Why don't aircraft black boxes just upload real time data to the net?

[amerikat]

Instead of recording it in a plane that has gone down.
Could real time data from each plane be broadcast to a
satellite and saved in a database?
Seems to me that it would be easier to Analise real time data than recovering
a black box from miles below the ocean surface.

[EFerrari]

1. Wouldn't that create security problems?

[amerikat]

3. Not sure what you mean.

It's just a data stream.

[cherokeeprogressive]

27. "just a data stream". The best challenge for a hacker I've ever heard. n/t

[Duer 157099]

2. Skynet!

Seriously. Don't worry, they'll get there.

[MadMaddie]

4. Yea, that creaps me out!

But the technology should be more advanced than it is. Airbus should spend the money to send a research team to get the box.

[amerikat]

5. data on the crash is better under the sea?

I'm just wondering why the data is .....nevermind.

[Submariner]

6. They'll find one or both boxes with an ROV

with a pinger receiver. If not, then seabed mapping will figure out the debris trail pattern, and then an ROV will survey the route. We have that technology pretty much nailed.

[nadinbrzezinski]

10. It will just take a few months, worst case

By the way, Remotely Operated Vehicle... not everybody gets the alphabet soup.

I wonder who will get the training? Worst case a few Deep Sea crews, assuming it did not go THAT DEEP, will get a training run or two.

[NYC_SKP]

16. Heard that the signal last only 30 days.

If true, they may never find it.

[Stellabella]

23. Plus it may be at a depth that is out of range of our technology.

It's at the upper limits, according to best guess right now.

[shadowknows69]

49. We've sent stuff five miles deep before

So I don't think that's a valid argument but it would still be looking for a needle in a literal ocean of haystacks.

[Captain Hilts]

15. Yes. The USN will find it. nt

[anigbrowl]

33. The French navy, actually.

France has an extremely capable blue water navy, and already have a ship on the way with two automated mini-submersibles. Trust me, the French don't have to take a back seat to us in technology.

[Captain Hilts]

41. Didn't say they did. But there's a lot of ground to cover in a month. There's plenty of work to go

around.

[MadMaddie]

26. I think I read somewhere the the ping will last about a month

I could be wrong. I will see if I can find the information.

[Sen. Walter Sobchak]

7. the technology has existed for many years

the issue is there are so many planes in the sky and satellites just don't have the bandwidth to relay all that information and radio isn't terribly reliable in many parts of the globe.

[amerikat]

8. I can get rickrolled but there is no bandwidth for airplanes?

No way.

[Statistical]

24. Are you connected via a sat connection moving at hundreds of miles per hour?

[nadinbrzezinski]

9. Because technically we are not quite there yet... and

have some holes still in satellite coverage.

That is why.

If you ever go on a cruise you will understand... as that is the BEST example any civie can get on just how many holes we still have in the SatNet... and the military and merchant marine gets priority.

[amerikat]

11. I guess we need to upgrade our sat coverage.

I grew up a few miles away from the worlds largest satellite test chamber.
It was RCA as I recall. Changed hands many times before it was
sold for scrap.

[Occulus]

13. What about networking the data among the planes themselves?

Thinking out loud here. What if each passenger jet were equipped with a system that would alert other nearby planes to their presence, with a 'carrier signal' of some sort that could transmit operating status? It could be set up such that if an unexpected 'hole' in the 'fabric' suddenly appeared, the position of the expected flight gets triangulated automatically. Such a system could have the added benefit of reducing or eliminating completely the 'near miss'.

It can't possibly be hard to transmit current arrival and departure times of all planes to all planes; that's a simple text log. Of course, this would only be usable under conditions in which there were many planes in the air, but over populated areas, that's no real issue.

Why haven't we put our massive computing power to work in air travel safety? Cripes- make it a literal data cloud using distributed computing, slap a dual quad-core system onto each plane, write the software, and deploy a system that takes care of all this!

This sort of system cannot possibly be completely beyond us....

[Statistical]

25. Most of the time the planes would be too far apart.

Developing a self healing mesh network where the nodes are traveling at hundreds of mph and often times may be separated by hundreds of miles isn't exactly easy.

[Occulus]

47. But is it *impossible*?

We shouldn't abandon the idea just because it's hard to do....

