midcap
Pattern Altitude
that's true....but I am assuming yo don't know when you are going to punch through until you get up there.
Auto Pilot and IFR
- Thread starter midcap
- Start date
that's true....but I am assuming yo don't know when you are going to punch through until you get up there.
What makes you think you treat them any differently than I do (I fly way more approaches by hand than I do coupled; I need the workout), except of course, for that bolded statement regarding something I never do.
If you do abdicate your responsibility when using an autopilot, if you don't know what the autopilot should do before it does it and confirm it does what it should, being ready to hit that disconnect button, in any phase of flight, you are a disaster waiting to happen.
I still think coupled approaches save workload. That's why they are available. I know in most cases (obviously not all) the basic ability of an autopilot to hold course and altitude better than I is a workload saver. YMMV.
Probably except for a few days here and there in certain geographic areas, most Part 91 light aircraft IFR takes place in blue skies. Even if the weather is technically not VMC.
Heck, if the Skew-T tells me climbing a couple thousand feet will bring me this, why the heck wouldn't I? This is a 2-hour flight that could easily have been more than half in the soup since the clouds were there all the way. As it was, only about 0.4 was actual instrument, most of it during an instrument approach which itself ended at about 900 AGL.
There are tools out there that will allow the prediction of cloud tops with a pretty good degree of accuracy. Just as pilots can file altitudes based on forecast winds, they can also file altitudes based on forecast cloud heights.
In that photo I posted, I had a pretty good idea that's what I would see and even told my wife to watch during the climb because breaking through the clag into sunshine is never boring (to say the least).
OTHO, on a recent currency flight with a friend on a good IFR day, we selected an altitude that we expected would keep us in the clouds most of the time and were not disappointed.
Definitely not foolproof. But perfection is not the goal. "More likely than not" is pretty good.
Good timing: the latest Callback, the periodical that reviews issues raised by NASA ASRS reports, discusses automation dependency. Primarily situations in which pilots have abdicated PIC responsibility to the boxes rather than taking a more active supervisory role.
ASRS often receives reports suggesting that aircrews believe their automation is accomplishing a desired task when, in actuality, it is not.
BarryCooper
Pre-takeoff checklist
90% of my flying is single-pilot XC, so my rule is IMC=A/P on
Checkout_my_Six
Touchdown! Greaser!
yup....said the guy who didn't press the CDI button on the coupled LPV approach.
now I know why it didn't couple.....
midcap
Pattern Altitude
where would I find those weather tools?
In part, look into Skew-T charts. They're a little intimidating at first, but once you learn to read them they're rather useful (and pretty accurate).
Indiana_Pilot
Line Up and Wait
I would love to have a better A/P in my 180 but it's a single axis.. I will somewhat hold a heading but can track a VOR or LOC fairly well. I never seem to have a problem without it but it would be very nice to be able to take some work load off me..
midcap
Pattern Altitude
ok thanks, I'll check them out.
midcap
Pattern Altitude
I'd imagine a single axis is a wing leveler and yours also tracks a VOR/LOC?
That seems like those two things would be very helpful while in hard IMC.
For an excellent intro, google "weather in the vertical" on YouTube. As @bradg33 said, they can be a little intimidating at first since one of the goals of the Skew-T is to depict atmospheric instability as a way to help predict the likelihood of thunderstorms. That's pretty important so there's an underpinning of stuff that can be heavy for some. But the use of the chart to show likelihood of clouds and clear and icing (always in conjunction with other tools, even the good, old standard weather briefing) is one of the simpler capabilities.
mulligan
Cleared for Takeoff
Sully uses auto pilot. Just saying
Sent from my iPhone using Tapatalk
Sent from my iPhone using Tapatalk
MAKG1
Touchdown! Greaser!
- Joined
- Jun 19, 2012
- Messages
- 13,411
- Location
- California central coast
- Display Name
Display name:
MAKG
He flies at least two of them.
BIG difference.
The equation is not at all the same between a twin engine swept wing large FBW jet airliner and a 2000-3000 lb single engine piston, even with the most advanced avionics available.
One big variable is redundant attitude references. For the autopilot in a light single, you generally have one AHRS or one vacuum AI (or electric TC), and that's it. Lose the one input, and the autopilot behaves oddly or not at all. That's not legal for Part 121.
