know. I did say that I had no idea what the deflection was. I was simply trying to illustrate the point that 30 degree deflection is available and perhaps required under certain circumstances.

                  there is no control on an aircraft that if it goes to max deflection capable of breaking an airframe.

                  Then why do manufacturers have maneuvering speed limits and why have we had planes crash where the wings or other control surfaces shear off due to improper control movements?

                  airbus can not do 500mph TAS its max speed is something like 240 TAS

                  True Air Speed for the A320 at 30,000 feet is typically and most de

        • Sounds like a problem with a SEU (single event upset) caused by gamma rays (not neutrons). It's unlikely the ECC failed since it'd either correct or report an error (e.g. SECDED memory will correct single errors and report double errors), more likely a bit flip in uncorrected memory or a signal line triggered a new code path that wasn't present in the older code. Since SEUs can occur at any time and any location you can design against them but can't always anticipate every possibility.
    • by Anonymous Coward

      This was after 15 were already injured due to this issue [cbsnews.com].

      • Looks like they got lucky.

      • by Viol8 ( 599362 )

        That was the incident that triggered it apparently. Tbh a 100 foot drop in altitude in 7 seconds is hardly divebombing territory.

        • Injuries were reported but yeah that doesn't sound too bad. I know a guy who was on China Airlines 006 (he was on tv in the Mayday episode "Panic over the Pacific"), he's told the story to me a few times, said the inside of the cabin was like someone had made a shaken salad (blood everywhere).
    • by gweihir ( 88907 )

      Indeed. Only one of the two major civilian airplane makers is not ok with blood on their hands.

  • Brooo (Score:4, Insightful)

    by locater16 ( 2326718 ) on Saturday November 29, 2025 @01:30AM (#65824135)
    Just do the Boeing thing and let the planes crash brooooo
  • They found out somebody accidentally installed Boeing bolts.

  • by Anonymous Coward
    “Intense solar radiation may corrupt data critical to the functioning of flight controls,” Airbus stated, adding that “a significant number of A320 Family aircraft currently in-service” may be affected.
    Replacing the software will take “a few hours” on most planes but for some 1000 aircraft, the process “will take weeks”, a source close to the issue told AFP."
    • They're not using ECC components on aircraft?
      • Maybe someone should do an experiment with something like RF Safestop on a computer with ECC memory and see how it performs? Military/aerospace would have more powerful equipment along the same lines and could simulate a more realistic situation. At one end of the spectrum, EMP from nuclear weapons supposedly knocks out electronics. Where's the line? RF Safestop can blast a directed energy beam at a car computer and cause it to malfunction, how much more insulated or secure is a planes flight controller?
      • Sure sounds like it, or the version of ECC was not sufficient. Also, silicon-based chips are more susceptible to radiation than GaAs, which is why GaAs was used in the early space program and maybe still today.

  • Not saying it is easy, because it really is not (despite how easy it looks when your PC, or your phone, or you Tesla performs an over-the-air update), but it might become a real selling point to allow a remote technician to over-the-air update (or downgrade) planes firmware. Sure, there will be exceptions where this will not apply, but it would save their customers money (or does Airbus pay for all those repositioning flight and lost revenue?). There ought be a way to design a process to do this safely with
    • by Viol8 ( 599362 )

      Yeah, great idea. Do a remote OTA update on a plane about to carry 200 passengers to 40K feet without bothering to test it first. What could possibly go wrong.

      Do you understand the difference in safety requirements between a car and an airliner?

      • Yeah, great idea. Do a remote OTA update on a plane about to carry 200 passengers to 40K feet without bothering to test it first. What could possibly go wrong.

        When you put it that way, maybe air travel should be banned for safety? Or at least severely limited (there's no constitutional right to plane flying). Have you ever looked around the underbelly of a airliner? A spaghetti mess of wiring, fuel, and hydraulic hoses. A lot has to not go wrong for the plane to take off, fly and land safely.

        • by Viol8 ( 599362 )

          Hardware can be physically inspected, software once installed cannot. All you can do is test it.

