I don’t know very much about this … but I do remember some years ago that a Voice Record had “rat’s nested”, meaning the mag tape was tangled up and was unusable. So, does anyone know, exactly, how this data is stored these days?
Logically, if it was cost effective to record video for the benefit received, the airlines would do so. But if it costs more to save a life than what that life is worth, then the airlines won’t. What would a video show that’s not already being recorded?
My point is that your “off the shelf” comment is out of whack. The components have to be a LOT more durable than what is provided to typical consumers.
That would be helpful for some crashes, like Continental 3407, but not for, say, United 811. Yes, I like Air Crash Investigation too.
Why? Currently, there is *no *video record of the cockpit interior. Even the cheapest consumer recorder (like one of those dashcams) would work some of the time. Sprinkling the aircraft with many of them would increase the odds even further. Obviously the equipment has to be qualified for aircraft use, but that is orders of magnitude cheaper/easier than building the unit to black box standards.
Your comment is a perfect example of “the perfect is the enemy of the good”. It may well be true that a video system as survivable as the current black boxes is not cost effective. That says absolutely nothing about whether less survivable systems could be cost effective or not.
The data does no good if it doesn’t survive the crash. By your reasoning we should just look for surviving cellphones and consumer grade cameras in any crash because why should the airplanes pay for what the passengers carry on for free? Which, I’m sure, is something already being done.
Lots of crashes don’t result in a fire or even significant structural damage. Even in relatively minor incidents, cockpit video would be useful in improving CRM best practices. There may be as much value in learning what to do (in incidents that ended well, like Flight 1549) as opposed to what not to do.
I’m sure you’re right that crash investigators look for cell phone video. None of them have a view into the cockpit; none have the potential for looped recording of the past N hours; none are properly mounted for good quality video; etc. Those silly Russian dashcams, however, have proven themselves useful in all kinds of ways (even when unintentional).
If the video recorder doesn’t survive X% of crashes, it means the data set is X% less valuable, not 100% less valuable.
Flash memory is quite durable, even in consumer grade devices, especially compared to mag tape, which was obsolete for that purpose 10+ years ago. And storage is less important if you frequently and continuously send the data to a remote site. This would have made a significant difference with solving the Air France crash in the Atlantic a few years ago; even the limited data transmitted in this fashion was a big help.
I don’t have too much to add. I think basically cockpit voice and flight data recorders had a huge benefit to flight safety so it was obviously a good idea to install them. Video recorders would only have a minor benefit as you can generally tell exactly what has happened from the combination of switch/control position from the FDR and the voice recording. Cases where it is important to know what the crew are doing when they are both not talking and aren’t manipulating the controls or operating switches are rare. With that in mind there just isn’t enough momentum for the idea of video recorders to overcome the general resistance to change in the industry.
That said, many aircraft do already have cameras positioned in various places. Ours has a camera in the forward galley so we can see what the cabin crew are doing. I’m not sure what the intended purpose of the camera is, we tend to use it to see if it’s a good time to interrupt them with a request for coffee/tea or to check that they are secured for landing. None of it is recorded, it is a simple CCTV.
Dr. Strangelove, the problem is that those systems aren’t free, they cost money to develop, install, and maintain. Unless the benefit outweighs the cost it isn’t going to happen.
With many/most commercial jets connected to the internet, they could conceivably just stream the data down to land somewhere, encrypted, at some government headquarters. No need for a black box at all. Now if something goes wrong with the transmitting equipment, etc, you’d need the black box.
Back in the day the CVR was a loop of magnetic tape similar to an old 8-track tape in operation. The FDR was a long roll of aluminum foil which was etched by pointers something similar to an seismographs that they show after earthquakes. Now days they are on more modern media such as solid state hard drives.
Left to their own devices, the airlines wouldn’t even have CVRs or FDRs. Government regulation is what forced such devices on aircraft. I’m not a big fan of government regulation, but there are times that it’s necessary. Finding out what caused a crash can prevent future crashes.
I would love to have a cite on this.
It is fascinating.
The OP’s question is why. Obviously such a system would be very helpful to investigators.
Do you have some connection to the industry? Because I’m not understanding your point either. Even the cheapest cell phone would probably survive a catastrophic crash inside current black box protection. It’s not the technology so much as the layers of protection it’s encased in.
