Autonomy in aerospace

[AdamSmith]

Interesting look at status and future of autonomous flight.

http://m.aviationweek.com/commercial-aviation/certifiable-trust-required-take-autonomous-systems-past-unmanned

[RA1]

I don't have time at the moment to fully respond to this article. There is just too much emotion, rumor and opinion as well as "facts" involved. Therefore, just a couple of points. In some ways we have already gone too far in automated flying. Some of the results include the 777 crash at SFO and others that indicate a lack of "stick and rudder" skills by current airline pilots (and others). After a hue and cry, there is a band aid approach. In other words, more emphasis on one particular aspect of flying. That is not a cure. It is simply emphasizing one thing this "month" or training cycle.

I question the statistic quoted about SP in Citations. This has been ongoing from the moment the FAA approved single pilot operations of jets. At first the insurance companies simply had a rate for SP operation that basically was the same dollars as hiring a second pilot. That was explained by lack of data. As time passed, the insurance companies further defined their "beyond what the FAA requires" qualifications for a reasonable rate. Today, not only does the aircraft have to be approved, but the pilot must undergo a defined re-current training course annually. One that is stricter than recurrent training for a two person crew. Then, that pilot/airplane will get an excellent rate which likely will be lower than the two person crew. Does this sound like I am defending a single pilot operation? Yes and no. As mentioned previously, today's training emphasizes one or two pilots monitoring the autopilot and other equipment, only inputting as required. Obviously this has good and bad points. Good that we have a lot of reliable electronic help. Bad that we do not necessarily have the skills from days before.

Answer: Hire a good old pilot and teach him new tricks.

Best regards,

RA1