I see a lot of investigations spend years digging through the black box to determine things that would have been obvious if a camera was involved in the cockpit.
In: Technology
As a preface… I am an aeronautical engineer. I’m an airframe guy and not a systems/electrical guy so my ballpark estimates are going to be a bit ballparky, but a lot closer than “rando internet dude”. There is a lot of things that come together to make it difficult to make such a system happen.
First off, any hardware design that goes on any aircraft has a lot more review than “regular” hardware. Any parts that go on an aircraft will have a documented paper trail available that goes back to at least the raw materials coming into the place that made the part, with anything that’s a critical component likely having documentation back to where it came out of the ground. Just this paperwork/tracking and quality checking alone can make something like a screw cost 140 dollars. [Example](https://www.reddit.com/r/mildlyinteresting/comments/bovm1m/three_screws_aircraft_grade_that_cost_13699/)
A $100 dashboard camera is likely a midrange unit designed to sit on a suction cup on your window. A cockpit camera would need it’s own installation hardware, connected to critical power lines like all cockpit hardware, and have a dedicated couple of data lines back to the 2 black boxes. The hardware would need to be vibration and shock proofed to a factor of 100 past your window mounted hardware. The camera setup with associated framing hardware for installation in a cockpit would be a few thousand dollars and the wiring harness additions would be in the low 5 figure range for a commercial aircraft.
Current standard Cockpit Voice Recorders (CVRs) hold approximately 2 hours of audio. The NTSB has been trying to push airlines into a 25 hour standard, but there has been pushback from pilot’s unions, airlines and aircraft manufacturers. Unions because they don’t believe that airlines need to have ears on what the pilots have to say, except for in the case of an accident, and 2 hours should cover most anything in the timeframe of an accident. Airlines aren’t hot on the idea of updating black boxes because it’s an expense. Current gen CVR’s cost between 10K and 40K per unit, with 2 units per aircraft. 12xing the storage on each unit has been estimated to double or triple the price per unit, since much of the cost isn’t the actual memory, but is in the associated hardware of the unit that lends to its survivability and radio beacons to recover. Each aircraft suddenly needing 40K to 240K of upgrades is pricy. Finally the manufacturers don’t want big changes because they’d have to certify that the new hardware works interchangeably with the old, OR provide the repair/retool procedure to make it work. Again, significant expenses.
Video files are AT BEST 3X the size of comparable length audio files, assuming you’re “only” going to something like 1080p. So if we’re keeping the video files for only 2 hours, I’ll be generous with estimating and say that the CVRs would only need replacing at base cost since it’s not the 12x of the storage only 4xish the size.
This would put my VERY rough estimate of adding only one 1080p cockpit camera, saved to the black box for a 2 hour timeframe, at around $40K per aircraft in hardware costs alone (and my gut says that’s likely a low number). If retrofitted into an existing aircraft I’d estimate it’d be around 3 day’s downtime to make the additions for removing panels and doing wire runs. This doesn’t include any testing or certification of said hardware. Those testing programs would likely cost around a million dollars to guarantee that the hardware doesn’t interfere with any other aircraft systems.
Each of the 3 largest airlines in America have around 1000 aircraft each. Taking my VERY rough, likely rounded down numbers means that hardware costs alone per airline would be around 40 million. Plus install costs, plus downtime for those aircraft would mean a retrofit program for each airline would be in the 100 million range.
So, it’s a lot of expense for the overall system, and for each aircraft that wouldn’t provide much more data than what was heard in the aircraft combined with control inputs as is already collected in the black boxes.
Latest Answers