[Statistical]

52. Would it be useful.

While aircraft is too far away from another aircraft (out of range) it crashes. No data transmitted to another plane.

It isn't a technology limitation it is a there earth has lots of sq miles and relatively few planes on them. The majority of the time the aircraft would be out of transmission range with another aircraft.

For example this aircraft was over the horizon of all radar stations. It may have been over the horizon of all other aircraft. Nobody to transmit to.

The black box system works fine and FAA is working with international community to improve it substantially.

[sailor65]

12. To an extent, they already do

It's called telemetry, and Airbus has it. The problem is the size of the full stream. The other poster is right; the ability is not quite there yet to get the full stream (Which is huge), but the aircraft does transmit telemetry, particularly on exceptions to normal operation.

[amerikat]

14. The plane phoned home to announce a total electrical failure.

I guess it had a few electrons to spare?
I just don't understand.

[The Velveteen Ocelot]

17. A "total electrical failure" in an A330 would mean

both engine-driven generators had failed, but if that occurs a thing called a ram-air turbine drops down from a wing and spins to provide some basic electrical and hydraulic power. In addition there are batteries that provide power to a few functions, but if all you have is battery backup you have a lot of problems because the flight control computers won't work, and you have very minimal control of the aircraft. Battery backup "law" gives you just enough control to keep the airplane upright until you can get a generator on line. If the airplane had a significant electrical failure in the middle of a thunderstorm, though, control would be a problem.

[amerikat]

19. I'm asking why don't we get real time data.

Why do we look to recording devises that went down with the airplane.

[The Velveteen Ocelot]

22. I was simply answering the question of how the aircraft was able to transmit

data regarding an electrical failure. Other posters have adequately answered the OP question. Many newer airplanes do periodically transmit technical data back to the airline's operations control facility, but it's not in a steady stream. There really isn't enough bandwidth for all airplanes to transmit all the time.

[AzNick]

58. Because they did not upgrade the technology yet

Tech advances go by leaps and bounces.

In this case, this plane was designed 20 years ago.

They are far from perfect and composite materials may not be so resistant after all.

[anigbrowl]

30. Oh dear

Look, planes and other things have multiple power systems. Just like the electrical systems in your car run off power from the engine, but if your engine seizes there's still the battery. Only on a much, much more complex scale. You can't, unfortunately, design away all risk, but engineers do design with the possibility of failure in mind. That's why air travel, despite sad episodes like this, is still very safe.

[armyowalgreens]

18. That would be waaaay to damn expensive. The amount of data transmitted would be huge.

Just think about the thousands of planes all transmitting non-stop data from a huge amount of data points. They would have to build a very expensive state of the art system to handle that much data.

And on top of that, it would all have to be coded to prevent hacking.

[amerikat]

20. Create jobs and save lives...........

sounds good to me.

[armyowalgreens]

21. This is capitalism. The point is to make a money, not help people.

[anigbrowl]

32. Oh pish. Where do you think the flight data recorders come from?

Safety products are just another industry. That they're manufactured by businesses doesn't make them bad, does it? No. If real-time data recording as the OP describes were so easy to implement, don't you think they'd be marketing the hell out of it?

[armyowalgreens]

35. You don't think that these massive businesses have thought about that already?

Like the OP is coming up with some genius idea that hasn't been mulled over by many people in these companies or organizations? That would be odd.

The only reason they have black boxes is because of regulations. Otherwise, I doubt they'd have them.

If something isn't profitable, it is nixed. That's how capitalism works.

Unless the FAA or other organizations mandates real-time data collection, there will be none. Assuming it doesn't somehow become profitable to do so and I highly doubt that will happen.

That is the cold hard reality of capitalism.

[anigbrowl]

36. Think again

Flight data recorders were invented in France, and different designs invented later in Australia, in both cases by private companies which are still players in the market today. They marketed their invention, and at first aviation authorities were indifferent. See http://en.wikipedia.org/wiki/Flight_data_recorder for a somewhat more detailed introduction to their history and use.

Another pioneer was James Ryan, who received the first patents for such devices in the US (privately, although he also taught engineering at the University of Minnesota) and lobbied Congress to require their use. He was also the inventor of the modern seat belt, and lobbied for the creation of the NHTSA.