I'm pretty sure the ASRS report I linked includes airlines.
MAKG1
Touchdown! Greaser!
- Joined
- Jun 19, 2012
- Messages
- 13,411
- Location
- California central coast
- Display Name
Display name:
MAKG
Of course it did.
Automation surprises can happen to anyone using automation.
But it's orthogonal to the observation that "Sully was using an autopilot," which is still irrelevant to a piston driver.
MAKG1
Touchdown! Greaser!
- Joined
- Jun 19, 2012
- Messages
- 13,411
- Location
- California central coast
- Display Name
Display name:
MAKG
All that does is dig your hole deeper. You can very easily get task saturated with an autopilot, especially in the context of automation surprise.
I agree with you completely but think you miss his point - if the system isn't perfect, if you actually have to watch what it is doing, it's useless.
That's not an incorrect perspective, just shows people are different.
gsengle
Pattern Altitude
- Joined
- May 9, 2016
- Messages
- 2,099
- Display Name
Display name:
Gsengle
No that's not my point at all. Monitoring is a different task than having to do it. It takes more brain power and can lead to task saturation, so it's entirely different.
Sent from my iPhone using Tapatalk
Sent from my iPhone using Tapatalk
Huh? It looks like that was in response to my post, but I guess not. Unless I missed your point completely. I though you said that the automation leaves you free to handle things better, including things that go wrong with the automation, which I agree with.
gsengle
Pattern Altitude
- Joined
- May 9, 2016
- Messages
- 2,099
- Display Name
Display name:
Gsengle
Your post said you agreed and then proceeded to disagree.
I disagree with you that something is useless because you have to watch it. You have to watch your engine too, but it sure is useful!
Sent from my iPhone using Tapatalk
I disagree with you that something is useless because you have to watch it. You have to watch your engine too, but it sure is useful!
Sent from my iPhone using Tapatalk
MAKG1
Touchdown! Greaser!
- Joined
- Jun 19, 2012
- Messages
- 13,411
- Location
- California central coast
- Display Name
Display name:
MAKG
Ever had an automation surprise on a complex approach?
They are a different type of loss of SA, but still a loss of SA, and very serious.
And when that loss of SA involves sequencing, it can interfere with going missed and getting out of there to sort it out. That's why your hole is deeper.
One nasty on I saw had the same VOR at three points in the approach, and it sequenced to the wrong one.
Task saturation is always serious and a complex management problem combined with an automation surprise is VERY likely to put you there.
Last edited:
mwagg737
Pre-takeoff checklist
- Joined
- Aug 13, 2016
- Messages
- 279
- Display Name
Display name:
Mooney man
I have a Mooney with the factory PC system. Presently,I am installing the ACCUTRAK autopilot. $1200.00
Sent from my iPad using Tapatalk
gsengle
Pattern Altitude
- Joined
- May 9, 2016
- Messages
- 2,099
- Display Name
Display name:
Gsengle
Sure, plenty of surprises including automation issues. And I know if Otto is flying it's easier to keep on top of the big picture and spot things that aren't right. ymmv
Sent from my iPhone using Tapatalk
Sent from my iPhone using Tapatalk
I guess that depends on what you mean by "complex". I've seen surprises in the middle of procedure turns, DME arcs, as well as ATC surprises that make the whole thing need to change. So far, never more than a momentary distraction in which to push the disconnect button. I have an approach chart in front of me and know what is coming next in the same way as I do when hand flying.
Got a specific example of an approach and where the automation surprise occurred? Since I like to do an automation failure when I give an IPC, it would be a big help.
DutchessFlier
Line Up and Wait
Bought the plane with a pretty basic Century IIB autopilot system. It worked okay, would hunt some, but had the radio coupler and components tested and upgraded. Much better. Added GPS roll steer and never looked back. Coupled approaches are pretty cool. Of course, you fly with the AP and without the AP to stay proficient in both modes. In hard IMC, especially around here single pilot with regularly getting course changes and re-routes it's almost necessary when you get the famous 'ready to copy full route clearance'.
MAKG1
Touchdown! Greaser!
