We now live in ages that even the data from billions of iphones or ipads is back-upped into the cloud, but airliners still use data-backup into odd FDR's inside the plane ?
In case of accidents it takes months and millions of dollars to explore the depths of the oceans for lost FDR's...
With actual technologies it is a piece of cake to frequently upload datastreams from aircraft to a central organization via satellite.
Satcom transfer is already in use for backup of engine data, why not use it for this information also ?
20 years ago i would agree it would have a high price, but now ? Every airliner from the 10.000 that are worldwide airborne, could stream easily his data trough satellite.
Within a few years we will do it even for every car !
I do not understand this. It seems as aircraft-builders are not so interested in this kind of solution, maybe because it could reveal information that they rather prefer to keep unknown in certain cases ?

Actually there is activity afoot to implement a system similar to you are thinking of. Your assertion that manufactures and, or operators is not valid. It's what we here in the US call BS:)

It's probably expensive and unless authorities require it airlines don't want to spend the money...to make it reliable it has to work via Satcom as you say and most short haul aircraft are not equipped with Satcom (they are operating within VHF range most of the time and if the have the odd flight outside of VHF range they have an HF radio fitted...), so it would have to be fitted as well as the cloud system you propose (you may consider running the cloud over VHF, but VHF datalink is highly unreliable - it usually works okay for getting weather or messages from operations, but VHF CPDLC is a total mess, it usually disconnects itself every 15 or 20 minutes or so...). So if you found a system that works okay and get the aviation authorities to demand it airlines would (have to) install it, but it may have a noticeable impact on ticket price, at least on short haul airplanes (which the traveling public may not be willing to pay considering that accidents are thankfully so rare).

I think you are under estimating the shear volume of data. FDRs and CVRs are on a continuous back up of real time data of 30 min. FDRs now record at least 18 parameters. Every commercial aircraft would have to have a on going SATCOM link to record that info in real time. That is a lot of satellite bandwidth. There would need to be a lot more satellites for world wide coverage.

The bandwidth to get that data off each individual aircraft, real-time, is substantial. Multiply that by however many long-haul aircraft are airborne at any given moment (add whatever "fudge factor" you'd like), and you've got a LARGE chunk of bandwidth consumed solely by HUMS data from airliners. Airlines aren't the sole users of the communications satellites that provide this bandwidth; they have to share with other users. Satellite bandwidth isn't exactly cheap.

Seems to me the amount of data wouldn't be that large at all.

I did an estimate, amateurishly based on the bandwith required in my field of audio.

This is based on 16-bit word-lenght, an estimate of 22,050 airliners with a parameter sampling rate of 1Hz. (Congrats those who spot the audio-parallell.)
100 tracks, eh, sorry, I mean 100 parameters.

That would end up with a global grand total of about
500 Megabytes per minute if every plane in the world is flying. Or about 80Mbps.

That is an average home broadband connection, or your iphone's 4g connection.

(Previous reply is FDR. CVR is a lot more bandwith demanding. )

Before this kind of thing is implemented, the question to ask is whether it will improve safety in a tangible way. Is it money well spent? Can we really say that having all this real time data will decrease the risk of future accidents? How many crashes remain unexplained because the data is unavailable? Sure, MH370, but that case may well have been pilot action (not confirmed but reasonable explanation). Collecting data won't help in that scenario. Also, that was one aircraft in over three decades. Hardly a significant risk to the traveling public.

There are plenty of areas where money can be spent to improve safety. To name a few: runway incursion prevention, (more) high altitude upset training, contaminated runway education, datalink in locations that don't have it, etc etc... The impact of more funding in these areas would impact safety directly and rapidly, as opposed to chasing data collection where the benefit is marginal.

How can we ensure data security if we upload all this sensitive data into the cloud? Can you imagine what will happen if it rains?

#PostLikeABhakt

Starlion, Real-time telemetry have spectacular potential for contributions to safety. Of course there needs to be secure encryption of this data as well as clear limits to it's use. This data belongs to the safety people and needs to be kept far away from the economy-department.

These are only slight challenges in exchange for an incredible gain in safety and knowledge of real-time, real-life conditions.

M564038 wrote:

Starlion, Real-time telemetry have spectacular potential for contributions to safety. Of course there needs to be secure encryption of this data as well as clear limits to it's use. This data belongs to the safety people and needs to be kept far away from the economy-department.

