Malaysia Airlines MH370: Possible debris in Indian Ocean | Hacker News Reader

Malaysia Airlines MH370: Possible debris in Indian Ocean

abc.net.au

82 pointsqzervaas88 comments

[EDIT] Adding a couple of more recent tweets from Micah Grimes indicating this may be the end of the search for MH370:

> JUST IN: @WrightUps from above Indian Ocean says US P-8 crew "getting radar hits of significant size;" trying to get visuals on hits.

> Australian maritime authority official calls objects credible and of "reasonable" size; largest object about 24 meters.

======================

ABC journalist David Wright [[0] is currently on the P8 plane that is searching for the debris.

Here's a tweet [1] from ABC's social media editor who (apparently) must have spoken to him over voice comms:

>.@WrightUps from Navy P-8 search plane: "We are just descending through clouds right now ... about 1,300 miles southwest of Australia."

[0] https://twitter.com/WrightUps

[1] https://twitter.com/MicahGrimes/status/446501269155618816

It'd be really interesting to know if Australia picked up some trace of the plane on the Jindalee over the horizon radar. [0] Check the range there.

The path of the plane may just have touched the radar's range at some point or some idea of the track may have been worked out.

[0] http://en.wikipedia.org/wiki/Jindalee_Operational_Radar_Netw...

Possible. Our Foreign Minister wouldn't comment except to say that the government is providing all resources at our disposal in the search.

Some more info: http://www.abc.net.au/radionational/programs/breakfast/could...

Probably not. The aircraft would have been at the extreme edge of the radar's range, and in an area that is probably not getting scanned terribly often.

Jindalee uses a directable beam to scan the area or interest. It takes hours to do a full sweep of its entire range (bear in mind that with over the horizon radars you have to do a 2D sweep, not the 1D sweep of a standard line of site radar). Generally speaking, the radar tends to be targeted on areas of specific interest, not just randomly sweeping (the sweep pattern depends on current threat profiles). Maybe if Malaysian authorities alerted Australia quickly enough they may have pointed the beam to the right area, but as I understand it the alert didn't start going out until much later. Disclaimer: ex RAAF radar engineer...

Unfortunately even if this is the case (and color me skeptical given the sheer volume of media misinformation that this story has generated), the cockpit voice recorder only records the last two hours of data, so regrettably we may never know what happened in those moments when the flight was diverted, even if the wreckage is found and the black box is salvageable.

EDIT - Clarification that the cockpit voice recorder (not flight data recorder) only captures the last two hours. Source: http://en.wikipedia.org/wiki/Cockpit_voice_recorder

We may be able to understand what happened through the last 2 hours of data though. If there's pure silence in the cockpit we could infer a number of things about what occured, plus the debris would give additional information as well.

I believe the Cockpit Voice Recorder (CVR) only keeps two hours, but the Flight Data Recorder (FDR) records for considerably longer. Both make up a black box. It will certainly be far more difficult to establish what happened without voice recordings though.

I guess they try to keep the airplanes cheap enough that they can't spring for a 16gb card in the CVR.

If you think the rule about turning off your electronics during takeoff and landing is conservative, try to imagine what it would take to certify a solid-state memory card for use in that application.

Why would doubling or quadrupuling the storage be a problem? Assuming you have the cost of a new chip paid for, I dont see the capacity issue. Or are you saying solid state is incompatible with the engineering parameters of a black box altogether? In that case, the answer is even simpler: just add more of what's already in there.

That's the point, it's not simply a capacity issue. They could plug in a $10 SD card that comes pre-installed to a Nintendo DS and solve the capacity issue.

Rather, it's a certification issue, and the certification process itself is so expensive that you don't simply upgrade every black box every 4 years as storage gets cheaper.

There's also a pilots union issue, they don't want every offhand utterance, gripes about management, gossip, etc. preserved for eternity.

This sounds more likely the real answer.

This is correct. Looking at the most recent FAA CVR reg. revision in 2008, which increased the required duration to 2 hours:

"Revisions to Cockpit Voice Recorder and Digital Flight Data Recorder Regulations" [1]

The FAA proposed that all CVRs be able to retain the last two hours of cockpit audio. Both the NTSB and the Transportation Safety Board of Canada noted that the short duration of available cockpit audio hindered the investigation of several accidents.

