Failure analysis

The loss of Air France flight 447 is a horrible tragedy. I can't even begin to imagine the pain of those who lost loved ones on this flight.

From a failure analysis perspective, it's an interesting study in what happens when there is very little data available. There was no communication from the crew before the accident. It happened over open water in inclement weather, making access to the accident site difficult. Most of the wreckage has presumably sunk in water that may be as deep as 15,000 feet, making recovery at best difficult and most likely impossible. For accident investigators recovering the black boxes is critical; the chances of such a recovery are very slim.

80% of airplane accidents occur three minutes after takeoff or eight minutes before landing, so a mid-flight accident of this type is pretty rare.

With the small amount of evidence available, people have been making lots of very sketchy presumptions as to what happened. The plane was flying through a storm, but the area actually has a low incidence of lightning, far less than what you'd see over land in the continental US. Planes are actually struck by lightning approximately once every 1,000 flight-hours, but the last lightning-related aircraft accident in the US was in 1969. Planes are pretty robust when it comes to withstanding lightning hits.

Some have suggested it could've been a terrorist attack, but there's no direct evidence that this is what happened. There was a bomb threat several days before against an Air France flight from Buenos Aires to Paris, but there was no evidence the threat was real.

In most airline accidents, which usually take place over land, engineers have an enormous amount of debris and data to analyze. Forensic investigators have become very adept at piecing the evidence together to determine an initial cause. In this particular accident, the scarcity of evidence and the low probability of finding much more mean we may never have a good idea what happened in this accident. At this point, any guesses about the crash are very much conjecture - the fact of the matter is that very little is known about what occurred. There's not much to do but let experts analyze what little evidence there is and hope they can reach a decent conclusion, but I suspect that's not how this story will end.