Title is essentially it:
Shouldn’t commercial airlines duplicate their “black boxes” in the cloud, in real time?
If not, then, why not?
…and if so, then what’s so hard about just doing it?
Seems to me an average satellite phone could manage it.
Is it all just politics?
A limited amount of information is uploaded in (somewhat) realtime, via ACARS.
Flight data recorders and cockpit voice recorders are designed extremely conservatively, and retrofitting new systems to them isn’t done lightly. Just saying “stick a satellite phone on it” isn’t really adequate, given the breadth of telemetry data recorded on modern aircraft FDRs.
It is an incredibly conservative industry and anything fitted has to be certified which can cost a lot of money. There has been talk of various solutions but nothing concrete yet. You are correct that it would not be difficult from a technical standpoint.
The newest black boxes record an incredible amount of detailed data, continuously. If this data was transmitted to the ground on a real-time or near-real-time basis, the traffic would be tremendous and the bandwidth needed, horrendous.
At least with today’s technology.
It’s actually not that far out a concept:
Bombardier and other companies have proposed similar technology. Mostly used for tracking aircraft in flight though. Generally, air crashes, particularly ones where they can’t recover the black box, are exceedingly rare.
This, pretty much. The basic technology (satellite comms) exists to do this, but tons of money would be spent on R&D to develop versions of hardware suitable for the many different aircraft that exist. That R&D includes not just making sure it basically works - it includes making sure it’s reliable, doesn’t ever interfere with any of the other electronics systems on the plane, has an acceptably low weight, and passes whatever certification hurdles the FAA might present. Once installed, there will be a program of inspection, maintenance, repair, and replacement, creating additional costs.
That all adds up, and in order to be worthwhile, the value you expect to get out of it has to be worth more than the expected cost. So the next question is: what is the expected value? It’s the number of lives you expect it to save multiplied by the value of each life.
You say you can’t put a price on a human life? Bullshit. That’s exactly what the FAA does:
So the last question becomes: how many lives do you think this modification will save over X period of time? Having studied the problem as intently as possible from my very comfortable armchair, I would have said that it’s not nearly enough to justify the cost (but, see below…). The only situation where live-streaming black-box data would provide a benefit are those crashes in which the old-school black boxes were either destroyed or unrecoverable. I can only recall one flight in my lifetime where this happened, MH370. And even if that flight had been live-streaming black-box data, it’s unknown how many lives might have been saved. It’s likely the water landing wasn’t survivable; if it was, the EPIRBS on the life rafts would have told rescuers where to look, so nobody on that particular flight was going to be saved anyway. Would black-box data from that flight have allowed us to learn something that might have prevented future loss of life? Maybe, maybe not.
As it turns out, MH370 did spur some changes. Live-streaming black-box data isn’t explicitly going to be required, but it’s apparently going to be allowed as one possible solution to the requirement that black-box data be somehow recoverable in the event of a water crash:
One thing to consider is the cost that would be saved if flight recorder data was streamed. If you’ve already got the data, you don’t need to spend the money to find and recover the boxes among the wreckage. When Air France 447 crashed in the Atlantic eight years ago, the black boxes weren’t recovered from the sea floor until two years later. I’m sure that wasn’t cheap.
Of course, the black boxes aren’t the only reason to find, recover, and examine the wreckage of a crashed aircraft. There are clues like metal fatigue and chemical residue that can help investigators. But finding the black boxes wouldn’t be crucial the way it is now.
Though you can certainly imagine circumstances where the streaming was cut off right when the data got interesting and useful, due to power loss, damage to the airplane, maneuvering, etc. so you’d still want to recover the black box in those cases.
A satellite-based solution would be required, since a significant chunk of commercial airline routing (and, arguably, the flights most at risk of disappearing without a trace) is beyond line-of-sight of a land-based solution.
Current satellite bandwidth is both limited and expensive.
At any given moment, there are thousands of commercial aircraft airborne worldwide (as I type this, flightradar24.com shows over 12,000 aircraft airborne around the world). A dedicated constellation of satellites would have to be designed, built, tested, and launched to support the bandwidth requirements of thousands of aircraft transmitting a real-time telemetry stream, without interfering with other satellite-based communications.
Of all the hair-brained ideas* I’ve ever heard this isn’t one of them. I assume this kind of thing is in development. All of the data that is recorded doesn’t need to be transmitted, but as much as is practical would save long searches for downed planes, and some problems that could affect other planes could be detected before a long and sometimes dangerous recovery effort.
*Or is it hare-brained? I think there was a thread about that.