The Air Line Pilots Association (ALPA) did not support the proposal to increase CVR recording time because the FAA did not propose any increase in the privacy protections regarding the access and use of information recorded on a CVR. The ALPA stated that existing protections are inadequate despite years of its attempts to change the standard.

[1] http://www.gpo.gov/fdsys/pkg/FR-2008-03-07/pdf/E8-3949.pdf

That actually sounds fairly legit, especially since the commenter below notes that the union has asked for more privacy protections but not received them. Hadn't realized the policy debate was multi-sided.

Just to be clear, I mean an actual aero-space chip specifically made for this application. Here is a cost estimate. Its $8K/per, but honestly that's a just a single pair of first class tickets...

The U.S. National Transportation Safety Board has asked for the installation of cockpit image recorders in large transport aircraft to provide information that would supplement existing CVR and FDR data in accident investigations. They also recommended image recorders be placed into smaller aircraft that are not required to have a CVR or FDR.[4]

Such systems, estimated to cost less than $8,000 installed, typically consist of a camera and microphone located in the cockpit to continuously record cockpit instrumentation, the outside viewing area, engine sounds, radio communications, and ambient cockpit sounds. As with conventional CVRs and FDRs, data from such a system is stored in a crash-protected unit to ensure survivability.[4]

Which is worse? A plane goes down and you have too little data because of strict regulations or a plane goes down and you have no data because of loose regulations?

yeah i'm sure you 16gb SD card data is made to survive several hundred Gs collision crash.

EDIT : adjusted Gs

It doesn't matter that much whether it does or not. Modern data storage is so cheap that we could stick in 8 of them in addition to the super-touch CVR at very little extra cost, and odds are that one of the cheapie units would survive. The hardened CVR system is great, but it should be a last resort sort of thing. We can have cheap, redundant, and high-capacity systems in addition to the CVR for peanuts, so why don't we?

Actually a small device with no internal moving parts will take 50Gs no problem, 1000s of Gs if you have some decent potting. It's keeping some sort of power source active through a high G crash that's the challenge.

SwissAir 111 impacted at an estimated 350g.

The sudden stop at the end of a fall is routinely that high. If you look at warranties for spinning disc hard drives, they'll be rated for several hundred G during transit. Sounds silly, but it's the difference between landing on a surface with give, like carpet, or landing on an unforgiving surface like concrete.

Source? I know nothing about this, but Wikipedia says:

"Most FDRs record approximately 17–25 hours worth of data in a continuous loop."

http://en.wikipedia.org/wiki/Flight_data_recorder#Design

EDIT: ah, it looks like you are talking about the cockpit voice recorder, which records 2 hours of data: http://en.wikipedia.org/wiki/Cockpit_voice_recorder

Several sources, here's one: http://www.washingtonpost.com/local/trafficandcommuting/mala...

Yeah - sorry I should have clarified voice recorder not the flight data recorder. Will edit.

I think you're thinking of the cockpit voice recorder [0], the flight data recorder [1] (also referred to as the black box) contains all sorts of data about what commands were sent to plane and, Wikipedia says, records "17–25 hours worth of data in a continuous loop" (presumably overwriting the oldest data once it's reached capacity).

[0] https://en.wikipedia.org/wiki/Cockpit_voice_recorder

[1] https://en.wikipedia.org/wiki/Flight_Data_Recorder

edit: just about everyone's beat me to it!

I'm not sure what the current physical medium / technology is, but with some generations, IIRC forensics could actually recover several of the most recent recording passes over the medium that travelled in a closed loop -- if there were enough incentive and resources to do so.

We could learn quite a bit, for instance, if the voices on the recorder were neither of the flight crew, or whatever the subject of what was said could suggest whatever had happened at the diversion

... unless whoever was smart enough to disable the transponders would disable the cabin mic, which would probably be the least informative condition

The CVR only recording two hours is absolutely inexcusable. Why does my iPhone - $850 retail - have capabilities that far surpass those of a $250 million modern passenger jet? Hopefully after this incident, public outrage over these ridiculous technological shortcomings will be the catalyst for legislation. Throwing in some extra satellite transmission, GPS, and voice recording capabilities that will hold up even in the event of an electrical issue or fire, and that cannot be disabled by bad actors, just cannot be this difficult.