- Joined
- Jun 19, 2012
- Messages
- 13,411
- Location
- California central coast
- Display Name
Display name:
MAKG
Well, it isn't anywhere near you, but the one I was referring to was the KCCR VOR 19R, and it sequenced to the inbound leg as we approached from COLLI. The PT was plotted, but it commanded a nearly 180 deg turn inbound immediately after crossing CCR VOR, and switched the CDI, while there was still a PT to do. IF you know what the hell it's doing, it's not hard to get out of. It's a big simplification to presume that's going to happen as you manage that mess of an approach. To make it worse, you WILL get an extra handoff, as there are two different TRACONs involved. Feeder is NorCal, approach is Travis. Guess when the handoff occurs.
The other local one was KLVK ILS 25R, where the GS repeatably captures early and places the aircraft aggressively pitch up in an effort to "get back on" the GS. I suspect a reflection off terrain; the specific location is in a pass. You don't see the erroneous signal in the CDI, so it's a complete nonissue when hand flown.
Thanks. I've seen moves by autopilots and flight directors similar to both of your examples.
That's the part we disagree on. I don't presume the automation will do it wrong but but I do know in advance what the correct thing to do is. Just as with hand flying, I am thinking, "when we get to CCR, we will turn to a course of 011 degrees". That puppy starts turning somewhere else, and I hit the disconnect and had fly.
What unit is that? If one of the Garmin or Acidynes that has a PC simulator, I'd like to see that one in action.
gismo
Touchdown! Greaser!
Given that less than 0.1% of your flying involves an autopilot I'm not sure you're qualified to comment on their usefulness.
IMO the value of an autopilot increases with the amount of attention required for onboard systems. When you're digging through the charts (or iPad) to analyze the latest re-route from ATC while operating the onboard radar and reprogramming the FMS like navigator, having a copilot or autopilot to keep the airplane right side up and pointed in the approximately correct direction is almost invaluable. Same with dodging TRW, gusty winds, and rain or managing ice on an approach. There are indeed pilots who can do all that while hand flying but I know that even though I'm comfortable with my ability to do that without help the autopilot can do a better job than I can.
It also depends a bit on the airplane itself. A Cessna 182 is so stable that you'd be hard pressed to tell if the autopilot had disconnected five minutes ago while a Bonanza will let you know within seconds by rolling off in one direction or another while climbing or descending a few hundred feet.
But autopilots have downsides beyond the cost to install and maintain them. A phrase familiar to almost anyone with significant AP experience is "What is it doing NOW?" because autopilots are a lot like computers, they tend to do what you tell them rather than what you want. Thus anyone flying with one really needs to spend some time learning the idiosyncrasies, limitations, and capabilities of the unit they're flying with. There's also the issue of skill atrophy. If you rarely hand fly you can become dependent on the autopilot and given that they can quit or worse yet malfunction at a critical time those skills need to be deliberately maintained through recurrent training or frequent use.
MAKG1
Touchdown! Greaser!
- Joined
- Jun 19, 2012
- Messages
- 13,411
- Location
- California central coast
- Display Name
Display name:
MAKG
It's a G1000/GFC700, and the simulator flies it correctly.
But I do think you're proving my point. You do all the same things with the autopilot on as with it off, except for the actual muscle. You have the same brief, the same scan, the same radio calls, etc. And you add a bit to make sure it isn't screwing up, plus a much more complex interface. In what sense does this save work on an approach?
Last edited:
gsengle
Pattern Altitude
- Joined
- May 9, 2016
- Messages
- 2,099
- Display Name
Display name:
Gsengle
Again, monitoring is easier and less brain cycles that keeping the wings level yourself, calculating wind angles, etc. it's not the same and with any experience with a good autopilot it's obvious.
Sent from my iPhone using Tapatalk
Sent from my iPhone using Tapatalk
RudyP
Cleared for Takeoff
I know you said generally and that's true of the legacy fleet but I think most newer planes (certainly almost all new Cirri) have dual AHRS and dual ADCs and will CAS any miscompares before acting up.
Given that less than 0.1% of your flying involves an autopilot I'm not sure you're qualified to comment on their usefulness.
He probably almost never flies IFR XC. Here is a fun little game to play with someone who is macho and claims they fly long distances IFR without an AP: Find their tail number and look up their flights.
Usually you find maybe one or two over the last three years.
Fearless Tower
Touchdown! Greaser!
That reminds me of the Baron. You could have it trimmed perfectly flying straight and level and then simply lean your body to the side and the airplane would roll into a turn.