• ggm 18 hours ago |
    I didn't know that was possible. Well sure, almost anything is possible but, this begs questions: how often does this happen, and is there a formal check at some granularity outside of crash events, to check if a black box recorder is working? What's the failure rate?

    I'm not pushing conspiracy theory, I'm just a bit aghast the mechanism designed to log things for disaster analysis itself can have .. catastrophic failure before an event.

    4 minutes is an eternity. This can't be down to buffer behaviour and the event itself surely?

    https://www.pprune.org/accidents-close-calls/663324-jeju-737... has discussion. Yes, catastrophic loss of power takes them off-line. Some kind of UPS like capacity seems necessary.

    • phire 18 hours ago |
      Because "the event" was the bird strike, and 4 minutes before the crash is the right timeframe for that. At least one of the engines was hit, and they appear to have lost AC power.

      The crash on the second landing attempt is just a the conscience of the bird strike, not the actual event.

      Apparently this aircraft was just old enough that the black boxes weren't required to be hooked up to any of the redundant power buses.

      • timewizard 17 hours ago |
        > be hooked up to any of the redundant power buses.

        There are multiple busses, but you almost always have an option to tie the two busses together, so one generator can drive both sets of loads. You can also add a battery through an inverter to carry loads. You can also turn on the APU and use it's output to drive loads.

        They may have had a more severe failure.

        • phire 17 hours ago |
          Tying buses together is a manual operation. All indications are that the pilots didn't have enough time to get that far in the checklist.

          It's also not that important to restore AC power. The standby bus automatically brings the standby instruments online in the event of a failure, and not that hard to restore the majority of instruments by connecting the main DC bus to batteries. The battery has enough power for 30min of flying, or a full hour if you have the dual-battery option.

          > They may have had a more severe failure.

          While we only have evidence of bird strikes on one engine, the actions of the pilots seem to suggest that they lost both engines. They were rushing to get back to a runway.

          • zmgsabst 16 hours ago |
            The struck engine was still spinning — you can see the visual distortion from its thrust reverser being deployed while on the runway.
        • f1shy 16 hours ago |
          You do not have time to start the APU. Moving any switch is a procedure with a checklist, at that hight, once you lost both engines… they had no chance.
          • timewizard 16 hours ago |
            It's just a button press. US Airways 1549 had the captain starting the APU within seconds after their birdstrike. Your biggest concern after starting is likely exhaust temperature.
            • f1shy 16 hours ago |
              Exactly! And he was called out because he did that without following the stablished procedures!

              Also is not instant, will take a minute or two. The APU is a turbine that takes time to start. Maybe they did turn in on, but 2 minutes too late. We don’t know.

              He was a super pilot, who had extreme good understanding of all systems in the plane, and many ours on that plane. You cannot expect that (and do both want that!) from an average pilot. It seems right now that this very accident may have been caused by not following the procedures.

              • phire 15 hours ago |
                Or they started the APU, but never got around to connecting its generator to the main bus. That's a second button press, after confirming it has successfully started.
              • nctk21 15 hours ago |
                If by "called out" you imply criticised, then that is not correct. Instead, the NTSB final accident report (https://www.ntsb.gov/investigations/accidentreports/reports/..., p. section 2.3.1) praises the specific quick action:

                "Although the flight crew was only able to complete about one-third of the Engine Dual Failure checklist, immediately after the bird strike, the captain did accomplish one critical item that the flight crew did not reach in the checklist: starting the APU. Starting the APU early in the accident sequence proved to be critical because it improved the outcome of the ditching by ensuring that electrical power was available to the airplane. Further, if the captain had not started the APU, the airplane would not have remained in normal law mode. This critical step would not have been completed if the flight crew had simply followed the order of the items in the checklist.

                The NTSB concludes that, despite being unable to complete the Engine Dual Failure checklist, the captain started the APU, which improved the outcome of the ditching by ensuring that a primary source of electrical power was available to the airplane and that the airplane remained in normal law and maintained the flight envelope protections, one of which protects against a stall."

                The final report also gave a new safety recommendation (A-10-66, p. 124) to develop more appropriate (shorter) checklist for dual-engine failure specifically for low-altitude incidents.

                • f1shy 14 hours ago |
                  Sorry I wanted to mean “it was pointed out”. I’m not sure of the connotation of “call out”

                  Thanks for citing the report, which I was too lazy to do. That is my point, as I was replying to “it is just one button” comment. If you do the procedures in the right order, it could be they didn’t have the time to start the APU

          • gorm 11 hours ago |
            Wasn't the main issue here that the airport authority had put an obstacle in an area that was supposed to be an obstacle-free zone?
    • MadnessASAP 17 hours ago |
      With a complete loss of electrical power there isn't much for the FDR to record other then airspeed, altitude, and attitude from the standby instruments. However that's already available from radar, ADS-B, and in this case, video.