Government agencies can and do call forth invention on behalf of social problems (a particularly famous example being the British government's sponsorship of a prize equivalent to several million dollars for a method of calculating longitude at sea, which led to the development of modern clocks). However, a great many inventions are developed privately and then offered to the public in hope of a profit - aeroplanes themselves being a case in point, when they turned out to have utility in World War 1.

Meantime, this idea is not likely to happen any time soon, not because of some flaw in capitalism, but because of basic economic: the engineering requirements are massively complex, while the ratio of crashes where the black box is unrecoverable relative to the number of flights taking place is something like 1:10 million. So you're looking at an extremely marginal increase in safety. We would get a greater improvement by spending more money on more training, recruitment and pay of pilots and flight officers.

[armyowalgreens]

37. You missed my point entirely.

Did I ever say that they weren't manufactured by private industry?

I said THE AIRLINES make no money by having black boxes. Every action they take is to make money or to appease the FAA.

If the FAA wasn't forcing them to put black boxes on planes, I doubt they would use them.

[anigbrowl]

45. You didn't bother to make it clear who you were talking about.

Airlines fly people for profit, and safety companies sell thing things for profit.

[Statistical]

28. It wouldn't save lives.

The majority of black boxes are recovered.

The majority of those that are not recovered are because it gets crushed not lost. Starting in 2011 a new blackbox system will be deployed and a backup blackbox will be used.

Instead of a seperate flight data recorder & cockpit voice recorder the 2 boxes will be combined and cockpit video is being considered. Everything on the first box will also be recorded on a se3xond box. In a crash situation the second box will be launched clear of the aircraft and slowed by gravity parachute.

This means it is unlikely both boxes will be lost. Hell after this maybe they will make the second box float.

The system you are describing is essentially economically impossible.

Everyday there are 5,000+ commercial passenger aircraft in the air at one time.

Bandwidth requirements would be immense. Like off the charts immense.

Communication ka band sats like spaceway-3 cost about $350 million each to build and launch. To cover the globe would take 36+. So setup cost is in the $10 billion range. Now you need to equip the 20,000+ commercial aircraft on the planet. Cost for an uplink that is FAA rated could easily run $100K each so thats another 2 billion or so.

Even worse would be issues with dopler shift due to the high speeds and changing angle between plane & sat.

Of course such a system would take up 36+ highly desired geo-sync orbital slots around the globe.

The system would fail around major airports. The number of aircraft their would overwhelm the system.

Likely you would need to put extra sat above major hubs to acts as a backbone.

The sheer engineering required for the uplink to track all 36 sats and then know when to switch sats and located and lock on quickly is a challenge.

Lastly.... rain fade. ka band is subject to signal loss in heavy rain. Like in a thunderstorm.

This doesn't even get into the concept of limited resources. Even if you had $20B or so to get this system up and running why would you? You could actually prevent crashes by using the $20B to improve maintenance standards, upgrade our aging air traffic control system, upgrade radars, improve pilot training and testing, improve mechanic training and testing, improve FAA simulations, improve compliance and auditing or build a system nobody needs and nobody wants.

[anigbrowl]

31. Exactly

I think $20 billion as an estimate is on the low side, to be frank. But even going with this, losing both black boxes etc is so rare that your safety increase is going to be maybe 0.0001%. Planes will still crash for unforeseen, and unforeseeable, reasons.

[anigbrowl]

29. I had a discussion with some engineers about this earlier today

The problem (in a nutshell) is that the FAA would be trying to deal with ~20 gigabytes per second of incoming data, or about 700,000 individual data streams (there are 88 things that get recorded on a plane's flight data recorder, and ~8000 commercial flights just in the US at any given time). If you consider worldwide flights, the data flow would be about triple that. Managing it all in real time would be a horrifically complex engineering problem (especially given that your data sources are flying around at about 500 mph) and would require an open network of the kind that would be very easy to attack, and which the military (understandably) would not want to be part of.

There is radio transmission of critical data already using a system called ACARS, but that's very very limited compared to the quantity and resolution of what gets recorded on a black box. It is fundamentally possible (that is, the laws of physics don't prevent it in any way) but to do it on such a vast scale would be extremely complex and expensive. The thing is, relatively few planes crash compared to the number of flights that take place, and in most cases the black box is recovered when they do crash. A crash like this where almost everything is lost beyond hope of recovery is extremely unusual - so unusual that coming up with a completely new system may not make any measurable improvement in air safety.

