Originally Posted by
HazelNuts39
It would be easier to understand your discourse, if you could tell us what those areas are.
Sorry HN. Did not see this right off. I was not ignoring your question, which is reasonable.
First, let me say I have gone through the type course and had plenty of simulator time and more on the A320 series not that long ago. The AI FBW is an ingenious system and while I will admit not being a fan going in, I came to appreciate the immense amount of very clever programming that goes into making the FBW system a transport pilot's best friend 99% of the time. I tell you this so you know where I am coming from.
One of the things I discovered during my 'indoctrination' was there are times when you are not fully in the loop and have to trust the process as you monitor. This is not a knock on the AI FBW system. It is a symptom of the larger problem of automation in general and it cuts across aircraft brands and types. Pilot’s groups and unions have known about this for over two decades, but as an industry, we are just beginning to understand the need for a tighter coupling of the man machine interface and the importance of keeping current our manual flying skills - getting back to the basics. Some never left that place of surety, particularly the old timers amongst us, but many of us could easily be caught in the trap of not “using it” and “losing it.”
I believe the evidence produced so far clearly indicates the crew of AF447 were out of the loop from the start, and it appears they were caught in a trap of mostly their own doing, however I also strongly feel their aircraft or training , most likely both, let them and their passengers down in what turned out to be the worst way possible and it was the combination of these factors that led these otherwise competent pilots, down the garden path. I strongly feel the BEA should have explored this aspect of the accident further. The report should have included data and analysis of the several documented instances of unreliable airspeed and known pitot issues across the fleet to better understand how this critical sensor failure manifests itself in real life.
Three paragraphs on, and finally to your question HN. What more could the BEA have done in preparing the report, outside of the above sentiments I offer? The published DFDR traces are entirely too coarse at roughly 30 seconds per ½ to ¾ of an inch. By statute, the DFDR on this category aircraft is required to provide a sampling of the pitch, yaw, and roll inputs and their corresponding control surface positions at a minimum rate of 2 hertz, with 4hz being standard for these channels. That's 2 to 4 samples per second. As to the other available channels, the raw CG trim fuel data might be interesting to know. An additional set of engine parameters beyond N1 might be useful. The AOA and airspeed plots are a mess as are the control input traces, the former group mostly due to the erroneous data but all three groups further muddled by the coarse scale that was chosen. The EFIS state is marginally presented, but a bit of a mystery remains as to exactly what was seen in the cockpit that night and when.
I see a renewed interest in the CVR transcript. This is one of the most sensitive areas of accident data. As far as I know, nearly all ICAO member countries share similar legal requirements for the handling of the CVR audio and transcripts. Other than photographic or descriptive evidence of human remains, the CVR audio is the most protected data set of the investigation. The BEA seems to have done a reasonably fair job in sharing this most sensitive part of the accident investigation.