These are only slight challenges in exchange for an incredible gain in safety and knowledge of real-time, real-life conditions.

I don't see the potential for significant safety improvement compared to the systems already in place today. If there is a real-time ongoing event engineering staff can access telemetry from the ground already. This is not by any means a new capability.

How specifically is any given flight safer because telemetry is sent from all aircraft to the ground in real time?

Telemetry at my airline (and at many others) is downloaded when the plane stops at the gate. This is then available for analysis after any particular event. If there is no safety related event, there is no need for the telemetry to be sent in real time.

Sure, MH370, but that was literally a one in a billion event. And it since it seems likely the pilot or pilots were behind MH370, sending telemetry to the ground is a moot point in such a case since it implies the aircraft was fine.

Happy to be proven wrong, but right now I don't see it. Just saying there is spectacular potential doesn't make it so.

Pluto777 wrote:

We now live in ages that even the data from billions of iphones or ipads is back-upped into the cloud, but airliners still use data-backup into odd FDR's inside the plane ?
[...]
I do not understand this. It seems as aircraft-builders are not so interested in this kind of solution, maybe because it could reveal information that they rather prefer to keep unknown in certain cases ?

The functions are completely different. Your smartphone back-up is to recover your day-to-day data which you may need to efficiently perform your daily business. Thus there is a desire to get that data back as soon as possible after losing or damaging your phone.
FDR data is meant to support investigations carried out methodically using verified inputs and careful logical reasoning. The inputs include recorder data, but also any other kind of useful data like maintenance logs, metallurgical study of debris, weather data....Which means :
1. the investigation will take months in any case, and gathering all the inputs will take time too. So there is little need to have instantaneous availability of recorder data. You could argue that it could prevent an immediate repetition of the issue if it can be known soon enough, however this brings me to my second point
2. the most important attribute of recorder data is its integrity, not its availability. This means the data must be trustworthy, generated/transmitted/stored through known protocols which are robust to external disruptions. You cannot guarantee that with radio transmissions, nor can you guarantee that the data stored "in the cloud" is not modified

There have already been cases of accusations of tampering with recorder data physically encapsulated in isolated boxes. Imagine what a mess there would be if the data was floating anywhere & everywhere.

One could say that immediate data availability could help the search & rescue efforts, but you don't need the full FDR data set for that. There are easier ways to do this with upgrades to emergency beacons, maybe triggering them on detection of a problem as proposed by the BEA after AAF447. https://www.bea.aero/uploads/tx_elyexte ... t.data.pdf

But as the BEA themselves underlined, just this relatively easy task already required better satellite infrastructure. And that's because when the information transmission is really needed, the aircraft is in the middle of nowhere. Not in a developed urban environment bustling with communication antenna like your smartphone would be.
And therein lies the main issue. If this capability were to be mandatory, it somehow would have to be guaranteed to work when it is needed. If it doesn't, then what is the point ? Yet to guarantee that, you need to pay for expensive new satellites to provide the proper coverage with the relevant communication systems, and pay to equip all relevant aircraft with communication systems with the proper reliability.
All that would be spent just in the name of 2 events which required long recovery efforts. Yet the AF447 recorders were eventually recovered. As Starlionblue pointed out, you're left with MH370, which is an outlier. Better to invest in more pressing safety issues.

So as Starlion says, there is just little point in this expensive idea, from a pure safety point of view. The transmission of data in real-time will eventually grow however, as predictive maintenance starts taking off.

flyingclrs727 wrote:

Why not just keep track of where planes are using GPS to determine latitude, longitude, altitude, speed, and direction of travel? If an aircraft crashes into the ocean, it should be possible to locate wreckage pretty precisely.

GPS can't track anything. It doesn't work that way. The GPS satellites don't know where the receivers are.

If the airliners is transmitting its position through ADS or SSR, you're good. That's why the AF447 wreckage was eventually found. Barring that, you're left with MH370, where the surveillance systems were off for some reason.

flyingclrs727 wrote:

Why not just keep track of where planes are using GPS to determine latitude, longitude, altitude, speed, and direction of travel? If an aircraft crashes into the ocean, it should be possible to locate wreckage pretty precisely.