      CVR might have some value, but again in most cases you're just getting a bunch of yelling and swearing before the crash.

      As far as understanding what went wrong with Jeju 2216, the interesting bits are going to be right up until they lost both engines, after that it's fairly straightforward to put together the chain of events.

      • timewizard 17 hours ago |
        > there isn't much for the FDR to record other then airspeed, altitude, and attitude from the standby instruments.

        You forgot the _most_ important data. The position of the flight controls set by the crew.

        > but again in most cases you're just getting a bunch of yelling and swearing before the crash.

        That has not been my experience. You can hear the pilots trying to work the problem until the last minute and hearing how they made decisions is important. You can also hear the engines, the wind noise, cockpit warning horns, and possible sources of pilot interference.

        • MadnessASAP 17 hours ago |
          > That has not been my experience. You can hear the pilots trying to work the problem until the last minute and hearing how they made decisions is important. You can also hear the engines, the wind noise, cockpit warning horns, and possible sources of pilot interference.

          Knowing the pilots were trying to prevent the inevitable crash right up until the end is a nice thing to know but now relevant for flight safety. The point of the investigation is to determine what went wrong before the crash became a certainty. Which for Jeju was the moment they lost both engines.

          > You forgot the _most_ important data. The position of the flight controls set by the crew.

          While manual flight controls are a thing, without hydraulic assistance they are really only useful if you're trying to maintain straight and level flight while you start backup power. W1ith compounding problems it gives you a little more control of which direction you will be crashing in.

          I won't deny that more data in an investigation is always useful, but the cost of ensuring that data is available has to be weighed against the potential value of that data. With this crash the data after the loss of both engines won't have much bearing on preventing the similar incidents in the future.

          To put it another way, should your commercial air liner lose both engines at 1000' in a descent the outcome will be crash, the objective is to prevent that scenario.

          • phoronixrly 17 hours ago |
            Some sources would be beneficial here.
          • robocat 16 hours ago |
            > is a nice thing to know but [not] relevant for flight safety [edited to what I think you meant]

            If the pilots made mistakes you want to know what mistakes they made so you can make changes to prevent future pilots from making the same mistakes. The voice recorder is extremely important to understand what mistakes (if any) were made. Even if they did everything correctly, reducing uncertainty is also important.

          • timewizard 16 hours ago |
            > The point of the investigation is to determine what went wrong before the crash became a certainty.

            That's the reason for starting the investigation. It is not the exclusive goal for it. Increasing safety for all future flights is a big part of why and how this is done. Bird strikes are inevitable.

            > Which for Jeju was the moment they lost both engines.

            It seems like you do not require any investigation at all.

            > should your commercial air liner lose both engines at 1000' in a descent the outcome will be crash

            The outcome will be a descent. You do see that the concern here is related to the fact that this failure happened very close to approach and even given this the pilots managed to successfully navigate it to the runway. With the gear up.

            I mean, if you touch down with the gear up, technically, I'll give you, that's a crash. If the gear was down it would look very much like a regular _landing_. What precisely went wrong considering the gear _had_ been down previously? So.. it's not hard to imagine how a working CVR and FDR here would help us answer that very question.

            Although, your investigation may have been this thorough, do you have the answer?

            • MadnessASAP 15 hours ago |
              The biggest question for me, is what happened between the bird strikes and both engines shutting down? Why was the decision made to go around instead of continuing the approach? How quickly did both engines fail?

              > managed to successfully navigate it to the runway. With the gear up.

              Sadly they did not, well technically yes, they did physically locate the runway, however they were also traveling too high, too fast, and crashed.

              > If the gear was down it would look very much like a regular _landing_

              It would have still been a very unstable landing half way down the runway with limited braking, no reversers, no spoilers, no flaps. It's very likely they would have still ended up off the end of the runway.

              > What precisely went wrong considering the gear _had_ been down previously?

              They chose to do a go-around, selected gear up, then with both engines failed the only way to lower the gear is manually, the handles are on the floor aft of the first officers seat. Their would not have been time to access them between when they had lost both engines and were on the ground while also flying the aircraft.

              I'm not saying I know what happened, I'm saying that the key to preventing future accidents like this will be in the moments before they lost both engines. After they lost both engines it's clear they aircrew tried to very quickly turn the plane around and land the runway, I have pretty serious doubts that there was any actions that could've made that a successful landing, and if there was, they'll be found in a simulator not on the FDR of the aircraft that DIDN'T make the successful landing.

