Where the 117 rules are obviously a complete and utter failure are circadian rhythm swaps, as has been mentioned above.
The phrase 'science supports the new rules' works for most of them but not this; anyone who has read a fair amount about the whole concept of the circadian rhythm and what is taking places in different phases of sleep can clearly see that the rules do absolutely zero to prevent issues in this area other than support a pilot to call in fatigued if they are tired (that part is good, of course). The science explanation is so ironic in this regard as the science could not be further from agreeing that the schedule below would be conducive to a rested, alert pilot.
Legal workday sequence:
Sleep from 23:00 to 07:00 at home.
Show up from days off for...
On duty 09:00 to 21:00.
Layover. Sleep from 23:00 to 07:00?
Awake from 07:00 to . . .
On duty 21:00 to 05:00. [note at this point the pilot would have gone on duty after being awake for 14 hours, two hours before than their usual bedtime, so they would not feel tired, but by the end of the duty period, they would have been awake for 23 hours]
Layover. Sleep from 06:00 to . . . 11:00ish if able? Not likely as their normal waking time was just 60 minutes after they laid down, and that's assuming they were able to fall asleep essentially the moment they walked into the hotel room (or possibly worse, home).
On duty 20:00 to 04:00.
Layover. Sleep from 05:00 to . . . 11:00ish again if able? Maybe more likely now as they are adjusting to a night owl schedule, albeit by missing two consecutive periods of deep sleep along the way, which is very problematic and science supports this notion.
On duty 03:00 to 11:00. Having awoken at 11:00, I think we could agree they'd likely be awake until at least 21:00 and more than likely 22:00-23:00 having done night work the previous two duty periods. Now they'd be setting an alarm clock for 01:30! That's 2-4 hours of useful rest.
I'll stop here after the fourth duty period as obviously this absurd sequence should not be allowed under 117, and even two more duty periods could be scheduled after these.
I am on days off typing from 8400 miles from home on a trip of nearly three weeks time off that I was able to create by cramming all of my flying into the other sides of the adjoining months. We've all done this type of thing. This is easier to do with 117 in my own experience. And I felt I bid responsibly and did not create anything that would needlessly tire me out. So, I like parts of these new rules for their 'side effects.'
But what's a pilot to do who's bidding 80% in base and gets stuck with whatever the system gives them?
Call in fatigued ten times per year?
Is that really a successful set of rest rules? Where someone could really, truly be in a position to have to do that?
I think the new 117 rules do a great job of looking at an individual workday as it's own entity and preventing a pilot from being on duty too long. The 16 hour workday is gone forever and that is great, and the 8.0 hour layover including transit is gone forever, and that is also great.
But as long as that individual work period is not fully tied into the days around it, then the success of the new stuff is only partial.
And partial success is a failure in this arena just as a baseball team who is in the lead for seven out of nine innings still loses the game.
It's the least common denominator that counts, and we've got crews doing sequences similar to the one I concocted above. The fatigue induced accident isn't going to happen to the crew that's doing some 9 to 5 daytrip. It only takes one crew out there doing something similar to above to have a problem and then we're back to square one. Unless it's a cargo crew, of course -- because they'd never mistakenly cross a runway in the dark, in fog, with rain covered cockpit windows, on their 15th hour on duty, and if they did, there's no way a passenger airplane loaded with insurance policies could be coming at them.
