Layers of safety in aviation

Last edited Fri Jan 25, 2019, 01:55 PM - Edit history (1)

I have worked as an engineer in the aviation industry. I was taught early on that it is very rare for air disasters to come down to one single point of failure. Margins of safety, redundancies and backup systems, human factors considerations, crew resource management, crew rest rules, maintenance program design and oversight, recurrent training, etc.

Our industry has good reason to be proud of its safety record. Over the decades, it has learned from its mistakes. Accidents are studied to determine root causes, and recommendations have led to real changes in equipment, training, crew interactions, and so forth. Vast swaths of people in the aviation community are absolutely dedicated to safety as their first priority.

These systems and rules are VERY detailed. For example, there are tiny cracks in virtually all metal parts, and ultrasonic inspections to detect cracks or defects in critical areas are performed at predetermined intervals. The engineers will calculate what size of a defect is detectable, and how fast it will grow to a size that causes the part to fail. Then, maintenance programs set an “inspection interval” that is short enough that at least THREE inspections will be performed after the point that the defect is detectable (and rejectable) but before the point it becomes a failed part. You don’t allow ONE improper accomplishment of an inspection to lead to a critical part failure; you design a program tight enough that THREE inspectors would have to do the inspection incorrectly before the critical part fails.

Another example. Special rules for “ETOPS” flight — if you have to do certain maintenance on both engines, you never do the maintenance on both sides on the same night with the same personnel, because if they make the same mistake twice in a row, the risk escalates.

Another example: crew rest rules. Five years ago, revised crew rest rules were implemented, driving a need for a major sudden increase in pilots, and greatly complicating operations. It was described to me that instead of two clocks ticking on each pilot (rest time / flight time), there were now “six clocks ticking” that have to be tracked. Not only do pilots have to get 10 hours off to rest, but the time to travel from airport to hotel and back have to be considered. Within a 7 day period, they must always have 30 hours off uninterrupted, and if they have had time zones changed, that must also be considered to account for internal “clock adjustment”. Night time flying has further restrictions. Keep in mind crews are scheduled a certain way, but weather or other problems cause changes to the schedules, so this is a very dynamic set of circumstances.

The crew rest rules like everything else is based on science, along with cost-benefit calculations and input from all relevant parties. An excerpt from the document:

It is infuriating that this refined system of safety responsible for safe passage of millions of passengers is at risk simply because a short fingered vulgarian in the White House and a racist greedy traitor in the Senate have conspired to shut the damn government down.

My impression is that Air Traffic Controllers have the hardest, most critical jobs in aviation. They could be considered an extension of the pilot. And I don’t think there’s a lot of redundancy in what they do. I just heard on TV that ATC controllers work 10-hour days, 6-day weeks, BEFORE the shutdown. I am shocked. But now, in addition to stressful 60-hour a week jobs, they are worried about paying bills, and some are trying to get second jobs to pay the bills. Not to mention the added annoyance of having some of the support staff in the facility on furlough.

If you watch Air Disasters shows, you may have seen tragic accidents wherein communication between ATC and the pilots contributed to a crash. There are 30 listed on this Wikipedia page — mostly, not in the U.S. Avianca, in Cove Neck New York, put in a holding pattern for too long, failed to use word “emergency”, until they ran out of fuel. Combinations of problems add up — weather plus pilot mistakes, or extra time spent on the runway plus ineffective de-icing procedures, or air traffic control failing to warn pilots about bad weather. Aviation is the “biggest team sport”, and every part of this system has to be working properly so that operations are orderly, predictable, and routine. A disaster might be directly caused by an ATC person who is over-tired or stressed out, or it might be caused “by weather” or “pilot error”, but in the latter case, an alert ATC guy might have been able to prevent it.

The GOP simply has no clue about the value of work or federal workers.