The problem has never been to determine position & speed of an aircraft. That has been done accurately and reliably for decades before the internet and smartphones even appeared
But apart from radar, most localisation techniques are run on-board the aircraft (technically called "dependent systems") and consequently the useful data is stuck on-board. The problem is getting that data to the people who can actually use it to direct SAR efforts, which are not on-board the aircraft but far away on the ground

Continuously transmitting data is no issue for your smartphone in an urban or semi-urban area of a developed country with antennas all over the place.
It is not so much of an issue for airplanes flying over such developed countries navigation and communication equipment all over the place. And there are often radars around, which can provide the information independantly from the aircraft

Transmitting data starts to be a problem, even for your smartphone, when standing out in the middle of nowhere. For example in any deep and hilly rural areas. Lots of place where the latest super-duper smartphone is next to useless because it has no signal.

It gets really tricky in the case of an aircraft over the middle of an ocean, 100s of miles from the nearest shore, probably going through bad weather, and most likely plummeting down with a random, rapidly changing attitude. Yet that is the case where the data is actually needed.
And to make it more difficult still, as airplanes are fast, the position has to be known till as late as possible, within a few seconds. AF447 sent a position transmission from cruise just 4min before impact, yet that was enough to result in a huge search area.

And that is why the comparisons with smartphones capable of switching on the lights at home etc... are completely besides the point.

Just to be clear, I don't mean that we shouldn't push for data transmission capabilities. It's just that such capabilities are most needed at precisely the time when implementing them is really difficult.

flyingclrs727 wrote:

Starlionblue wrote:

flyingclrs727 wrote:

Why not just keep track of where planes are using GPS to determine latitude, longitude, altitude, speed, and direction of travel? If an aircraft crashes into the ocean, it should be possible to locate wreckage pretty precisely.

GPS can't track anything. It doesn't work that way. The GPS satellites don't know where the receivers are.

If the airliners is transmitting its position through ADS or SSR, you're good. That's why the AF447 wreckage was eventually found. Barring that, you're left with MH370, where the surveillance systems were off for some reason.

Of course the aircraft would have to upload their tracking data to satellites. It's a much smaller problem than trying to upload all the black box data continuously for all the airliners flying around the world. In fact engine data for MH370 continued to be sent to the manufacturer periodically for hours.

On the other hand, with all the automation on modern aircraft, why can't aircraft monitor the status of the pilots? Couldn't avionics determine that the pilots are not talking to each other or that they might be disabled? In such a case couldn't the black boxes be programmed to upload their data?

Avionics already "check" on the pilots. If we don't touch any cockpit controls for a determined number of time, warnings start to sound.

As mentioned airliners already upload position data to ground based stations. This has been happening for years. Doesn't make a difference what the pilots are doing.

BawliBooch wrote:

How can we ensure data security if we upload all this sensitive data into the cloud? Can you imagine what will happen if it rains?

#PostLikeABhakt

I agree, I wouldn't trust my data in someone else's computer, besides it's probably a CIA run thing to keep tabs on everyone.

Because the cloud is lame...?

M564038 wrote:

Starlion, Real-time telemetry have spectacular potential for contributions to safety. Of course there needs to be secure encryption of this data as well as clear limits to it's use. This data belongs to the safety people and needs to be kept far away from the economy-department.

These are only slight challenges in exchange for an incredible gain in safety and knowledge of real-time, real-life conditions.

How?

People spouting this but no one ever posts how it will improve safety. The cost:benefit ratio is terrible.

Data and Voice recorders are build phenomenally well. But is there a way to make them better? Longer lasting, stronger pingers etc. Just throwing it out there. Better money spent I think.

Constant uploading doesn't seem like it will make things safer just more expensive to fly, and the data ? Why encrypt it? Its not like anyone has broken into the cloud before, but what's the value in it?

C.

EDIT: for additional thought..

M564038 wrote:

Starlion, Real-time telemetry have spectacular potential for contributions to safety. Of course there needs to be secure encryption of this data as well as clear limits to it's use. This data belongs to the safety people and needs to be kept far away from the economy-department.

These are only slight challenges in exchange for an incredible gain in safety and knowledge of real-time, real-life conditions.

