Nothing of substance has been posted in the last 80 pages or so.

Type rating discussion leads to SFAR discussion then the MU2 comes up. It goes in circles but it's a valid topic.

Lots of better examples for type ratings like the Eclipse, Phenom 100 or Mustang.

Reading the FAA FSB Reports http://fsims.faa.gov/PICResults.aspx?mode=Publication&doctype=FSB%20Reports will give you the FAA thinking and training requirements behind any aircraft requiring a type rating. Before the SF50 is delivered there will be an FSB Report on it.

One thread over, a Beech 1900 thread is in the process of being diverted to a Mu2 thread. We have already gotten to the 'autopilots are for chumps' stage.

You are correct to some extent, especially WRT the level of automation (and associated redundancy) found in most GA airplanes.

That said, well designed automation is likely to result in fewer, not more accidents because significant automation failures are far less frequent than scenarios where pilots come to grief while flying without (or failing to use properly) useful automation. Just considering the rather unpleasant number of accidents picked apart in crash talk over the last couple years I don't recall a single instance where it appeared that an automation failure was a significant factor. IOW I strongly believe that if it were possible to follow two otherwise identical groups of pilots for several years where one relies heavily on automation and the other avoids it, the former group would have fewer serious accidents. Unfortunately such a study hasn't (and won't) happened so I can't offer any evidence to support my belief.

And just because a pilot chooses to employ automation most of the time doesn't necessarily mean they're unable to handle an automation failure. Any of us can have the best of both worlds by becoming comfortable operating the avionics in all appropriate phases of flight in addition to putting sufficient effort into recurrent training to maintain hand flying proficiency. Another mitigation strategy would be to avoid challenging situations where possible if the automation is on the fritz (e.g. if your autopilot goes TU on a flight that would terminate with an approach in low IMC, divert to somewhere with better weather).

You are correct to some extent, especially WRT the level of automation (and associated redundancy) found in most GA airplanes.

That said, well designed automation is likely to result in fewer, not more accidents because significant automation failures are far less frequent than scenarios where pilots come to grief while flying without (or failing to use properly) useful automation. Just considering the rather unpleasant number of accidents picked apart in crash talk over the last couple years I don't recall a single instance where it appeared that an automation failure was a significant factor. IOW I strongly believe that if it were possible to follow two otherwise identical groups of pilots for several years where one relies heavily on automation and the other avoids it, the former group would have fewer serious accidents. Unfortunately such a study hasn't (and won't) happened so I can't offer any evidence to support my belief.

And just because a pilot chooses to employ automation most of the time doesn't necessarily mean they're unable to handle an automation failure. Any of us can have the best of both worlds by becoming comfortable operating the avionics in all appropriate phases of flight in addition to putting sufficient effort into recurrent training to maintain hand flying proficiency. Another mitigation strategy would be to avoid challenging situations where possible if the automation is on the fritz (e.g. if your autopilot goes TU on a flight that would terminate with an approach in low IMC, divert to somewhere with better weather).

How was this an automation failure? It was an older PC12, not an NG so it wasn't packed with the latest and greatest. The pilot flew into a storm. This crash had nothing to do with automation.

How was this an automation failure? It was an older PC12, not an NG so it wasn't packed with the latest and greatest. The pilot flew into a storm. This crash had nothing to do with automation.

Pascals MU2 had a failure that required hand flying. Was that an "automation failure"?

This is a circular argument.

The more pilots crash manually flying, the more they should have used the automation. The more they use the automation, the lower their manual flying skills. The lower their manual skills, the more they crash flying manually.

It isn't about the automation failing (though there are plenty of examples of this), it is about the automation damaging the manual flying ability of the pilots. This damage is insidious, happens without notice, so that when your manual skills are required, they aren't up to the task.


If you can't or won't fly the plane to some airport manually but would let the autopilot do it, you are an autopilot dependent pilot and you should not take that flight even if you knew the autopilot would function for the entirety of it.

