In the 6000 + fleet of Cirrus SR2xs this has happened only a couple of times
I bet it is much more than 2×. I have probably heard of 2 directly face to face from pilots.
There are loads of crash stories from pilots I have met, from static when flying in IMC, in “plastic” planes. The DA40 had this a lot.
The magazines will never print any such story because Garmin owns much of the avionics advertising revenue.
Overall I don’t think glass avionics are significantly (I mean say 10x) more reliable than the old stuff (in simple terms of failures per airborne hour) but there are many caveats e.g.
which can mean that individual experiences will vary a lot.
Not about the PFD but the MFD, but I recently had an interesting experience in the SR22. Just as I was flying towards the Golden Gate Bridge and starting my turn back due to low clouds, the TAWS in this (Avidyne) Cirrus started saying “obstacle”. Next time I looked, the MFD wouldn’t show any useful information; most pages were greyed out, I think only two were selectable but didn’t contain any information. One of them was the TAWS page which just said “TAWS failure”. No more moving map or engine data or trip information was available. Pulling the MFD circuit breaker restored everything again. But that was an interesting example of a software system that seems to work flawlessly for hours, and right when you might need it it fails.
The magazines will never print any such story because Garmin owns much of the avionics advertising revenue.
That’s not true, Peter. Magazines will always write about such stuff, and I did it for 20 years. What’s true: Sometimes somebody from the advertising department would come to the editorial office and ask me " do you really have to write that". And as and editor I answered: yes.
This happened to me when I reported about the lousy climb rate of the first Thielert Cessna 172. The Thielert company knew that I had not been so impressed with the test flight and called my boss and said “that they would stop all ads” in our magazine. The article was published as I wrote it.
The trick is: Once your magazine is your market leader they will think twice about only their competition having ads in your magazine. That’s how germn SPIEGEL news magazine can have a big artikcle about corruption at some big company – AND their ad :-)
I would say that with 100,000 aircraft-years or more of operation in G1000 systems, if these glitches were a serious security hazard, it would be clear by now.
If you look at airliners with computers, knowing which CBs to pull if it “does that thing again” is part of the routine, but overall the safety improvement from generation to generation (classic airliner —> FMS, TCAS, GPWS —> EGPWS + envelope protection) was noticable.
C.
Peter wrote:
Actually I am absolutely staggered that avionics don’t seem to contain a watchdog.
What makes you think that? To me this would be so self-evident that I’d like hard evidence to the contrary.
I have some schematics 
Not that I have any experience (I did touch a G1000 once…) but in a glass cockpit a watchdog wouldn’t necessairly have to be on a schematic, could be software based. And lots if the potential failure modes are in software, the hardware ones will either be obvious (black screen) or picked up by the software (lack of GPS signal).
Peter wrote:
I have some schematics
You win. :-)
There could be an on-chip watchdog of course. Or the watchdog could be in the OS. The chance of the basic OS scheduler hanging is really very, very, slight.