As Wacker1000 asked - how? How would cloud storage "have spectacular... ...contributions to safety"? That is, airline safety is already amazingly good. How does creating another data stream and storage system improve upon the current system? I do believe that such a system is technically possible, but other than MH370, how does such a system improve today's level of safety? Simply stating that it do so, doesn't answer the question and is not an explanation.

After my years - admittedly some years ago - of flying "the pond" and not always having reliable SatCom connections, I'd be worried about depending solely on a data streaming and storage system.

I just don't know why people have to call it "the cloud" ugh. I mean, I'm only 32 so not exactly old - but I was building computers when I was 14-15ish, getting into data and IT transfer & systems when I was 17-18 with different network cert. classes the academy program offered in HS. They're just servers. Nothing has changed - just huge banks of computers on racks in a warehouse building somewhere. This 2015+ marketing term of "cloud" is rather funny. Okay end rant - anyways..

Basically, we could back up data to a server on-the-fly (no pun intended..) but there are definitely bandwidth issues to address. Not capability-wise, but cost-wise - ie. no one would ever do it if the connection and data fees for sat. data were astronomical (err again - no pun intended!)

I think a simple few lines of code with lat / long, altitude, and airspeed, perhaps a code system (a database of 3-letter codes for example, each pertaining to a situation, error, excessive vertical speed +/-, cabin pressure code, etc).

I mean how many combinations can you get with 3 digits alphanumeric? That would keep each update within a few KB of data. Not a ton of information - you aren't going to get engine parameters or anything, but who cares? If you can at least locate the aircraft and have an idea of its final trajectory in addition to the coordinates, you could actually get your hands on the FDR w/ complete data set.

Such a small amount of data would be capable of being integrated into updates VERY often. Someone smarter than me would have to figure out intervals - but ideally, you'd want the intervals as far apart as possible, while limiting the search zone as much as possible if a dramatic maneuver was performed after the latest update (ie. straight level @ cruise speed, sends an update, then dives and does an unusual 120 degree turn @ high speed) that scenario would yield a difficult search area based on last update, unless it was often enough that even at very high IAS w/ maneuvers, the search zone is still small enough to pick up an FDR / CVR ping quickly.

Honestly I'd imagine all the systems to do something like this already exist in most aircraft (at least overseas / long-haul) - it's just a matter of programming a simple module to integrate with the QAR or FDR and output some stuff through the ACARS link or other sat system. Plus certifications and integration, training, blah blah - but the simplicity of it all versus other options might sell some airlines, or certainly insurance companies.

estorilm wrote:

I think a simple few lines of code with lat / long, altitude, and airspeed, perhaps a code system (a database of 3-letter codes for example, each pertaining to a situation, error, excessive vertical speed +/-, cabin pressure code, etc).

Iridium and Aireon are already doing this on Iridium Next. https://aireon.com/resources/iridium-next/ It won't require anything beyond ADS-B which all airliners have/will be required to have.

USAirKid wrote:

estorilm wrote:

I think a simple few lines of code with lat / long, altitude, and airspeed, perhaps a code system (a database of 3-letter codes for example, each pertaining to a situation, error, excessive vertical speed +/-, cabin pressure code, etc).

Iridium and Aireon are already doing this on Iridium Next. https://aireon.com/resources/iridium-next/ It won't require anything beyond ADS-B which all airliners have/will be required to have.

I'm sure the aircraft will need some sort of communication gear to communicate with the Iridium satellites in addition to ADS-B. That will be a cost that the airlines may balk at if it's not mandated.

it has been noted that the data that is proposed to be uploaded is useful retrospectively ie after an incident
ie the vast majority of data from flights that were routine need not be uploaded in real time

but why not program the system to upload in real time in the event of an incident?
ie warning lights sudden changes in flight parameters or instrument settings
severe turbulence lack of communication between crew and/or between crew and ground

that way much less data is needed and in the event the flight eventually ends uneventfully it can be purged.
even if the flight lands, the data can be used and certainly if a flight is lost at least some data from
the time period immediately prior to the event is available in real time without scouring the ocean floor.........

entdoc wrote:

it has been noted that the data that is proposed to be uploaded is useful retrospectively ie after an incident
ie the vast majority of data from flights that were routine need not be uploaded in real time

but why not program the system to upload in real time in the event of an incident?
ie warning lights sudden changes in flight parameters or instrument settings
severe turbulence lack of communication between crew and/or between crew and ground

that way much less data is needed and in the event the flight eventually ends uneventfully it can be purged.
even if the flight lands, the data can be used and certainly if a flight is lost at least some data from
the time period immediately prior to the event is available in real time without scouring the ocean floor.........

