1ºMike Merlo that's the thing. The only way to prevent and possibly save lives is a live uplink. With today's technology it could easily be monitored electronically for key words or phrases. Then start recording at that time and send an alarm to human to verify if there is indeed a current threat.
As the original post states the CVRs need to be updated or expanded. Honestly if there are no issues on the flight. The pilot or cockpit crew should be able to erase or record over with the next flight. The article is totally correct with today's technologies there is no need to limit to just 2hrs.
As someone else points out. No limit or cloud storage may help track maintenance issue history. If a plane crashes the whole history could be reviewed. Revealing the pilots noted this mechanical issue on previous flights..
2ºPatrick Daniel why are you approaching this from a "one-upping" someone perspective? Why don't you type something useful instead so I can get a factual response to my question of why transmission of such data is not a possibility? If planes are capable of transmission of internet (albeit they'll need to be transmitting more than receiving, granted) for hundreds of people onboard then why, with compression as good as it is these days (most of the telemetry being highly-compressible text along with audio that can be highly-compressed in realtime using lossy codecs), can they not transmit this data to be stored on land? Of course it would be storage of a lot of data but given that the current retention of black boxes is only 2 hours then even if data retention was set to 24 or 48 hours, there are storage platforms that deal with BigData far greater than this. Sure, it would take some international agreement as to where it would be stored, or is this therefore more a political problem rather than a technical one?
3ºJames Hanje Being recorded doesn't stop hijackings. That's my point. We have the recordings from UA 93 and the CVR worked as intended yet it still happened.
I have no problem with the recordings. It's a tool to help investigators when there is a crash. Current CVRs record for a period of time (I think about 2 hours) and then record over themselves. If the airplane crashes and power stops the recorder stops. The last 2 hours of what was happening is saved.
My opposition is to live uplinks of conversations that could be listened to by anyone with a password or that could get into the computer system. There are clear rules for pilots about "sterile cockpit" where we only talk about important/immediate flight related topics. When we are at a low risk time we are allowed to have personal conversations while still monitoring the aircraft.
4ºTodd Chapman it’s currently a satellite bandwidth issue. It’s on the horizon and in some limited instances we have streaming flight data, but we are a ways away from streaming the entire flight. Most folks don’t realize the massive amount of data a modern flight recorder captures. In binary form, we capture over a thousand parameters with most of them at 2-4Hz or two to four times per second. Some parameters are sampled at 8hz. A typical flight recorder download would export into 32,000 *seperate* Excel spreadsheets because of the one million row limit of Excel. It’s a massive amount of data. Streaming all of that just isn’t possible yet.
5ºGary Buchanan may not know what you want right now, but the best wishes for a beautiful evening will bring joy and happiness to your heart. Good evening friend! I have looked through your profile and you seems to have an impressive profile. Well I’d like to be your friend if you don’t mind since you know that friendships are built one day at a time they take time, patience, honesty, passion and ability to compromise and grow! So with true love, is by sharing bad and good times together. Your friend request is required if you don’t mind being friends with me. Have a good day today.
6ºJames Hanje Air Marshalls are occasionally on board, that's something that certainly could have helped. That goes back to the airplane being a bubble. If the Marshalls are on board they are the ones that can deal with the threat but they must be on the aircraft before the door closes and the airplane takes off.
If a hijacker wants to talk to someone the aircraft already has radios, satphones, ACARS (a text data link) etc. If they don't want to talk to anyone then they turn the radios off. It's really that simple. No live link would change that.
7ºMark Visser It's incredibly advanced. Two hours was the arbitrary number decided when the original 30 minute recorders were deemed too short. So we went to two hours. No one foresaw investigations on incidents that could have happened 14 hours ago and potentially overwritten. So now, 25 hours is the new arbitrary recommendation. One day someone will ask why we don't have 50 hour recorders. I can absolutely tell you it's an advanced recording system, I work with them daily. Video recorders are also a recommendation but come with an incredible amount of union pushback. Some operations have them, most don't. I'm a proponent of them, but we don't ask for "video". Video implies moving picture with audio. We already have audio (CVR). A single image grab every second would be a great start. Cockpit Image Recorder would be the appropriate term.
8ºAdam Foxton we don’t use excel for our primary means which is the premise of my statement. We decode the binary data into engineering units.  It would take far too long for me to explain how you arrive around 32,000 sheets but just trust me when I say it is accurate.  excel is not capable of accurately displaying frames, subframes, words, and bits and that’s in the order in which they are kept in the recorder. The way it would have to be exported into excel would necessitate my statement of how many separate spreadsheet are needed. A 25 hour recorder with a rate of 1,024 words per second would roughly fill 30,000+ excel files when the binary data is uncompressed and converted to engineering units.
9ºNeil Cassar Eldridge here you go. Here is your rationale: the average dataframe today records 128wps x 12 bits per word / 8 bits per byte. However the standard is now changing to 256 and 512wps. But for conservative estimate sake, lets stick with a 128wps recorder storing data at about 190bps. With an average of 15,000 flights in the air at any given time, that 3,000Kbps may not seem like much but on an Iridium satellite network that averages 3-4bps, you begin to see the problem. Now we havent even began to discuss streaming cockpit audio which is comprised of 4 audio channels. So now those 15,000 flights are transmitting the equivalent of 60,000 live phone calls every second, while also transmitting the aforementioned flight data. We have already started to stream SOME data, but the entire stream is just too taxing and costly at current satellite structures.