          • kermatt 6 hours ago |
            > The point of the investigation is to determine what went wrong before the crash became a certainty.

            Data about recovery efforts after system failures is also valuable. It can help improve procedures for similar future events.

      • phire 17 hours ago |
        > However that's already available from radar, ADS-B

        Not in this case. ADS-B was lost at the same time as the flight recorders.

        If there was a primary radar in range, they might be able to recover groundspeed and a 2d flight path. But those have been falling out of fashion, and were never that good for things near the ground.

      • Dalewyn 16 hours ago |
        >you're just getting a bunch of yelling and swearing before the crash.

        Even that is valuable information for discerning what went wrong and how.

    • f1shy 16 hours ago |
      > 4 minutes is an eternity.

      private pilot here: 4 minutes is nothing while flying and coping with an emergency! NOTHING!

      • ggm 16 hours ago |
        Yes. My comment was to the amount of data loss. I'm not a pilot and I respect pilots. What a terrible situation to be in.
        • f1shy 16 hours ago |
          Sorry. I misunderstood the context.
  • marze 17 hours ago |
    How much would it cost to create a data recorder with a built-in battery? Like a mini UPS that would power the recorder for one hour, if external power failed.

    The antiquated technology in jets is mind boggling.

    • Taniwha 17 hours ago |
      remember they turn them off when they're not being used, a 1 hour battery would go flat every day
      • Over2Chars 17 hours ago |
        a nuclear battery would last a few years.
        • f1shy 17 hours ago |
          Hopefully the radiation does not clear the memory…
        • baby_souffle 10 hours ago |
          And finding the black boxes at a horrific crash would be easier. Just go in the direction that your Geiger counter makes the most clicks.
      • anshumankmr 16 hours ago |
        Keep recharging the battery on the plane,, switch to battery when there is a loss of power.
      • DoctorOetker 7 hours ago |
        aren't most passenger planes in near continuous use, unless scheduled for repair or sold?
        • magic_smoke_ee 6 hours ago |
          Define "continuous use". The ones used daily sit around about half the time in major airports' temporary parking areas away from the ramp.
    • f1shy 17 hours ago |
      The problem is old, new planes don’t have that problem. The the technology is antiquated because the planes are antiquated, because they are (or were) built with good quality and lasted long.

      Appart from that, how much could it cost? Well, the battery maybe a couple hundred US pesos, but the whole test, verification and validation until approval, plus changing all planes? You have to ask Catl Sagan.

      Last but not least: I think the key to understanding the accident will still be there (they shut down the wrong engine, for example) if that is the case, even with this little problem, the boxes dis their work… so, no need for new things.

      • kylehotchkiss 5 hours ago |
        I for one would like a few hundred US pesos
    • skirge 16 hours ago |
      battery can explode and you have two problems
    • brian-armstrong 5 hours ago |
      If a battery is infeasible, perhaps a hand crank could be added for power? In such cases where power was lost, one of the crew could be instructed to provide continuous crank power to the black box.
      • kylehotchkiss 5 hours ago |
        Yeah; just spin this thing so your surviving family members can know the exact sequence of your demise. That seems maybe a little contrary to survival instinct. Isn’t the first rule of aviation to fly the plane first; last; and always? If you have the ram air generator lowered and the airspeed; you have controls to glide (see gimly glider)
  • SteveVeilStream 17 hours ago |
    One possible scenario is that they had a bird strike on one engine but then accidently turned the working engine off instead. Full loss of power from both engines could have taken the data recorder offline in an older aircraft like this one. With both engines off, they may have panicked to try to complete a tight turn to return to the airport and to maintain altitude to make it to the runway. It's possible the landing gear was forgotten but also possible that it was intentional (to extend glide.) It's too early to know anything with certainty but I suspect that the investigation will show that a different set of choices would have allowed them to put the plane down safely (for example by continuing with the initial approach.) Even in that case, they may not have been to blame - for example they may have been following a standard procedure that should be revisited.
    • Over2Chars 17 hours ago |
      I was under the impression that these CDR/FDRs were independently powered. But a quick Quack on DuckDuckGo didn't answer anything.

      That two different recorders both went titsup at the same time I find mind boggling and very sus.

      edit: apparently they have both AC and a battery backup, if the internet is to be believed.

      Which makes a simultaneous loss of two devices with battery backups... curious.

      Did they get hit with an EMP?