Autopilot failure should not be an emergency, IMO.

Mike C.

Pascals MU2 had a failure that required hand flying. Was that an "automation failure"?

Storm had little to do with it. Pilot could not hand fly when the automation disconnected.

http://ntsb.gov/_layouts/ntsb.aviation/ ... 4234&key=1

The failure of the pilot to maintain control of the airplane while climbing to cruise altitude in instrument meteorological conditions (IMC) following disconnect of the autopilot. The reason for the autopilot disconnect could not be determined during postaccident testing. Contributing to the accident was the pilot's lack of experience in high-performance, turbo-propeller airplanes and in IMC.

The pilot was using the autopilot as a skills crutch. Then it wasn't there. Now he and 5 others are dead.

Mike C.

http://ntsb.gov/_layouts/ntsb.aviation/ ... 1605&key=1

Following a second instruction to turn to heading 270 degrees, the pilot indicated he was, but was unable to disengage the autopilot.

2 dead.

http://ntsb.gov/_layouts/ntsb.aviation/ ... 3615&key=1

The pilot’s in-flight loss of airplane control due to spatial disorientation while operating in dark night instrument meteorological conditions, which resulted in the exceedance of the airplane’s design stress limitations and a subsequent in-flight breakup. Contributing to the accident was the pilot’s distraction by the reported malfunction of the autopilot system.

1 dead.

http://ntsb.gov/_layouts/ntsb.aviation/ ... 0220&key=1

After noticing the airplane turn about 180 degrees, the controller queried the pilot, who had not recognized the turn and stated that his autopilot had disconnected. The controller subsequently issued the pilot a heading back toward ASE, which the pilot accepted. However, the airplane continued to turn left and then began a rapid descent to impact.

1 dead.

http://ntsb.gov/_layouts/ntsb.aviation/ ... 3745&key=1

Contributing to the accident were (1) the complexities of the autothrottle and autopilot flight director systems that were inadequately described in Boeing's documentation and Asiana's pilot training, which increased the likelihood of mode error

3 dead, numerous injuries.

http://ntsb.gov/_layouts/ntsb.aviation/ ... 4708&key=1

possibly his sudden need to fly the airplane without the aid of the autopilot would have been conducive to the development of spatial disorientation.

3 dead.

http://ntsb.gov/_layouts/ntsb.aviation/ ... 3524&key=1

If the resistor was in an open condition at the time of autopilot engagement, the autopilot would appear to engage with a mode annunciation indicating engagement, but the pitch and roll servos would not engage.

1 dead.

http://ntsb.gov/_layouts/ntsb.aviation/ ... 1302&key=1

It is likely that the reason the pilot requested to “leave the frequency” was to leave his seat and attend to something in the airplane. While leaving his seat, it is plausible he inadvertently disconnected the autopilot and was unable to recover by the time he realized the deviation had occurred.

2 dead.

NTSB search of fatal accidents with the word "autopilot" produced 492 entries. The above are just snippets from the first few where the autopilot was mentioned as possibly playing a role.

The main difficulty in detecting autopilot related crashes is that there is little to no evidence after the fact to say the autopilot was being used, engage, or failed. In some cases, a survivor can provide evidence, but in most cases, you never know who or what was flying. We just don't have good evidence in GA.

Mike C.

It did? How do you know that?

Mike C.

Storm had little to do with it. Pilot could not hand fly when the automation disconnected.

http://ntsb.gov/_layouts/ntsb.aviation/ ... 4234&key=1

The failure of the pilot to maintain control of the airplane while climbing to cruise altitude in instrument meteorological conditions (IMC) following disconnect of the autopilot. The reason for the autopilot disconnect could not be determined during postaccident testing. Contributing to the accident was the pilot's lack of experience in high-performance, turbo-propeller airplanes and in IMC.

The pilot was using the autopilot as a skills crutch. Then it wasn't there. Now he and 5 others are dead.

Mike C.