Total Autopilot Failure
 

Hi all,

There are two interesting threads running on the use of automatics; Auto flight vs hand-flying & I’m starting to think…

Question: has there ever been an example of total autopilot failure on modern airliners? I’m talking A320/B737 (lol modern)/A330/B757 Etc to present day.

The 320 has two independent autopilots, the 737 is the same. I think the 757 has three?

I have flown with the autothrust INOP on the 320 for a day.

But has there ever been a case of BOTH APs failing during normal operations? Or dispatching with both APs INOP?

And to be clear; I’m asking about only complete AP failure - NOT failures that lead to loss of AP - for example dual hydraulic loss on the 320.

Regards,

BD

so you thought you'd start a third?

Never dispatched without autopilot (pointless as we wouldn’t be RVSM). You could search for the report of the A320 at EGPH which lost all the automatics thanks to their runway being too rough at take-off (and the bouncy runway has nothing to do with the Airbus-shaped dent I left in the touchdown zone a few years ago).

I once dispatched with both A/P's inop. Gatwick to Athens and back on a 737 classic. Quite difficult operationally - both the F/O and I shared the flying doing about 15 or 20 minutes at a time.


Kind regards
​​​​​​​Exeng

I’ve been dispatched without an AP and without any FMC. My friend flew an A320 without AT the other day. None of the events made the news because nothing happened.

A couple of years ago my mob flew a 767 across the Atlantic without any A/P. Probably below FL280. Needless to say the pilots were reportedly quite tired :)

Of course. I’ve seen a dual FMGC fault during an ILS approach on a 320, which led to no AP and no AT. I’m sure I’m not the only one. The FO was flying by the way, he didn’t bat an eyelid and completed the approach manually, in full IMC!

Yes, there are some very straightforward single failures on the 737 that prevent either autopilot from working, losing a single IRS would be such. It could also be a simple electric or a mechanical failure, such as a sticky AP disconnect switch or a closed circuit for that switch - can't engage AP if one the disconnect switches is pressed (or sensed as pressed).

737 up to including NG can be dispatched with both AP inop (outside of RVSM etc.), but on the MAX, at least one must be operative to provide MCAS functionality.

C141(L300) PANC-RJTY, +/- 1976, A/P would not engage on the climb out. Took turns hand flying for 8 or 9 hours at FL350. Not fun.

Gotta say, the fact that you need to commend that fact as remarkable clearly shows the sorry state of things in aviation at the moment...

I hear what you say, but I wouldn't really call it a sorry state. To me it's a glass half full/half empty situation: we can all fly and land the aircraft without the automatics to a safe outcome, it just not a routine operation for good reason. So I commend those who keep up the skill to a high degree like the FO on the day our autopilot failed, it gave me more options to manage the situation efficiently.

Most definitely. I’ve had it in the cruise where both failed never to return with no other indications of a problem. ( 737 NG)

I have lost the whole MCP (so no AP, FDs or autothrottle) on a 737 several times, and would have to refer to my logbook as to when and where. I have also dispatched without one, although I can’t remember why!

On the ‘bus or other Boeings, not yet, although I have had an FCU channel fault a few times which is only one failure away from being like a day in the sim.

I hand fly raw data or visually a couple of times a week in conditions (and colleague!) allow; a loss of an AP gives you an aeroplane that is in trim and shouldn’t be any issue at all beyond RVSM and low vis/RNP considerations.

Something like unreliable airspeed or or an uncontained fire cause me far more concern than losing the AP, which is a workload management tool at the end of the day.

I recently had a colleague tell me of two separate incidents where the 777 AP failed. It happens.

Been there in C-5 on the NATS, once before RVSM and once afterwards. F280 is a lot easier than even F330. Also done four crossings in the A-10, no AP installed. Kids these days.

lol that’s how I feel sometimes.
I remind myself that the previous generation said the same thing about me. I recently read an account of Guy Gibson trying to light a cigarette and maintain 100ft AMSL at night over the ocean while leading a formation flat out. In the end he got his second pilot to light it as the nose kept on dipping each time he tried.
I made the mistake of staying up at FL370 for over an hour when fuel wasn’t an issue. I’d definitely drop down a bit if it happened again just to reduce workload. It wasn’t a problem as such but neither of us could trim it to a nice stable hands off state.

I’ve experienced this in the sim once or twice 🤣

In the days before RVSM:-

Flew an A320 manually at FL390 for several hours. Not an autopilot issue but vibration in the rudder when it moved out of the neutral position.

Had both autopilots failed on a 737-300. We flew from NW Europe to the Canaries and back.

I started my career flying aircraft that had no autopilot installed. We flew all weather IFR.

Humans expected to 'fail'; automation not
 

Flight guidance system failures are rare occurrences, particularly in those aircraft designed to be operated with advanced technologies. FGS;- FMC, navigation, FD, autopilot, thrust management.

If operations without tech assistance are as described in the preceding 'heroic' posts, as onerous high-workload, unexpected situations, then the circumstances and outcome should be formally investigated.

In the span of a career, advanced technologies have improved safety, but with this improvement a significant change in the operational perception of these systems.
e.g. reacting to a GPWS alert was a safety success - appropriate human action; now-days an alert requires formal reporting and investigation, with the connotation of failure by someone / something.
The objective is to learn, but what; that complex systems have unexpected, successful outcomes, more often defying explanation of why people reacted as they did and saved the situation - as expected.

In an increasingly safe industry the implication is that (explainable) tech failures in technology-depended aircraft should be investigated with similar vigour as inappropriate human behaviour - judged after the fact.
The objective would be to identify and avoid safety issues where changes in standards of human performance, education, training, experience, etc, are unable to provide an equivalent level of safety in modern operations.
mm

​​​​​​​You usually choose your words better than that Safetypee