Hacker News new | past | comments | ask | show | jobs | submit login
Jeju Air jet black boxes stopped recording before crash: transport ministry (koreatimes.co.kr)
24 points by gslin 14 days ago | hide | past | favorite | 26 comments



"The analysis revealed that both the CVR and FDR data were not recording during the four minutes leading up to the aircraft's collision"

How does this even happen where two systems would stop recording data. This story seems to get weirder and weirder. I'd always been under the impression that black boxes were really highly secure and extremely rigorously developed to specifically prevent data loss in situations like this.

Bad enough for the accident to occur, but to not have accurate telemetry and recorded data from the onboard systems is definitely going to make an already hard investigation very much harder


According to 737NG technical diagrams I found online, the CVR is powered by the XFER BUS 2, and the FDR by the DC BUS 1. Neither of these busses can operate from the battery, but both should be operating if either engine or the APU is running, provided all switches are in the correct position. On the surface it seems odd to not run the CVR/FDR from battery, but on the other hand it’s not a critical flight system.

This suggests either dual engine failure, or single engine failure followed by shutting down the wrong engine with the fire handle (which cuts off electrical power), or (unlikely) a catastrophic electrical system failure. This isn’t a new interpretation, since the ADSB transponder also shut down at about the same time which led many to assume the electrics had gone out for some reason. My guess remains that they shut down the wrong engine given that the reported bird strike occurred (reportedly) a few minutes before the plane seems to have lost power, but the timeline is still not certain.

However, it did seem like the engine that had the mobile footage with the compressor stalls was still running at landing which should usually be enough to generate electricity since even windmilling is enough, which adds another wrinkle. Perhaps the bird strike destroyed the generator?


Some recorders have a built in battery for exactly these cases (called Recorder Independent Power Supply), and (as I discovered while trying to figure out whether this exists) Canada seems to require them since 2023.


It might still give us some hints. In the other thread they say that the black box might have stopped recording because of a power failure - and planes manufactured before 2010 (this plane was from 2009) didn't need an UPS.

The hint that we get from this news is that the plane (or better, the blackbox) was missing electricity for 4 minutes prior to the crash.


I think you have a typo either on 2019 or on 2010.


2009, sorry. Fixed! Thanks!


Plane was built in 2009.


Easiest explanation is that they were not powered. They are on a bus that doesn't get power from the battery on 737, they need either APU (doors of which were found half closed on the wreck) or one of the engine generators. That would explain lack of gear, slats and flaps, too.


I'm learning today that black boxes don't have any sort of built-in emergency battery power, apparently. For a device that is meant to be a data gatherer up until the "last gasp" of the airplane, that seems like a pretty obvious design flaw to me.

Does anyone know why?


Money. Adding a battery adds cost and weight (and complexity, but we know how to engineer and maintain things properly).

The cost/weight would have been a lot more prohibitive decades ago when the rules were made, and regulators are slow to adapt.


And for every additional kilogram or pound of weight, a typical 2000mi / 3200 km flight will consume an equivalent mass of fuel.

Conventional (fuel-powered) aircraft, and to an even greater extent battery-powered aircraft, are absolute marvels of light-weighting engineering, driven by both materials and designs.

As FDR/CVR are used only once, and that in a vanishingly small fraction of aircraft, the cost considerations of carrying additional mass are high. It's not just the aircraft that crash, but all aircraft across the entire global commercial fleet. That's on the order of 32,000 aircraft presently, each with multiple flight cycles per day on average.[1] That's on the order of 50--100 tonnes of fuel consumption per day if the total added weight is limited to 1 kg.

Black boxes must also survive crash conditions. A battery itself might be a fire risk, and would require additional material to protect both itself and the data storage of a black box in the event of an accident.

That said ... one would think that current lightweight battery packs might make short-term battery operation reasonably viable.