In my view, the moment this plane went off course, or a fire alarm went off or any other irregularity occurred, a Malaysian Airlines control center should have been alerted and been able to bring up video of the cockpit, maybe the passenger cabin as well, and been able to remotely view diagnostic and positioning data. All of this could easily and cheaply occur via a satellite internet connection.

Edit: For the record, I am not suggesting that plugging a commercial iPhone into a cockpit would work - I'm comparing the capabilities of consumer technology to cockpit technology, which seems woefully inadequate to me.

>Why does my iPhone - $850 retail - have capabilities that far surpass those of a $250 million modern passenger jet?

Because your iPhone isn't designed to survive a plane crash.

The requirements for both design and certification are so different that comparisons to consumer electronics (especially regarding price) are meaningless.

So the thought that it might be possible to install some more memory on the CVR is completely outrageous?

While correct. I for one don't see a reason they could not supplement the highly fault tolerant Cockpit Voice Recorder with a backup system that can record longer periods of time. worst case you only get the two hours.

The airlines can, but they don't because if its not mandated they aren't likely to do it unless it is advantageous for them.

There is an additional data recorder that operates like this the "Quick Data Recorder" designed to be able to be removed and read regularly by the ground crew to better diagnose and fix the plane.

Unfortunately the voice recorders are really only used when the plane crashes, so the airlines aren't going to pay to put them in.

Add to this the fact that anything being hooked into critical plane systems (and a lot of them are considered critical) needs to go through hundreds of hours of testing means your not talking a cheap option.

Not sure why you're getting downvoted for this comment, but I agree with what you're trying to say regarding storage. Granted, of course there are some survivability and hardening considerations to take into account for this application vs. your iphone, but at it's core we're talking something not too dissimilar in concept. For reference, this is the link to one of these products: http://www.uasc.com/products/cvfdr.aspx

It clearly states that the data for modern CVRs are held in flash storage. For something that costs thousands of dollars, and is absolutely critical to after action flight accident discovery, there's no excuse not to capture the voice data from every flight in full given the current prices of flash storage.

Not sure why you're getting downvoted for this comment

Probably because I said I had an iPhone :). Either that or the people downvoting it thought I was suggesting that we plug an iPhone into the cockpit and all will be well. I was simply saying that many of these problems have been solved for use in far less important applications, so it should not be that big of a stretch to get them into commercial aircraft. The suggestion that it might be possible in 2014 to install additional memory into a CVR is apparently such a technological leap that it is worth 5 downvotes :). I cannot wait for the day when it might not sound so crazy.

No, it's because you're not considering properly the problem at hand. You're comparing apples and oranges by comparing consumer hardware capabilities and the capabilities able to be achieved for a hardened service environment solution.

It's the same reason why we don't just go and install a new quad core x86 on rockets for their guidance.

So you are saying that there is a hard technological limit of 2 hours on a CVR because it must be in a hardened container. Got it.

No, thanks for taking my words out of context. I'm saying that there's a whole litany of processes and requirements alongside storage capacity that necessitate limitations. In the case of storage, it will be that certian memory chips, etc. have not been through the process of hardening them. Just like how the latest core processor could run all of the equations required to guide a rocket to orbit, but it sure as hell isn't done on one of them because a number of other features would be neglected.

Is it possible to get more than 2 hours of memory in a CVR? Sure. Why hasn't it been done? Due to a variety of other considerations on hand.

The initial downvotes were because the comment was much shorter and less informative when it was originally posted.

I think it's worth having the discussion of having a permanent locator on aircraft that can't be disabled, remotely-triggered hot mikes, and longer CVRs. But I downvoted you for falling into the "if we can send a man to the moon, why can't we ____" argument trap, which is more or less the straw man.

> just cannot be this difficult ... easily and cheaply

Ever get a feature request from someone who, in a single sentence, describes months or quarters of development time and introduces all kinds of risks and dependencies, but thinks it can be done by next Tuesday?