And how can we decide what data to upload and what's not?
And until what time the data should be kept? I remember 737 rudder hard over accident and how they need to publish another reports, years after the accident.
Besides, call me paranoid, but with highly accurate real-time data available online, it's only a matter of time until people hack those data and design anti-aircraft missiles to access those data online and shoot them down.

RetiredWeasel wrote:

USAirKid wrote:

estorilm wrote:

I think a simple few lines of code with lat / long, altitude, and airspeed, perhaps a code system (a database of 3-letter codes for example, each pertaining to a situation, error, excessive vertical speed +/-, cabin pressure code, etc).

Iridium and Aireon are already doing this on Iridium Next. https://aireon.com/resources/iridium-next/ It won't require anything beyond ADS-B which all airliners have/will be required to have.

I'm sure the aircraft will need some sort of communication gear to communicate with the Iridium satellites in addition to ADS-B. That will be a cost that the airlines may balk at if it's not mandated.

I listened to the Airplane Geeks Podcast 445 Iridium Communications and Space 2.0 a few weeks ago, and as I recall its just the standard ADS-B signal that it picks up.. no added equipment needed on the airplane.

benbeny wrote:

entdoc wrote:

it has been noted that the data that is proposed to be uploaded is useful retrospectively ie after an incident
ie the vast majority of data from flights that were routine need not be uploaded in real time

but why not program the system to upload in real time in the event of an incident?
ie warning lights sudden changes in flight parameters or instrument settings
severe turbulence lack of communication between crew and/or between crew and ground

that way much less data is needed and in the event the flight eventually ends uneventfully it can be purged.
even if the flight lands, the data can be used and certainly if a flight is lost at least some data from
the time period immediately prior to the event is available in real time without scouring the ocean floor.........

And how can we decide what data to upload and what's not?
And until what time the data should be kept? I remember 737 rudder hard over accident and how they need to publish another reports, years after the accident.
Besides, call me paranoid, but with highly accurate real-time data available online, it's only a matter of time until people hack those data and design anti-aircraft missiles to access those data online and shoot them down.

what data to upload - anything out of the ordinary even a warning light. so the rudder hard over incident would be uploaded.
till when to keep - if there was any sort of incident even a warning light the data is downloaded from the cloud until whatever the law (statute of limitations or whatever other) requires maybe forever.
if they get a CVR or FDR and use its data how long is that kept in hard copy?
same for cloud download. except that data is accessible minutes after whatever happened.
not after days of searching a crash site or months or years on the ocean floor.

your paranoia is quite possibly spot on. but i wonder if there are not ways to make such a system one way only ie download. (from the aircraft)
no upload possible (to the aircraft)

maybe there is a way of having such a system be totally closed to the outside ie a part of the deep web (NOT the dark web)
many government financial and other institutions (from what i read) have intranets that can be considered part of the deep web and are not accessible from the rest of the world. do not know if such a thing exists for what we call the cloud, although even the cloud bottom line uses physical hardware. so maybe it is possible.

at least in Hollywood there was a CSI (i think) where a passenger hijacked a planes computers via the IFE system and then the CSI techs hijacked it back from the ground also via the IFE. but that is so far just Hollywood

Some aircraft manufactures don't want to modernize planes to the extent where they can be accessed from the outside and hijacked by a terrorist in his basement on a laptop.

~Zac

Zac agree but as i said wonder if it could be a closed system. one way download from aircraft and only in the event of a triggering incident. no upload possible to the aircraft. and one way download when needed (ie right after the incident that triggered the upload) from the cloud to a regulatory agency ie NTSB or similar in other countries. (once downloaded from the cloud it is kept by the agency and purged from the cloud) Not even accessible to an airline until after released by the agency. and NO OTHER access from the rest of the world. just wondering.

entdoc wrote:

Zac agree but as i said wonder if it could be a closed system. one way download from aircraft and only in the event of a triggering incident. no upload possible to the aircraft. and one way download when needed (ie right after the incident that triggered the upload) from the cloud to a regulatory agency ie NTSB or similar in other countries. (once downloaded from the cloud it is kept by the agency and purged from the cloud) Not even accessible to an airline until after released by the agency. and NO OTHER access from the rest of the world. just wondering.