According to Quora, 12 Gb per aircraft per flight would be required. Remember that not all aircraft have FDRs, though. You could also further alleviate that load by using terrestrial radio and only use sat uplink when out of range of other comms and still have the FDR/CVR on board as a backup. It’s doable but to get over the regulatory hurdles will be the hard part.
Slightly off-topic… but common sense says that the FDR and CVR data streams should both go to the two recorders. This way you have some redundancy… if one is destroyed you still have all the data in the other.
Historically this may not have been possible but I’m sure that it’s possible now.
It’s virtually certain that the MH370 flight plowed into the ocean and killed everyone onboard. So realtime data streaming wouldn’t have saved them. However, something went wrong on the flight. It might not have been pilot error, there could have been some type of equipment failure.
But, whatever happened the MH370 hasn’t been observed in any other flights. (some event requiring them to turn off transponders and possibly become incapacitated). So it must be extremely rare. And there’s a base 80% chance it was pilot error.
Let’s assume that because the FDR for MH370 wasn’t recovered, there’s a 50% chance another plane will be lost, and 150 people will be killed.
So you multiply the 0.5 * 150 = 75 people. At 6 million a head, the corrective measure needs to cost less than $450 million dollars.
It wouldn’t. It would cost more. Satellite bandwidth is expensive, and retrofitting existing aircraft with the additional avionics would cost hundreds of thousands each.
That 12GB number is highly suspect, even if it had been stated as an average (which it was not, at least the way I read the Quora post). An A380 flying DXB-JFK is going to collect loads more data than a CRJ-200 flying OKC-MSP.
14 CFR 91.609 requires flight data recorders and cockpit voice recorders on US-registered aircraft that have 20 or more passenger seats, or those that have six or more passenger seats, are turbine-powered, and require two pilots (basically, every aircraft that fits the commonly-understood definition of “commercial airliner”). Additionally, most business/corporate aircraft, including helicopters, have CVR/FDR installed due to insurance requirements, even if they’re not technically required by 91.609 to have them installed.
Thanks, everybody. Just checking back in to let you know I haven’t abandoned the thread. Very insightful replies.
Somewhat off topic, but is there any headway on preventing pilots from turning off their transponders?
It’s kind of related, because I’m somewhat convinced that the MH370 flight was deliberately flown into the Indian ocean until it had no fuel and was beyond recovery. Also the 911 hijackers turned off transponders.
What if the black box just broadcast an update its position every five seconds or something. Wouldn’t that at least make finding it easier?
It’s my understanding that a major reason why not is that the pilots won’t stand for it. A but of googling got me an article touching on that.
There are a couple of reasons we want to be able to turn off transponders.
If the electrical bus it is connected to has a major problem then that bus needs to be able to be turned off so that any chance of fire is reduced.
Sometimes (rarely) transponders are faulty and give bad information. This is bad for controllers and pilots alike as incorrect position and/or altitude information would be displayed on the radar screen and in the cockpit. In this case the transponder needs to be turned off.
If you thought about it long and hard you could probably get around these issues, but it’s not as simple as just hot wiring the transponder into the electrical system. The question then is whether it is worth the time and effort to design transponders in this way just to prevent very rare accidents? Would it even have prevented the accidents? If MH370 had its transponder on it could still have trundled on down to find a quiet piece of ocean to crash in. We just would have had a better idea of where to look. Likewise the 9/11 aircraft probably still would have been able to complete their mission.
That article was written 6 years ago by somebody with either no clue or with an axe to grind.
I’ve been flying professionally for nearly 30 years. I have never seen nor heard of a pilot pressing the erase button.
Rather the opposite. We have procedures to pull the CVR circuit breaker after parking following a number of different incident types. Precisely to ensure the recording isn’t overlaid by the next 2 hours of silence while the airplane sits at the gate. The CVR breaker is specifically marked to make it easy to find in the giant array of circuit breakers.
And before you ask, no, I’ve never seen nor heard of a pilot pulling that breaker inappropriately.
In addition, all FDR data now *is *downloaded and analyzed for statistical performance. If the crew force collectively is doing something wrong, it’s noticed and fixed. If one particular flight does something enough out of the ordinary there will be an enquiry.
The critical difference versus that proposed legislation is that the enquiry is done by a joint team of FAA, airline management, and union people. With the intent to discover what happened, learn from it, and discourage it from happening again. Both to that crew and all others.
If the finding is that crew was screwing around, they are fully liable for FAA & management punishment, up to and including permanent loss of job and/or license to fly at all. If OTOH, it was an innocent mistake or lack of skill then appropriate retraining gets done.
Rest assured nobody sane wants to go through that harrowing and embarrassing process.