      • What do you think happens when they upgrade the firmware at the service facility? Do you think they have test pilots with parachutes and chase planes following take it up to test it over some range where if they were to ditch the plane it would not hurt anyone?

        As for a difference between cars and airliner, well, OTA being applied to Teslas for example will update maybe 100K cars in one night, perhaps as may a million (if not first deployment wave), so I guess a lot more car passengers are exposed to its m
  • by zmollusc ( 763634 ) on Saturday November 29, 2025 @04:10AM (#65824263)

    The issue is a software thing, that is the easiest thing to conceal and misdirect.
    Why can't they just falsify the update record, then if planes crash and people die simply shift the blame to the pilots? I thought Airbus were trying to be a major player in modern aerospace? So unprofessional. Sad.

    • The fact that this defect managed to proliferate through their entire fleet of 6000+ aircraft sold to major carriers ought to show you that they're well on their way!

  • Did a cosmic ray flip one of the 20 millions cells in their Excel spreadsheet ?
  • by test321 ( 8891681 ) on Saturday November 29, 2025 @08:55AM (#65824451)

    "This emergency action stems from JetBlue Flight 1230, which experienced a sudden altitude loss on October 30, 2025, resulting in at least 15 hospitalizations and triggering global regulatory scrutiny.

    On October 30, 2025, JetBlue A320-200 N605JB operating Flight B6-1230 from Cancun to Newark encountered an uncommanded pitch-down event while cruising at FL350, approximately 70 nm southwest of Tampa.

    * The aircraft dropped roughly 100 feet in about seven seconds
    * The autopilot remained engaged throughout
    * The crew diverted to Tampa International Airport for an emergency landing
    * At least 15 passengers, including children, were hospitalized with non-life-threatening injuries

    Although the aircraft remained controllable, the event revealed a critical failure mode inside one of the flight-control computers."

    https://aeropeep.com/easa-issu... [aeropeep.com]

  • I read last night that some airlines had already finished their software updates yesterday and are flying today as normal.
    This will be over soon, but of course it is not fun for those who had their flights right in the middle of it.

  • The aircraft owners are doing the software change and most should be done within a few days.

    • The aircraft owners are doing the software change and most should be done within a few days.

      The software change is reportedly to roll back to a previous version of the software (which did not have this particular issue, although the updated version was presumably introduced to fix some other issue).

      Most airplanes already require regular daily checks (think of it as checking the oil levels, but there is a lot more to check on an airplane than the fluid levels) usually performed at night when the planes are not being flown. Adding a few hours of software rollback to an aircraft's daily checks is

  • by Mirnotoriety ( 10462951 ) on Saturday November 29, 2025 @04:29PM (#65825101)
    ELAC L104 Role:

    The Elevator Aileron Computer (ELAC) software version L104 on JetBlue Flight 1230 (A320-200, registration N605JB) malfunctioned due to solar radiation corrupting flight control data, causing the uncommanded pitch-down and rapid descent from FL350 near Tampa on October 30, 2025. This system processes pilot inputs to control elevators (pitch) and ailerons (roll).

    L103 Immunity:

    ELAC L103 resists solar-induced data corruption through more robust fault tolerance and data handling that prevents bit flips from triggering unsafe commands, unlike L104's vulnerable processing changes.

    Upgrade Purpose:

    Airbus released L104 under its "Safety Beyond Standard" program to bolster A320 safety with enhanced flight-envelope protections, including pitch attitude limits in alternate law to avert stalls and better failure recovery, mirroring A350 features.
  • Major recall is OK, as long as it's not Total Recall.
    They can always try to turn them off and on again.

  • I find it interesting anyone would offload error detection and correction to application software. Not only are you needlessly increasing local complexity in doing that any possible machinery you implement to accomplish this in software is itself subject to failure from same sources of arbitrary corruption.

    Why would someone do this instead of using hardware with some sort of RAS with memory mirroring, pool scrubbing, multi-bit error correction...etc? If you are extra paranoid just add more memory and or C

We all agree on the necessity of compromise. We just can't agree on when it's necessary to compromise. -- Larry Wall

Working...