This was NBAA week in Orlando where some of the most exotic aircraft on the planet are displayed along with all of the hardware and software that make them nearly goof proof. Nearly. Chris Hart, acting NTSB Board Chair, noted at the opening session his concern about some air carrier accidents that we’ve all read too much about:
- Asiana, in San Francisco where two highly experienced captains were unable to make a visual approach on a perfect day.
- Air France 447, which involved the mere loss of air speed due to icing (which shouldn’t have happened) that confused the flight data computer so auto-throttles and autopilot disconnected. The crew, bewildered by myriad warnings of symptoms, not cause, pulled the Airbus 330 into a deep stall and settled into the ocean from well over 30,000 feet.
Chris then went on to note that most subway trains these days are so automated that, under normal operation, the only thing the operator is allowed to do besides monitor is to close the doors—the train does everything else. It’s a thankless job because what little skill is involved in dealing with nearly non-existent emergencies is rapidly eroded by mindless boredom. Every so often there is an accident that requires the driver to get involved, and sometimes they don’t do very well. So it is becoming with many cross-country aircraft and some trainers.
I thought about what avionics might be upgraded in my own 27 year-old machine and what I could afford. It involved a long look back. Transitioning from the most basic instruments in 1970’s era trainers to the state-of-the-art instrumentation at the time was like magic: Two axis autopilots, a horizontal situation indicator, DME, and a flight director revealed that the high and mighty of the corporate world needed far less ability than the junior CFI and his students, who were manually doing everything and having to look about four places at once. The corporate world just went a lot faster and higher with a potty on board!
With the exception of potties, we’ve progressed far beyond that. Even in the basic training machines, if desired, you can program both horizontal and vertical guidance so that shortly after takeoff the magic is engaged, only to be shut down on short final. In research, aircraft have already demonstrated autoland capability and a new term, OPV (optionally piloted vehicles), doesn’t just apply to the local airport bozo.
Now, before getting too disdainful of all this, it must be acknowledged that the corporate safety record is second to none—including the airlines—and in some cases it is better. The automation doesn’t fail often, synthetic vision certainly helps to reduce CFIT, the fishfinders (aka traffic awareness systems) spot way more traffic than we ever will, and moving maps/GPS make finding airports a non-issue. On-board weather helps us decide—somewhat—how far we’d like to poke into phenomena that before we’d never have even approached.
But, like the subway train operator, is something being lost? What are the trade-offs and where is the optimal blend of skill versus automation? The light GA safety record is very gradually improving, but in way too many cases we’re still cracking up aircraft for all the old reasons.
It’s your turn to weigh in. How do you fly? How should we maintain skills and satisfaction while, in most cases, the automation will fly more precisely than us? Is there an optimal level of engagement for pilots that is below what the microprocessors can do? How should pilots stay mentally engaged and ready to take command rather than being a bewildered system monitor all the way to impact? As for “Human Factors,” any ideas? Kindly do NOT get too graphic!
In addition to the poll, I’d really like to hear your thoughts on this, so please submit your comments.