4 ms·

Slowly but surely the system is moving towards an additional layer of communication using text. The system is CPDLC. The major issue with using it exclusively

Slowly but surely the system is moving towards an additional layer of communication using text. The system is CPDLC. The major issue with using it exclusively is that anything outside of ordinary canned messages takes time to type and that’s often exactly the time you do not have time. Go back to the Miracle on the Hudson. Their radio transmissions are very brief and concise because they are extremely busy in the cockpit. They definitely didn’t have time to fiddle with a tiny keyboard to transmit all of that information. I think text is going to be a very important layer in the future but it will never completely replace voice transmissions.

tpmx 7y ago
So what's needed is a hybrid voice/text system. Preferably languge-agnostic on the text/data part. Text for the 99.9% normal things, voice for the 0.1%. After listening to ATC recordings between e.g. JFK and Air China pilots I have no idea how major accidents don't happen every week.
- __s 7y ago
  For reference: https://www.youtube.com/watch?v=1NDqZy4deDI https://www.youtube.com/watch?v=1NDqZy4deDI Transmitting text from ATC would also be able to have checksums which are then played by an automated voice. Pilots would become accustomed to the automated voice, as opposed to varying accents Regardless, if you think "I have no idea how major accidents don't happen every week" then it's likely because you don't recognize how much what you're referencing is an outlier, or you're disregarding what else mitigates issues
  - tpmx 7y ago
    Thanks for finding a link to that infamous recording. I think a hybrid scheme could work well. Transmit data in text, greetings via voice. Make sure to do the greetings every time, just to make sure people are used to using voice, for when the emergencies happen. > Regardless, if you think "I have no idea how major accidents don't happen every week" then it's likely because you don't recognize how much what you're referencing is an outlier, or you're disregarding what else mitigates issues You do realize you are commenting on a story that's about.. exactly that?
    - tpmx 7y ago
      I'm sure something nice could be built with frequency pairs: one for voice, one for data. It should be possible to build something backwards compatible. Use processes to fall back to voice for when data is lacking, etc.

tpmx 7y ago

So what's needed is a hybrid voice/text system. Preferably languge-agnostic on the text/data part. Text for the 99.9% normal things, voice for the 0.1%. After listening to ATC recordings between e.g. JFK and Air China pilots I have no idea how major accidents don't happen every week.

__s 7y ago
For reference: https://www.youtube.com/watch?v=1NDqZy4deDI https://www.youtube.com/watch?v=1NDqZy4deDI Transmitting text from ATC would also be able to have checksums which are then played by an automated voice. Pilots would become accustomed to the automated voice, as opposed to varying accents Regardless, if you think "I have no idea how major accidents don't happen every week" then it's likely because you don't recognize how much what you're referencing is an outlier, or you're disregarding what else mitigates issues
- tpmx 7y ago
  Thanks for finding a link to that infamous recording. I think a hybrid scheme could work well. Transmit data in text, greetings via voice. Make sure to do the greetings every time, just to make sure people are used to using voice, for when the emergencies happen. > Regardless, if you think "I have no idea how major accidents don't happen every week" then it's likely because you don't recognize how much what you're referencing is an outlier, or you're disregarding what else mitigates issues You do realize you are commenting on a story that's about.. exactly that?
  - tpmx 7y ago
    I'm sure something nice could be built with frequency pairs: one for voice, one for data. It should be possible to build something backwards compatible. Use processes to fall back to voice for when data is lacking, etc.