I have, but I don't think that's relevant here. This isn't a "next Tuesday" issue. Planes have been flying over oceans for decades, and these issues have been contemplated and examined by allegedly bright people for at least that long. Yet, somehow, here we are in 2014 searching for a jet with 230 people aboard and are clueless as to where it is or how it "vanished".

I can assure you that whatever happened, it didn't vanish, and that those people likely suffered horrible, violent deaths. It is frustrating that we wouldn't apply modern technology to learn about and potentially prevent such tragedies. Even if seeing what happened in real-time couldn't have stopped this incident, knowing what happened could prevent similar incidents in the future.

Legislation is what set the CVRs at two hours in the first place. §121.359(i)(1) [1] sets the minimum. It should hardly be surprising that these manufacturers don't exceed it. My strong suspicion is that there's a design that meets the requirements and people aren't putting money into developing better ones until more regulation requires it. Obviously I know nothing about international law, but given how closely developed nations seem to coordinate commercial air travel regulation, I would suspect that changing this would take forever.

[1]http://www.law.cornell.edu/cfr/text/14/121.359

You're right of course on this being a legislatively driven requirement.

This presents an interesting case study on how the sclerotic nature of a legislative (or bureaucratic) process sometimes restricts adoption of beneficial technology.

"Floors become ceilings" as they say.

That's certainly true, but you could also make a strong case that flight recorders aren't necessarily a place where you're going to have a race to put as much technology in them as possible. I doubt we'll see a YC startup for this any time soon.

I could make the most awesome flight recorder ever seen, but nobody will buy it (speaking of which, who actually buys these? The airline like they do with engines or the manufacturer like they do with, say, the batteries?). They're going to keep buying the Honeywell or the Teledyne one because that's just who they buy recorders from.

I can't even fathom what would be the cost to enter this market. The engineering and testing and certification costs must be immense. Even if you do this, I don't think you'd be able to charge a premium for a better product. Imagine yourself as the Boeing 797 project manager: you're hundreds of millions over budget, the shareholders are getting upset, etc. Now your data recorder engineer comes up to you and says "here's a list of data recorders that meet federal and international standards, sorted by price." Which end of the list do you start from?

I'll bite.

> Why does my iPhone - $850 retail - have capabilities that far surpass those of a $250 million modern passenger jet?

It is not unheard of, or unknown, for iPhones to crash, have unrecoverable data loss, or suffer from hardware issues like defective flash. A CVR needs to work all the time, without fail. People will excuse their iPhone crashing out from time to time. This is inexcusable on a CVR. You're talking about planes that run flight computers with three different CPUs and instruction sets to prevent bugs in the underlying silicon from manifesting.

> fire alarm went off or any other irregularity occurred, a Malaysian Airlines control center should have been alerted and been able to bring up video of the cockpit

Even without the technical constraints it's simply not worth it. Why would a control centre need live video of the cockpit in the case of a fire alarm going off? If there's a fire the crew are not phoning home. Communication with people outside the plane is a low priority. The 'Malaysian Airlines control center' is not going to be getting in touch with the pilots. That's not how it works.

> Hopefully after this incident, public outrage over these ridiculous technological shortcomings will be the catalyst for legislation

I have worked on hardware that goes into commercial jets. These rules may seem silly to outside observers, but those observers frequently don't realise exactly how harsh the environment is, and the constraints you're dealing with. What they are not are 'ridiculous technological shortcomings'. Commercial airline safety is already one of the most highly regulated forms of travel out there.

We don't use 'domestic' components for the same reason they don't use them for satellites: failure is not an option. The 'live video feeds' and 'constant satellite links' people keep bringing up are only of use in relatively minor emergencies. They serve no purpose during things like catastrophic mechanical failure, because you've already lost the satellite link at that point. The plane is going to go down. We will find the plane. We will find the CVR.

Again, I didn't suggest that plugging an iPhone in is the answer. It was a comparison to the capabilities of consumer electronics vs. those in cockpits of supposedly advanced aircraft. We can, and we should, install more memory into CVR's if they can only record two hours of an international flight.