      • f1shy 17 hours ago |
        If the motor was just shutdown, it should still make power off windmilling. If they discharged the extinguishers, then there is no more power. In any case, that should be recorded.
      • SteveVeilStream 16 hours ago |
        Battery backup was not required until 2010. This aircraft was manufactured in 2009. https://www.ecfr.gov/current/title-14/chapter-I/subchapter-G...
        • jiggawatts 15 hours ago |
          The general populace seems convinced that black box recorders are made of adamantium and can return data even if the plane falls into a black hole.

          Meanwhile every time I read about a crash, I'm horrified by how primitive the requirements (and implementation) appear to be.

          This reminds me of my favourite method for reviewing software I had never heard of: I check the release notes. If version v17 has "now uses transactions" then that means that the developers were happy to release the first sixteen major versions with data corruption as a feature. (Conversely, if the notes only mention fantastically obscure scenarios being fixed, then the basic issues have been fixed long ago.)

          Battery backup being introduced in 2010 as a requirement is absolutely insane. It's not like batteries or battery backup are new technologies! This stuff has been around forever. Why on earth was this not a requirement decades ago!?

          • lotsofpulp 10 hours ago |
            I find it just as crazy that many or most personal cars come equipped with cameras and sensors all over, but are not required to record any of it for analysis of a collision and attribution of fault.

            The most unqualified person can be driving a 10,000 pound (4,500kg) electric Hummer and have no measures in place to use the built in cameras and sensors to determine what went wrong in the event of a collision.

            Hell, there are cameras inside the vehicle that can tell us whether or not a driver was using their phone when a crash happens, but it does not seem politically popular to enforce reviewing this footage, even though being in a personal car is far more dangerous than being a passenger on a plane.

            • SR2Z 9 hours ago |
              This is not strictly true and most cars have black box modules which record control inputs and IMU data.
            • GoToRO 8 hours ago |
              The airbag module does record acceleration, seat belt status and the general scenario used to deploy the airbags/tighten the seat belts.
            • kjellsbells 5 hours ago |
              The car situation is at the unholy nexus of bad behavior, money, and privacy. My insurance company would dearly love to have access to the existing telemetry in my vehicle. As an alternative, they continually beg me to install a telemetry device in the OBD port in exchange for lower premiums. But their incentive is so obviously to use the data against me in the event of a claim ("ohho! 31 in a 30 zone! Denied!") that I see no reason to do so.

              Dashcams, by contrast, provide the operator with reasonable evidence of other drivers' behavior, and consequently are popular - people pay out of pocket to install them.

              I suspect that if laws were passed requiring use of telemetry and dashcams all over, all that would really happen is that rates would rise to reflect new data on how badly most of us drive, and claim denials would skyrocket.

          • nneonneo 10 hours ago |
            Because adding a battery backup means one (or two!) more batteries that have to be tested and certified to fly; more batteries that have to be checked before flight; more points of failure that could ground an aircraft or cause incidents themselves; more weight on board (especially relevant for smaller craft); etc. etc.

            The battery issue isn’t idle speculation; a recent flight I was scheduled for was cancelled when the airline accidentally allowed a battery to deep discharge, necessitating a new battery to be flown in from another city (delaying the flight so badly it was cancelled instead and passengers delayed to the next day).

            • tomcam 3 hours ago |
              Fascinating. How did you find out?
          • dmitrygr 8 hours ago |
            > Battery backup being introduced in 2010 as a requirement is absolutely insane

            I welcome the formation of your company selling already-certified batteries that will not damage anything around them (like the precious recorded data) if they go from 500 to 0 in no seconds flat, are subjected to pressures of a few kilometers of water, or to a post-crash fire.

            • rixed 4 minutes ago |
              The batteries do not have to be located close to the recorded data
        • themaninthedark 10 hours ago |
          Did some airplanes have battery backup before that?

          It's my recollection of several crashes, where the plane is not immediately recoverable that there is always a concern that the black boxes will overwrite themselves.

      • dist-epoch 16 hours ago |
        Even if the FDR is powered, if the flight computers/systems are without power there is nothing to record.
    • dzhiurgis 17 hours ago |
      Batteries are there for at least 30 minutes
    • f1shy 16 hours ago |
      Note that just shutting down normally the engine would not cause a power loss. Only in the case of a turn off with extinguisher. Of course after bird strike you shut it down with extinguisher.

      Now there is clear procedure, with checks that has to be memorized, where you first identify the engine. Pilots are regularly tested for that procesures. Why it went wrong?

