Air France 447: Eight Years Later

Eight years ago on the night of May 31^st, 2009, Air France flight 447 from Rio de Janeiro, Brazil was lost in the Atlantic Ocean north of Brazil.

Virtually every aviator today knows the number 4-4-7 and the unnecessary tragedy it represents. The initial reaction of many was “how could this happen?” How could three airline pilots possibly lose control of their aircraft in under a minute and kill everyone on board within 4 ½ minutes of the autopilot disconnecting—essentially because the autopilot disconnected?

But the world of aviation has not sat idly by in its aftermath. The NTSB has a saying: “From tragedy we draw knowledge to improve the safety of us all.”

Much like the June 1975 crash of Eastern flight 66 at JFK due to a microburst, and the follow-on windshear crashes of Pan American 759 in New Orleans (1982) and Delta 191 in Dallas-Fort Worth (1985), these accidents lead to increased research and education on the weather phenomena and the development of windshear warning systems, both on the ground and as airborne equipment.

Also like the December 1974 crash of TWA 514 into the mountains outside of Washington, DC, which lead to the increased awareness of minimum safe altitudes, the installation of the basic Ground Proximity Warning Systems (GPWS) on all transport aircraft and the subsequent development of the modern look-ahead GPWS systems.

In each of these cases, the development that resulted in the aftermath of these accidents significantly reduced or virtually eliminated those causal factors in the years since.

Loss of control in flight (LOC-I) now remains as the leading accident factor in the few accidents and incidents we see in recent times.The loss of Air France 447 and another LOC-Iaccident earlier that same year, Colgan 3407 on February 12, 2009, spurned:

• Increased awareness of the need for manual flying practice and more concentration on hand flying skills.
• The need to be aware of normal pitch attitudes and power settings for all flight regimes.
• The roll of the startle factor in pilots being able to cope with unexpected unusual conditions.
• The roll of automation dependency in this increasingly automatic industry.
• Increased training, now mandated, in upset recovery, high altitude stalls, with a revised emphasis on correction of a stall by reducing angle of attack, not just power application.
• Increased minimum flight times for new airline pilots.

And while most of the emphasis is focused on human flying skill (and rightly so), technological advancements have also been taken, almost surely with this accident in mind. The new Airbus A350’s autopilot now can remain engaged in all but the most extreme situations, thus far reducing the instances where pilots are unexpectedly handling a causal factor (e.g., weather), system malfunction, and autopilot disconnection all at the same time. The system allows the airplane to recover back to within the normal flight envelope from extremes of speed, bank, and pitch and reduces the other recovery maneuvers that must be hand flown by providing for the capability of the autopilot to fly windshear and TCAS traffic resolution advisory maneuvers automatically, and allow the pilots to take over when they are ready. Enhancements to the weather radar allow for better analysis of a storm and provide an aural warning should one several minutes ahead go unnoticed.

Yet, still, the problem of LOC-I has not been eliminated, even though the lessons of AF447 seem loud and clear. Air Asia 8501 in December 2014, stands as a stark reminder that continued diligence is required.

Commercial aviation is very safe, it is in fact amazingly safe, but it is not inherently so. It is safe because a lot of people work hard every day to make it that way, and to keep it that way.