Originally Posted by
Bergerie1
Can somebody check my maths (never one of my strong points), but looking at the graph on page 14 in the link, in 1961 there were 9 hull losses per year and around 1 million flights. In 2017, there were 35.8 million flights per year. If the same accident rate had persisted, there would have been 346.5 hull losses in 2017
https://accidentstats.airbus.com/sit...-1958-2021.pdf
A stunning improvement by any measure even though there remains much to be done to improve crew training. Any thoughts or corrections?
The 60's and 70's were brutal for plane losses. The temporal change and basic nav and displays were not conducive to safety. Early jets and fanjets were quickly more reliable than big pistons and turboprops, but the performance changes challenged the training and alerting systems. As alerting and SA systems improved, the crew coordination and decision-making became evident as problems, and post Portland and the Everglades, coordination seemed to get a boost in emphasis.
The concern in the early 70's was that the rates of losses were so high that the expected expansion of global aviation would result in unacceptable headlines every day, as you indicate. EGPWS/TAWS/TCAS/EFIS/ECAM-EICAS/GNSS etc helped, and additional protective warning modes were added to warning systems, which also became conditional to assist the crew in making appropriate decisions.
The crew have been losing control of aircraft since the start of aviation, the concern remains that loss of control generally results from a failure of SA and we have much more advanced tools for the crews to mitigate that yet it continues. The losses in most cases make depressing reading with very few novel manners of losing the plot being disclosed. The HITL is still the current most effective way of mitigating the gap between planned design behavior and system reliability. Automation inherently places the human into a process that they are less effective at over time, that of monitoring, and frequently that is a primary factor in the divergence from the desired state to the actual state. The recovery from an anomalous state often is at the end of the spectrum of responses from nothing or inadequate through to excessively aggressive with possible adverse consequences, e.g., Lauda's TR event was time-critical, AA587 needed a gentle response, both events had critical information that was not self-evident before the event which in hindsight resulted in changes to procedures. SR111 is another case where there was a mismatch of procedures and needs, which generated some changes post-event.
Competency in dealing with recognition of time-sensitive anomalies is a challenge but offers potential relief from such things as 260 kt low flying on takeoff, or wild rides in recovery from upsets.
Much like the '70s. management needs to look to new initiatives to reduce the residual events severities that exist now with current aircraft operations. Some stuff will remain depressing, like the turning off of all IRS's in IMC conditions... hard to mitigate unfathomable actions. Deliberate unlawful actions by the crew need more response than has occurred to date, and remain a blight on the industry. Post 9/11, the compliance with security protocols when audited varied from excellent to patchy, so protection from crew-caused illegal actions could also be variable depending on how seriously mitigation protocols (if any) are taken.