[krispos42]

40. What about some kind of "burst" transmitter?

Operating through the conventional communications system? Say, every minute or every 5 minutes the plane could upload a compressed data file of the previous, say, 15 minutes (so there's an overlap, you see) plus a GPS position/altitude/direction/speed fix.


If it takes 10 seconds to upload, you're cutting bandwidth by a factor of 30.


:shrug:


Just thinking out loud.

[Baclava]

43. Exactly - this is an extreme "worst case scenario"

Recording all the parameters that CVR/FDRs do and transmitting them via satellite or ground station would require unreal amounts of bandwidth, to say the least. It's just not economically viable; most of that data would never even be examined, because let's face it -- the only time they're needed is in times like this. It's possible, in theory, but costs would be (literally) astronomical. It's just not worth spending so much money for such a small gain.

And again...GPS is one way from the satellites to the GPS device. Nothing goes from the device back to the GPS satellite.

What we see on TV as GPS tracking is from the device to the cell phone network. It is the cell phone network tracking the device, but using the GPS determined position of the device for accuracy.

This aircraft was far out of cellphone range.

Even if the plane did break-up at altitide I would be surprised if some of the ACARS messages received from the aircraft did not have a pretty precise location.

But still, after falling from 35,000 ft - the wreckage impact with the ocean could be 10 or more miles from the last reported position.

So you are looking at a circle at least 50 miles in diameter even if you have a perfect position report.

Transmitting position data is of little use to an airline. If an aircraft is hurtling into the sea, transmitting position data to Paris doesn't matter because there is nothing they can do about it. Last known position data is not really necessary because it is fairly easy to estimate where the aircraft went down (which does not happen very often anyway) and as demonstrated in the case of AF 447, debris can usually be located within a day or two.

[trthnd4jstc]

34. You are onto something, yet please consider this...

The GPS Satellites are quickly filling up, and further, there is starting to get to be too much space junk, where a good portion of our GPS Satellites are now in orbit. But yes, we can have real time GPS tracking of planes. I think you are right.

Analyze.

Spell Check.

Just a suggestion.

Anyway. Space, the necessary expansion of the Human Population into the Intra-stellar space, decreasing the Earth's population of Humans and having industry (jobs) in Space to keep the people occupied and continuing to work on our progress as people...

Whatever.

[Canuckistanian]

38. GPS satellites are "filling up"?

I don't think you understand what GPS satellites do.

The satellites are transmitters or "broadcasters" ONLY. They don't get information or take "requests" from land based GPS units. It doesn't matter how many people are using GPS - it makes no difference to the satellite.

All they do is transmit their position and a very accurate time. Using 4 or more satellites, a GPS receiver does some math with those 4 signals and can give you your exact location.

[krispos42]

39. You could... as long as everything is operating normally

However, trying to maintain a datalink when the plane is doing all kinds of weird aerobatics is virtually impossible. Hell, a stiff breeze can make satellite TV signals go off the mark.


We'd know how the plane was doing right up to the point that everything turned to shit. Past that point... :shrug:

A satellite link would be a find suppliment to the conventional black boxes. After all, parts of the trucking industry has GPS real-time tracking devices for their fleets. But you still need something that will independently record radio transmissions, cockpit audio, and airplane systems status.

[MicaelS]

42. It's the oldest answer of all, money.

Look at aircraft accident statstics

http://www.planecrashinfo.com/cause.htm

Airliner 4.03 Fatalities per million flight hours
Commuter Airline 10.74 Fatalities per million flight hours
Commuter Plane 12.24 Fatalities per million flight hours

It's simply not cost effective to do so.

[mnhtnbb]

53. Bingo. This why cockpits had not been secured prior to 9/11 even though

it had been recommended.

[Edweird]

44. Why don't we just make the 'black boxes' the size and shape of the planes?

:silly:

[Baclava]

46. Speaking of the underwater topography of the region where the plane fell

I found this...