In your example of a catastrophic mechanical failure, a battery powered satellite link wouldn't necessarily lose touch until impact. At the very least, the plane's position would be known and rescuers dispatched. Critical evidence in this incident is likely long gone because of the time that has passed - even if we find it today. Video feeds etc. would at the very least assist in preventing future tragedies and help us understand exactly what happened. We would know, for example, the specific tactics of hijackers (if that was indeed the case) or that the pilot just went nuts, or if there was a mechanical or fire risk issue that must be addressed on other planes. We would know answers to questions we would likely never otherwise be able to answer.

You're illustrating my point about how people without experience in the field are severely underestimating the harshness of the environment.

> a battery powered satellite link

Battery powered? Batteries are big fire risks (ask Boeing, they know all about it). Where are you going to put the battery? You can't put it next to the transmitter (which, needing line of sight, would go on the top of the plane), because you want the battery to be serviceable with relative ease. So now you've got a power line running from the battery to the transmitter, which is no different to a power line running from the main power feed. Let's scrap the battery.

> wouldn't necessarily lose touch until impact [...] at the very least, the plane's position would be known and rescuers dispatched

Satellite linkups are flaky, they need line of sight. Planes at cruising altitude travel extremely fast. The plane's exact position would not be known.

> Critical evidence in this incident is likely long gone because of the time that has passed - even if we find it today

If the recorders can be recovered this is unlikely. Flight recorders are designed to capture the information needed to figure out what happened. Air France 447 was lost for two years, and yet investigators were still able to establish with near certainty what happened.

> Video feeds etc. would at the very least assist in preventing future tragedies and help us understand exactly what happened.

I can see how video feeds could possibly be useful in establishing what happened in a terrorist attack or hijacking. In terms of telemetric data and establishing mechanical failure causes they are not very useful.

I get that a lot of people on HN have lots of great ideas, but this is an area where you really need to know exactly how harsh the environment is before you can start throwing suggestions around like they're basic common sense.

When we look at the costs of investigating this disaster and the search for AF447's black boxes (and finding them was not a certainty, as far as I recall),I'm sure the cost equation will shift in favor of more extensive (certificated) in-flight monitoring technology, after all, if this lead turns out to actually be the plane, it will be because of a remote monitoring system that the pilots couldn't turn off.

Does the voice recorder still function if they pull all the fuses? It's possible, if that's the case, that they will get the lead up to the event on voice recorder. seems like planes have redundancy though and this particular system is probably on a different power supply.

Well, Finally a credible source of information. Things we take granted in "western" news organizations and research come to light when you see how investigation is done in the hands of some people. No disrespect to the Malaysians I love their country and culture but the authorities did a disaster of a job handling it all. More disinformation than anything.

"credible source of information"

you're not Australian are you? ;)

I think the Australian Maritime Safety Authority is a pretty credible source.

Very true, I was just making cheap funnies :)

It's not reddit, Oliver.

Reddit for redactors that don't want to sort through the garbage.

Friend, hundreds of people may have died. Do you think cheap funnies are really appropriate?

It wasn't a joke aimed at those poor people, it was a political joke aimed at our Prime Minister. I apologise if you though I was making light.

More seriously, does anyone find it odd that the PM of Australia is making this announcement? For example, if the US Navy found something, I would not expect President Obama to announce it.

He's announcing it to the House of Representatives. Same thing would happen in the UK with the House of Commons and PM David Cameron. Designed to provide ultimate transparency within the system. Both PM's are obligated to inform the house as their primary responsibility.

Westminster parliamentary system.

He "announced" it to the parliament during Question Time, there's a real press conference by Maritime Safety in a few minutes

Isn't Obama busy with Crimea?

Yes, it's embarrassing. It shows the un-professionalism of the situation.

One assumes the real press conference was delayed so he could do the grandstanding.

That said, there should have been nothing said at this stage anyway until some sort of real conformation was performed.

I guess it's all about entertaining the masses who demand their ongoing live entertainment fix.

You have allowed your political views to cloud your judgement.

Details of the satellite images and retasking of search aircraft would surely have leaked and the Australian government would have been rightly criticised for withholding that information had it not been announced.