I think you still need small amount of data to be uploaded for digital handshakes, even for the most secure one.
If we design the system to one way only, the airplane needs to broadcast it's data continuously, because it cannot determine the status of receiver, and the broadcast is simple enough to track (IMO).
That way I believe it's still not feasible, unless maybe we can rig some highly secure satellite constellation solely for that purpose, and let's not talk about the space junks yet.

Btw entdoc, are you ear-nose-throat specialist? Just wondering from your username.

Alas that I am

Uploading data to the cloud can help in cases for instance after an airplane crash the airplane can no longer be found like MH370, or if somehow even the blackbox get destroyed.
The data transfer rate needed should be minimal, for cockpit voice data it is just a voice data recorder which i don't know what spec it used but if similar to a mobile phone then that would only be about 20kbps. and for flight data that would usually get recorded into FDR the data rate seems to be a magnitude smaller than that. So about 22kbps. That's actually minimal amount of data for a flight with satellite internet or even just satellite phone on board, far less data usage than an ordinary passenger browsing webpages on the plane. As for data sensitivity, those are data that would be recorded into recorders anyway, sending them via satellite probably encrypted won't make a different i think.

Why not backup airliners data into the cloud

Because the cloud is supported by highspeed DSL, Cable to share 4096 x 2160 pictures of your dog or x264 movies with your girlfriend on vacation in near instant fashion.
A aircraft is supported by xx kilometres of empty air to reach a nearest station. It only has so much broadband to feed data to all the instruments and transmit it's own data to ground, or other nearby aircraft.

This broadband doens't support all the variables per second of data recorded by a FDR.
You would need a massive radio dish on your aircraft just to transmit and receive the necessary data.
Which will cripple weight/aerodynamics and then the airline company will go bankrupt soon.

So in short, bad idea!

semiless wrote:

Why not backup airliners data into the cloud

Because the cloud is supported by highspeed DSL, Cable to share 4096 x 2160 pictures of your dog or x264 movies with your girlfriend on vacation in near instant fashion.
A aircraft is supported by xx kilometres of empty air to reach a nearest station. It only has so much broadband to feed data to all the instruments and transmit it's own data to ground, or other nearby aircraft.

This broadband doens't support all the variables per second of data recorded by a FDR.
You would need a massive radio dish on your aircraft just to transmit and receive the necessary data.
Which will cripple weight/aerodynamics and then the airline company will go bankrupt soon.

So in short, bad idea!

In short, that's it. You don't even need to be way out over the ocean for a simple TAF report from four airports to take 5 minutes to receive. We're talking a couple of thousand characters. The infrastructure simply isn't' close to having the capabilities that most people think exist.

As I said above, while technically possible, uploading the proposed volume of data would be massively expensive. More importantly, the safety improvement would be marginal at best. Not a good use of funds.

Starlionblue wrote:

semiless wrote:

Why not backup airliners data into the cloud

Because the cloud is supported by highspeed DSL, Cable to share 4096 x 2160 pictures of your dog or x264 movies with your girlfriend on vacation in near instant fashion.
A aircraft is supported by xx kilometres of empty air to reach a nearest station. It only has so much broadband to feed data to all the instruments and transmit it's own data to ground, or other nearby aircraft.

This broadband doens't support all the variables per second of data recorded by a FDR.
You would need a massive radio dish on your aircraft just to transmit and receive the necessary data.
Which will cripple weight/aerodynamics and then the airline company will go bankrupt soon.

So in short, bad idea!

In short, that's it. You don't even need to be way out over the ocean for a simple TAF report from four airports to take 5 minutes to receive. We're talking a couple of thousand characters. The infrastructure simply isn't' close to having the capabilities that most people think exist.

As I said above, while technically possible, uploading the proposed volume of data would be massively expensive. More importantly, the safety improvement would be marginal at best. Not a good use of funds.

But many aircraft, especially intercontinental widebodies, already have necessary transmission equipped, and allow passengers to shuf the net at mbps speed over the ocean. Transferring data to the ground obky need a tiny fraction of this as explained in my post above.