--------------------------------

Notes:

1. "Information supplied by aviation data specialists ch-aviation shows that at as of July 7, there were 31,734 civil aircraft in the global fleet being used by more than 750 different airlines. " <https://www.key.aero/article/how-many-commercial-aircraft-ar...> Story appears to be from 2020 per Internet Archive: <https://web.archive.org/web/20200929012536/https://www.key.a...>.


Thanks for that explanation. I didn't realize the fuel costs were quite so bad, and extrapolating it to all aircraft per day makes sense as well.

Yes, that's a key point many advocates fail to consider (or express), and those concerns carry their own risks and costs, including of lives, though often indirect.

As an example, what is the contribution of air travel to climate change and wildfire, such as those presently underway in Los Angeles and environs, now costing > 12,000 homes and 24+ deaths? Aviation overall is about 6% of global petroleum consumption as I recall, and safety systems would then be a pro-rated fraction of that.

Another lesson of my IT career has been that a frequent cause of outages and risks is safety equipment itself. Failover routers that fail, failover power supplies that wipe out power, load balancers which lock up, firewalls that wall legitimate traffic, logging systems that fill disks, alerting and monitoring systems that overwhelm pagers and staff, etc., etc. Every technological mitigation has both benefits and costs, often manifesting in different timescales and conditions. We tend most often to adopt measures which promise quick and manifest results, or whose negative consequences are temporally distant and vague, but this also means that we underutilise those which offer distant and vague benefits (exercise, healthy eating, as the classics), or whose costs are manifest and immediate (again, exercise, various medical interventions such as exams and vaccinations, many safety measures and mitigations).

Commercial jet aviation is remarkably safe, but that's come through an intense focus on safety at all levels, from aircraft to operators to air traffic to ground response and most critically to thorough evaluation of incidents. One consequence is that many notable recent events have come from one or more of those factors themselves: pilots intentionally crashing aircraft (GermanWings 4U9525, 2015; probably Malaysian Airlines 370; EgyptAir 990, 1999), by passengers (most notably the 9/11 attacks in the US, 2001), or by safety or compensatory equipment (most notoriously the Boeing 737 MAX MCAS system, and the crashes of Lion Air 610 and Ethiopian Air 302). As with many other technical and economic phenomena, arguments of marginal utility strongly trend to developments reaching a pain threshold where harms increase to balance benefits, making naive analysis difficult.

(A realm in which this is probably evident to many HN readers is the tendency for software and online services to enshittify, in Cory Doctorow's marvelous coinage, to the point that they (attempt to) perch delicately on the threshold of fatal annoyance. A chief problem with such tactics are that the threshold itself is a remarkably unstable point, and induces its own further interactions and harms.)


It’s a mystery, and according to a cousin post, black boxes from planes manufactured recently do have backup power.


Federal Express Flight 705 comes to mind, in which a crew member disabled the CVR by pulling it's circuit breaker before attempting to hijack the plane.


It is quite a coincidence.

Is it possible that "some buffer just did not flush" as in software development may happen.


I’ve studied this in older aircraft, and no that was possible because it was recorded directly to tape that records over itself in a loop.

I’m not sure if that’s still true on newer planes.


There was a case where a black box tape recorder missed the last 6 minutes

https://en.wikipedia.org/wiki/Swissair_Flight_111?wprov=sfti...


4 minutes is too long for a "missed flush". This looks like bad software simply. One may wonder if it is also developed by Boeing.


Boeing only integrates the black box to the planes they don’t develop them. They are made by companies like Honeywell Aerospace


not the same article, but same subject with more discussion: https://news.ycombinator.com/item?id=42664097




> Investigators have pointed to a bird strike, faulty landing gear and the runway barrier as possible issues.

That runway barrier, “possible issue”.


The runway barrier didn't make the plane fail to extend its landing gear and overrun the runway. It certainly contributed to the severity but it doesn't explain the incident.


It should most definitely be included among the “possible issues”




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: