NW_Pilot's Trans-Atlantic Flight -- All the scary details...

Stefan writes:

Ok, so you want the FAA jump in and require full redundancy on all
instruments for all privately operated light singles to be considered
airworthy? I'm not sure you really want this.

It wouldn't be a bad idea for computer equipment, but it's not enough.
It would be necessary to require not only redundant systems, but
systems with independently-developed software, so that failure modes
of one system don't match failure modes of the back-ups.

--
Transpose mxsmanic and gmail to reach me by e-mail.

Neil Gould writes:

I think its reasonable to think that some of the G1000's wiring or the
unit itself was damaged during this hack.

That would not generally cause software reboots.

It's much more realistic to think that there are fundamental defects
in the G1000.

Even attaching the power to the
ADF or entertainment system in a way that caused the power to the G1000 to
be flaky or intermittent could account for the drastic failure modes he
experienced.

There should be no drastic failure modes at all, under any conditions.

--
Transpose mxsmanic and gmail to reach me by e-mail.

John Theune writes:

It would appear that applies to all aviation for you, it's a mystery,

But computers are not. And with what I know about them, I would not
entrust my life to a completely computerized cockpit.

--
Transpose mxsmanic and gmail to reach me by e-mail.

"g1000_eng" wrote in message
ups.com...
1) All avionics software implements internal isolation to prevent one
part of the system from taking down another part.
2) A faulty fuel reading cannot cause the system to reboot. In addition
to testing every possible faulty fuel value, I've tested every
combination of faulty sensor readings related to this thread and am
unable to get anything out of the ordinary to happen. The picture of
the fuel sensor with the red X is correct behavior when a gauge is
fauly or giving erroneous data.
3) When the system reboots due to a software error, a very obvious
message with a very obvious color is displayed on the screen prior to
the reboot. Was this seen? I have seen no mention of it.
4) FYI to a few: the CO message is indicating an error in the detector,
not CO in the cabin.

What was going on with the second display?
Was the "Initializing System" message being displayed each time it
'rebooted'?
During the 15minute intervals between reboots, how operational was the
system?

I won't delve into the actual debate issues of whether to go glass,
realtime reliability vs. features demanded, benefits vs. risk of
various situational awareness methods, or anything like that. I'm just
trying to get the facts straight. No software engineer would claim a
flawless system, but the facts so far do not allow for a simple answer
such as the fuel gauge or airspeed indication being the only cause.
Something very strange had to be going with where that escaping fuel
was going. If it was affecting three gauges (airspeed, co detector,
fuel) in a measurable way, who knows what it could have been doing to
less obvious internal wirings of the aircraft. I've never heard of a
report of a continuously rebooting system, and there are a lot out
there. The somewhat drastic customizations and the newness of the
aircraft add to suspicion. That said, there's no excuse for a failure,
wherever in the aircraft that failure is determined to be.

PS: I appreciate the balanced feedback and analysis of most of this
group. Don't feed the 20% trolls.


Do you expect me to believe you are a Garmin Engineer using a hotmail e-mail
addy and a 1 time poster under this username :-)

It's kind of funny once the aircraft was on the ground for 24 hours and
there was no further fuel venting and every thing dried up the G1000 worked
flawlessly until the tach failed some hours later during the flight because
of a Bad Sensor "not confirmed yet just an educated guess" There was no
alert about the reboot it was like some one just cut the power to the
displays then restored power the system did a standard reboot things started
to fail then it would reboot again. I will be talking with the customer
again upon delivery of future aircraft and will report the avionics tech's
findings on that airplane that is if they have anyone that is certified and
has the knowledge to work on and diagnose the G1000 over there!

"NW_Pilot" wrote:
"g1000_eng" wrote in message
ups.com...
1) All avionics software implements internal isolation to prevent one
part of the system from taking down another part.
2) A faulty fuel reading cannot cause the system to reboot. In
addition to testing every possible faulty fuel value, I've tested
every combination of faulty sensor readings related to this thread
and am unable to get anything out of the ordinary to happen. The
picture of the fuel sensor with the red X is correct behavior when a
gauge is fauly or giving erroneous data.
[ ... ]

Do you expect me to believe you are a Garmin Engineer using a hotmail
e-mail addy and a 1 time poster under this username :-)

Just FYI, the poster used Google Groups, which tacks on a header indicating
the IP address of the poster; in this case that header is:

Injection-Info: e3g2000cwe.googlegroups.com; posting-host=69.247.168.62;
posting-account=YoF8rQ0AAABFBsk62Tpp2wBJ_FD_CVG_

I used the "dig" utility on one of my Linux machines to see what host name(s)
might be associated with that 69.247.168.62 address. It came up with this:

c-69-247-168-62.hsd1.ks.comcast.net.

Looks like a Comcast account in Kansas, based on my understanding of how ISPs
like Comcast assign host names to IP addresses. Garmin is headquartered in
Kansas. So the possibility can't be immediately ruled out based on where the
poster is posting from.

Jim Logajan wrote:
"NW_Pilot" wrote:
"g1000_eng" wrote in message
ups.com...
1) All avionics software implements internal isolation to prevent one
part of the system from taking down another part.
2) A faulty fuel reading cannot cause the system to reboot. In
addition to testing every possible faulty fuel value, I've tested
every combination of faulty sensor readings related to this thread
and am unable to get anything out of the ordinary to happen. The
picture of the fuel sensor with the red X is correct behavior when a
gauge is fauly or giving erroneous data.
[ ... ]
Do you expect me to believe you are a Garmin Engineer using a hotmail
e-mail addy and a 1 time poster under this username :-)

Just FYI, the poster used Google Groups, which tacks on a header indicating
the IP address of the poster; in this case that header is:

Injection-Info: e3g2000cwe.googlegroups.com; posting-host=69.247.168.62;
posting-account=YoF8rQ0AAABFBsk62Tpp2wBJ_FD_CVG_

I used the "dig" utility on one of my Linux machines to see what host name(s)
might be associated with that 69.247.168.62 address. It came up with this:

c-69-247-168-62.hsd1.ks.comcast.net.

Looks like a Comcast account in Kansas, based on my understanding of how ISPs
like Comcast assign host names to IP addresses. Garmin is headquartered in
Kansas. So the possibility can't be immediately ruled out based on where the
poster is posting from.
And I would not expect a engineer from Garmin to use his real name and
email address because as a engineer I doubt he is authorized to speak
for the company. That does not stop him from commenting in a much more
knowledgeable way then a lot of the posts about this issue.

"Jim Logajan" wrote in message
.. .
"NW_Pilot" wrote:
"g1000_eng" wrote in message
ups.com...
1) All avionics software implements internal isolation to prevent one
part of the system from taking down another part.
2) A faulty fuel reading cannot cause the system to reboot. In
addition to testing every possible faulty fuel value, I've tested
every combination of faulty sensor readings related to this thread
and am unable to get anything out of the ordinary to happen. The
picture of the fuel sensor with the red X is correct behavior when a
gauge is fauly or giving erroneous data.
[ ... ]

Do you expect me to believe you are a Garmin Engineer using a hotmail
e-mail addy and a 1 time poster under this username :-)

Just FYI, the poster used Google Groups, which tacks on a header
indicating
the IP address of the poster; in this case that header is:

Injection-Info: e3g2000cwe.googlegroups.com; posting-host=69.247.168.62;
posting-account=YoF8rQ0AAABFBsk62Tpp2wBJ_FD_CVG_

I used the "dig" utility on one of my Linux machines to see what host
name(s)
might be associated with that 69.247.168.62 address. It came up with this:

c-69-247-168-62.hsd1.ks.comcast.net.

Looks like a Comcast account in Kansas, based on my understanding of how
ISPs
like Comcast assign host names to IP addresses. Garmin is headquartered in
Kansas. So the possibility can't be immediately ruled out based on where
the
poster is posting from.

Well, this is Usenet and I think that if it was really a garmin rep/engineer
why would they hide themselves with a hotmail account!

It will take a lot for Garmin to win my opinion and respect of them and
their products/systems if ever. Due to maybe a software flaw in their system
it could have taken my life or someone else's and probably will take a
someones in the future lets hope it's not mine.

I will still fly the G1000 for customers that hire me me to fly them I will
Just be more cautious of them!

Would I take my family or friends in one IFR conditions no way not ever!!
Not until Garmin/Cessna can prove to me that the G1000 will operate for 100+
hours without some sort of bug or failure!! Asking me to place my family in
one in IFR conditions would be the same as asking me to put a gun to their
heads and play Russian Roulette with them and I take my family in IFR
conditions in my cessna 150!

I have also been in contacted by an aviation publication about my experience
on the G1000 not sure if I want to do the interview or not? I know I should
just to expose that there maybe a potential fatal bug/flaw in the system!

If Garmin would like more info from me I am instructed to tell them to put
it on paper and place the request in the mail!

Now I am done talking about the G1000 and it's problems, Lets all talk about
other aviation related things!

"John Theune" wrote in message
news:M6jVg.8501$pS3.1688@trnddc01...
Snip And I would not expect a engineer from Garmin to use his real name
and email address because as a engineer I doubt he is authorized to speak
for the company. That does not stop him from commenting in a much more
knowledgeable way then a lot of the posts about this issue.

As I said in my last message If Garmin would like more info from me I am
instructed to tell them to put
it on paper and place the request in the mail!

Stefan wrote:

Matt Whiting schrieb:

I'd prefer it for all flights given the importance of fuel supply in
an airplane and given the fairly high rate of fuel exhaustion
incidents. I especially want redundancy with a system as fragile as
the G1000 appears to be.


Ok, so you want the FAA jump in and require full redundancy on all
instruments for all privately operated light singles to be considered
airworthy? I'm not sure you really want this.

I never said I wanted the FAA involved. Where'd you fabricate that one
from?

Matt

I have also been in contacted by an aviation publication about my experience
on the G1000 not sure if I want to do the interview or not? I know I should
just to expose that there maybe a potential fatal bug/flaw in the system!

I think you should. But... be sure it's a respected publication with
even handed coverage, and be sure you have all your facts right as well
as you can, all your speculations identified as such, and all your ducks
in a row.

Jose
--
"Never trust anything that can think for itself, if you can't see where
it keeps its brain." (chapter 10 of book 3 - Harry Potter).
for Email, make the obvious change in the address.