Go to the Brazilian version of the wikipedia page of the Sao Pedro Sao Paulo Rocks, accessible from the English one

http://en.wikipedia.org/wiki/Saint_Peter_and_Paul_Rocks

Then scroll down to Morfologia Submarina. There is a 3-D plot of the ocean floor mapped by latitude and longitude:


The wreckage is located at W-NW to NW of SPSP Rocks - about 90 miles from it. The ocean depth is 3000 to 4000m.

Mountainous portion of the mid-Atlantic ridge - no flat seafloor here. Difficult terrain to say the least, if indeed this is where they are looking.

[nadinbrzezinski]

57. ROV (remotely operated vehicle) or deep submersible territory

somebody will get a practice run SOON, and yes that might very well be the US or British DEEP Sea, or the French DEEP SEA team... and yes them training runs are good for all... perhaps all three, given the extent of the terrain.

[RamboLiberal]

48. Article on the idea

Richard Healing, another former NTSB board member, said Wednesday that a burst of real-time data transmitted in emergency situations "is a very good idea" and "technically, it's totally possible today." According to Mr. Healing, some helicopter fleets rely on sophisticated maintenance-management systems that keep track of unusual inflight events and "can notify an operator to ground an aircraft or complete a specific repair before flying again."

Potential obstacles include the cost of changing the software and the possible need for more satellite-transmission capacity, which can be expensive. Monitoring information as it arrives would also require more sophisticated systems than those used now for looking at maintenance updates hours or days later.

With surveillance planes and an armada of international ships gathering to try to find the black boxes, industry officials already are looking ahead to extract lessons from the tragedy. Airlines may opt to install more-expensive but more-capable models of weather radar, and they may look at ways to stream real-time weather information into cockpits.

Brazilian government officials and safety expert familiar with the details said the plane was beyond the reach of ground-based weather radars but had no ability to receive real-time weather data from satellites. "This storm blew up like an atomic bomb," one safety expert said, and the crew apparently didn't realize the extent of it earlier in the flight.

http://online.wsj.com/article/SB124402430768380893.html

Heard discussion two of idea of black boxes that would eject on impact & float.

[Baclava]

50. "Heard discussion two of idea of black boxes that would eject on impact & float" Why?

So they could float away with the currents? Planes don't just go plunk right to the bottom anyway. There is always some location error involved....and all the pieces are important.

The present system works just fine - maybe it's just the short-attention span public that wants everything wrapped in a few days or they lose interest?

Experts have stated that the black boxes will signal their location for 30 days, but the recorders themselves can last indefinitely underwater, even at extreme depths in salt water. With all the modern equipment available, they still have a good chance of locating them.

(edit)

Let them do their jobs - we'll wait.

[Statistical]

51. They misunderstood.

The #1 reason for lost black box more than all other reasons combined is it gets crushed.

Current system has a seperate box for:
flight data recorder & cockpit voice recorder.

This was a limit of 1970s technology. Today solid state drives enable much more storage and one box can record everything (plus even more sensors). There is even talk of recording flight video (camera behind pilots recording actions, and view out cockpit of horizon).

New system will be 2 boxes per plane. Both identical in storage (voice, data, video).

One remains in the plane, one ejected from plane.

The probability that both will be lost, destroyed is essentially 0.

The system hasn't been finalized but FAA is pushing for it and is hoping something in 2011 timeframe.

[Baclava]

54. No system is perfect.

But progress is always appreciated. The public has a right to know what happened.

Maybe this episode will get the airlines moving forward here.

[Angleae]

55. The FAA might be pushing for it but the airlines will be pushing back.

The airlines are broke and the FAA expects them to modify two systems that would cost about $100,000 and up per plane.

[Pavulon]

56. Solution. (i want my cut bastards) Privacy, and telemetry

data is black / red separated (cleartext / cyphertext) on the aircraft. Data is written to box and then bounced out of plane after encryption. Key is generated by FAA / NTSB only to be decrypted in event of crash, requires 2 party agreement. That way the pilots do not have to be hen pecked. So a key is generated by a trusted agency (non airline) the data in encrypted and transmitted using aes 256 or better. This is one way transmission, there is no "hacker" aspect because it is just a stream of data. even in clear text it presents no threat.

In event of crash the primary key is used to decrypt the data for ntsb investigation.