I am a Malaysian, and I concur with your assessment on Malaysian "authorities".

As a fellow Malaysian, we should cut the government some slack. This was a thoroughly difficult case - even with international assistance, we still couldn't locate the airliner.

That said, our government is incompetent, that can't be denied.

I wouldn't equate this situation with incompetence. Lack of resources and training. Nor should we expect a country of that size of have the training and resources but that expectation is being placed on the Malaysian government. Its really tough spot for the Malaysian government, people want answers and finding an airplane is like finding a needle in haystack the size of rhode island. In comparison, Air France wreckage took 2 years to find.

It doesn't help there's all these theories about hijacking or the pilot going rogue. People/Media ignore the most probable theory that the plane had mechanical issues and crashed. I had calm my mother's fears on the phone couple days ago as she heard from a media source that plane was going to be used to deliver a nuclear weapon to the US mainland which is a completely absurd theory. Real incompetence is how the media spun this event out of control in the name of ratings and ad revenue.

Regarding AF447, small correction, the first major pieces of wreckage were found within 5 days[1]. The black boxes took around 2 years to recover.

[1] http://en.wikipedia.org/wiki/Air_France_Flight_447

Good thing we've got this covered here on HN, because there's been like a media blackout on this subject...

Largest piece is 24 meters, location in the zone identified by the NTSB.

AMSA: One of the objects on the satellite image has been assessed to be 24 metres in size #MH370 #Malaysia

Src: https://twitter.com/ABCNews24/status/446508379783974912

From their press conference, held about 30 mins ago.

From the photos of the two objects posted on @MicahGrimes (https://twitter.com/MicahGrimes)

http://t.co/gFZJhGfTNG http://t.co/SRYeeZfkc2

Distance between: 43:58:34S 090:57:37E 44:03:02S 091:13:27E

is 22.72km according to WolframAlpha

https://www.wolframalpha.com/input/?i=distance+between+43%3A...

There are just so many stories in the media it is hard to know what to believe. This airline disappearance is becoming a place for all sorts of conspiracy theories.

I hope people take cognizance from the disappearance and learn how such a thing was even possible in our high tech world of today.

What are the chances that it is just ocean debris? I would say it's quite high.

> http://en.wikipedia.org/wiki/Indian_Ocean_garbage_patch ?

24m is too large for some junk?

As per that, the vast majority of mass is small and sub-surface, not visible from satellite or ship.

Hoping this is true and the plane's black box can be located as well.

Even if some debris is found, it might still take years to locate the actual aircraft wreck deep underwater, and recover the black box(s) and get any real answers (like in the Air France crash).

They still have about 2 weeks before the pinger runs out of battery. So there is a decent chance they could find it within 2 weeks, but if the wreck isn't found by then, it could definitely take years to find it, if ever.

I suspect this kind of sonar application is why TI is making analog-to-digital conversion chips with 31 bits of resolution: http://www.ti.com/lit/ds/symlink/ads1282-ht.pdf

Compare to 16- to 24- bit resolution in your typical audio processing chain.

In the case of Air France 447, it did take 2 years to recover the black box. I'm guessing this one will be quicker because of the potential terrorism angle.

It would mean no survivors. So sad.

Unless the debris includes a liferaft.

But I'm not keeping my hopes up.

I really hate to be negative about this but isn't this highly unlikely considering the rafts on the planes have emergency locator radio beacons?

Other things on the plane also have similar beacons (the flight recorder, most notably) but have yet to be found, so it's not completely out of the question that they be deployed but not located.

Anytime an airplane disappears the probability to find anybody alive is small. The day it happened I was already feeling sad about it. I just can't imagine how the families must feel every time news are announced.

Where do you think survivors could possibly be? A mysterious tropical island, lost in time and space?

How about a widely held view so far: Possibly held captive by hijackers or terrorists? Not desirable, but better than certain death indicated by debris (if confirmed).

New theory #239858935: The copilot killed the pilot with an axe, set the plane onto auto pilot towards the southern seas, then cut his own throat and let the rest of the passengers die helplessly.

Reply on news.ycombinator.com