This is why a video would be totally unreliable. If it showed both pilots humping the stewardesses, well, that’s one thing. Otherwise, it leaves far too much open to interpretation to be of any value.
I don’t believe anyone has made a really good case that it would be very helpful to investigators. There is one case mentioned where knowing whether the Egyptian pilot was purposely putting the plane in a dive would be helpful to understand what happened. But beyond that, what additional information gathered from video would be instrumental in determining the cause of and prevention of crashes?
Every piece of information has a potential use. I can dream up a scenario where a pilot’s arm brushed against a control, possibly a device who’s position isn’t recorded, and a video would reveal what happened. Video can also make it easier to tell who is speaking in a harsh and noisy environment.
Government agencies have been doing audio recordings for years. Now some are doing video as well; sometimes facial expressions or human, non-verbal reactions can augment the audio. Saying “audio is good enough” is like saying “No one needs more than 640K RAM,” or “Why record a meeting? We have the minutes!”
I’m not sure I understand the distinction you’re trying to draw. To paraphrase what I think your point is: “We don’t need video because we already understand the cause of crashes well enough without it”
My response is that I bet some of our current understanding is wrong, and we just don’t know it. There have been lots of cases where we think we know what happened, and new evidence comes to light that proves we were mistaken. Look at the results of applying DNA testing to past murder convictions. We “knew” what happened well enough to convict someone, but adding additional data indicated that, actually, we were incorrect.
My suspicion is that adding video would have a similar result.
I wasn’t thinking of that. I was thinking of the case where, in an emergency, the pilot misses seeing a particular warning because he’s looking elsewhere. Eyetracking studies have produced all sorts of cool results for optimizing the design of information display systems. And, of course, we’d see some where the pilot is asleep or otherwise obviously not paying attention.
I don’t understand this way of thinking at all. Obviously, a video is never going to give us 100% reliable knowledge of exactly what happened. But it’s clearly better than not having video.
Video would not be “totally unreliable”. Video would explain some things and fail to explain some other things.
My US$0.02 is that the data recorded is captured from instruments that are already in-line with the rest of the electronic systems required for flight. Information on altitude, attitude, rate of climb (+/-), etc. even communications and navigation data, (all necessary for a flight) are easily-captured as these things are already *electronic *data. Video would be a different source of information, taking analog information (reflected photons) and digitizing it to send to the recorder (not difficult at all; just a separate source).
Here’s the really critical part, with the really really critical part bolded:
“…these occasions [when only video could tell us something that would save lives in the future] are likely too rare to justify the added expense.”
I laid out the two all-important questions earlier in the thread:
How many lives would cockpit video be expected to save?
How much would it cost to develop, certify, install, and maintain those systems fleetwide?
If the dollars-spent-per-life-saved is higher than some arbitrarily chosen set point, then the measure is not justifiable.
It’s not that we don’t “need” cockpit video (whatever “need” means in this context), or that it wouldn’t be useful. It’s just that it needs to be useful enough in order to justify the costs involved.
Imagine, for the sake of argument, that a hypothetical crashworthy video recording system required each aircraft to sacrifice an entire row of seating. Now each aircraft loses $6000 worth of passenger revenue every day, or $2.2M per year. For the entire commercial aviation fleet, that’s…a shitload of money. Billions of dollars of lost every year, for the benefit of video footage from the cockpit. If the information gleaned from video of one crash results in policy or engineering changes that prevent one future crash, thereby saving 100 lives, the net cost per life saved is:
[billions of dollars] / [100 lives]
Is the cost of such a system justified? Depends on the value you assign to a human life. I suspect most would feel it’s not a good investment.
Clearly this hypothetical giant video recording system is unrealistic, but it serves to illustrate the decision-making process: you can’t just say “this system provides us with useful information, therefore it’s worthwhile” without considering the costs involved.
I have no problem with your proposed analysis, but I can state up front that the costs involved are likely to be minimal. For $300 retail, I can buy a GoPro camera that takes up 1/10,000 of your hypothetical single-row of seats, and provides exceptional video quality. Redesigning for multiple cams and industrial quality is a trivial task for the right manufacturer.
Spoken as someone who has designed (roughly) similar (military-grade aviation) systems many years ago, so I have a little expertise, but not current. Since the trend is downward in cost, I feel confident that such a system is incredibly cheap compared to life costs.