      • asmor 16 hours ago |
        I remember that Boeing changed which pack provides air conditioning to the cockpit between the 737-200 and the 737-300, which lead to a few similar situations in which pilot were confidently wrong which engine was bad based on smell alone.
        • gorm 15 hours ago |
          It was changed in 737-400. Before it only took air from the right engine, but 737-400 took from both. In the Kegworth air disaster, which was caused by a failed blade and the system causing a fire trying to compensate by injection of more fuel, the pilot assumed the fire was in right engine and turned it off as crew never notified them which engine was burning. When they discovered the error the speed was too low to kickstart the engine.

          https://en.wikipedia.org/wiki/Kegworth_air_disaster

          • yetihehe 12 hours ago |
            You would think that such a complicated machine would have some better method to notify pilot which engine failed, instead of "does it smell like right or left engine".
            • 0_____0 10 hours ago |
              There are a series of gauges associated with each engine, but in a high workload situation it might take some effort to make sense of them.
    • m463 15 hours ago |
      It could also be a flock of birds damaging all the engines.
    • mysterydip 14 hours ago |
      Does the APU automatically start or is it a manual process?
      • thedrbrian 12 hours ago |
        Manual process.
  • aaron695 16 hours ago |
    It was a 2009 plane so didn't have to have battery backup which was 2010 - https://www.pprune.org/11803679-post1676.html

    Age - https://www.planespotters.net/airframe/boeing-737-800-hl8088...

    Source it stopped recording - https://www.molit.go.kr/USR/NEWS/m_72/dtl.jsp?id=95090593

  • et-al 16 hours ago |
    It seems like we're in a period of realizing that airplane systems aren't as robust as we've assumed. E.g. MH370's satellite uplink also lost power.
    • fsh 10 hours ago |
      In the case of MH370, this was most likely intentional, not due to a system failure. Hard to defend against a pilot with a death wish.
  • throwaway290 15 hours ago |
    Same physical airplane was in the news a couple of days before the crash for declaring an emergency. Probably coincidence? https://www.ekn.kr/web/view.php?key=20241228028449548
    • kermatt 6 hours ago |
      Google Translate suggests: "A Jeju Air passenger plane made an emergency landing at a nearby airport after a patient was found to be in critical condition during the flight."
  • pfannkuchen 9 hours ago |
    Can we believe the SK government here? If the popular theories about what happened turn out to be correct (e.g. panic, wrong engine turned off), the black box contents would produce an incredible loss of face for SK. Convenient for it to have failed as well.
    • sushid 8 hours ago |
      Did you read the article? The South Koreans couldn't decipher the content due to the extent of the damage and sent the black box over to the US. Do you think they colluded? Please take off your tinfoil hat on the way out.
      • pfannkuchen 7 hours ago |
        The contents of the article are irrelevant beyond the fact that SK says the recorder was damaged.

        I’m not asserting collusion, but they easily could have sabotaged it before sending it to the US. If the US finds sabotage on the recorder, it becomes a political decision whether to release this information against an ally.

        This is well within the realm of realpolitik international political brinksmanship, no tin foil hat required.

  • DoctorOetker 7 hours ago |
    If there was a loss of power, would'nt the passenger cabin have gone dark?

    If that is really the case, I would have expected the passengers to have collectively sent text messages informing friends or relatives that "omg, loud noise, lights went off!".

    4 minutes would be plenty of time to send messages.

    We would have read this information much sooner after the crash, ergo I don't believe there was a power loss.

    I fear the data was deleted by some party.

    • brokenodo 7 hours ago |
      It’s a fair question and I find your comment to be an interesting example of resorting to a conspiracy theory because you don’t understand something that’s easily explainable.

      The Boeing 737-800 has batteries that power certain things in the event of a dual engine failure, including cabin lighting.

      Those same batteries, on 737s manufactured more than about 15 years ago, do not provide backup power to the flight data recorder or cockpit voice recorder.

      • DoctorOetker 2 hours ago |
        That seems totally backwards.

        In the event of power loss, the data and voice recorded have lower priority than cabin lighting?!

        And the cherry on top is that this was the case on 737's until ~ 2010?!

        Defusing my bullsh!t alarm with a second layer of bullsh.t?

        I would swallow it all ad fundum if you can provide HN with reliable links to manuals, procedures, ... describing how electrical energy for the passenger cabin is prioritized over the data and voice flight recorders!

        • wlonkly an hour ago |
          I don't know if that's correct or not, but if I were to prioritize, I'd put safe passenger evacuation ahead of the recorders.
          • DoctorOetker 39 minutes ago |
            if the parent poster was talking about merely emergency lights being battery powered, and hence main passenger cabin lighting turning off a full 4 minutes before the crash, we go back in a circle and have to explain the apparent absence of passengers texting friends and family about a scary power loss during descent (people would quickly try to figure out if power sockets for their consumer electronics were no longer providing power either and learn from each other that it is the case).