We need to rethink the 'Infinite Flight autopilot'

In terms of implementation, the specifics of the particular type of aircraft is presumably a relevant factor: how to tailor AP response (the “umbrella AP” controlling all IF aircraft using AP) to unique engine response time and usable thrust relative to inertia, plus the unique drag profile at the time of the power change (for the particular aircraft in use at the time).

If a gradual throttle-up is not “tuned” appropriately for a particular aircraft, for example, one aircraft may be fine, but another might have an excessive drop in IAS before it reaches equilibrium of forces, for example.

The above is not claiming smoother power response can’t be done, it’s outlining potentially, why the issue involves some complexity (consumption of resources to implement).

The most effective UI might not be complicated in concept on the surface, by being paid for with a lot of unseen complexity “under-the-hood.” But human-factors-friendly is obviously hugely important.

I assume you mean VNAV doesn’t set altitude in the irl ALT mode. But VNAV obviously sets altitude internally to its own independent function (the profile):

Again, the above relating how the independence of function you describe adds to the overall functionality: the independent modes provide more powerful combined functionality by each being independent of the other.

Again, your example of the power of independent mode functionality:

But somewhere you have altitudes entered before the time of transferring each such goal to the AP, so that you don’t have to be scrolling through numbers each time you need the next alt?

Yeah, I think I misunderstood someone else’s point on that recently.

Again, it would be very cool to see that.

A given aircraft at a given load, density altitude, power setting, and IAS, has a unique pitch for a given VS (in unaccelerated flight).

So, your point is I presume, in spite of the above, for any aircraft as long as IAS shows any hint of deceleration, the system feedback is: keep lowering pitch until reaching the VS necessary to halt the IAS decrease*, and then eventually (at some altitude) you arrive at the pitch for VS = 0? Even if thrust output changes with altitude, so aircraft dependence should, one presumes, be little to none? (Am I missing something I might think of later?)

Though there is still the question of how the above might affect “gameplay” across a wide user group.

*edit: I don’t know with complete certainty there wouldn’t be some comparative rate of response issue in the IAS’s feedback for how different aircraft are modelled (a lot of things seem pretty simple until you test the implementation).

Thanks, I’ll have a better look at these comments shortly and respond as needed

The simplicity of this feature request is this: Infinite Flight has come a long way… But considering the lack of attention to this over the past 5+ years, I would like to see Infinite Flight take a better look at how real world autopilots function, and, in accordance with any findings, implement any features into the app within reason considering the limitations of the system, people’s devices, and in consideration with IFs users

Voted. I mean… some of these changes are complex (e.g. A/T vs Speed vs Pitch control loops for various aircraft), but your nine original posts in May are low-hanging fruits that, in my opinion, deserve their overdue attention.

1 Like

@Cody_M, one other pain point that I think could be added to the low-hanging fruits list:

When HDG/LNAV is engaged, the bank angle shall be limited to 30 degrees or that that produces the standard rate turn, whichever is smaller.

This seems to be a straightforward logic to implement and could be applied to all aircraft models. While I can manage without creating a hold in the FMS (or rather, the Flight Planning page), it would greatly improve my flying experience if I didn’t have to manually fly the hold due to the inability to perform a standard rate turn, especially on Cat C and above aircraft.

Additionally, executing a VOR approach with a procedure turn would be less challenging as I might be able to avoid overshooting my inbound radial for once. :D

1 Like

Micro bump

1 Like

bumping this. we need some dev to let us know what they think about it. doable or not?please

3 Likes

I am so annoyed by the current autopilot, I vote with all my might for the implementation of this idea!

2 Likes

How vertical speed works irritates me. Adjusting fpm either way (+ve or -ve) to a certain value is almost never gradual.

Case 1
Going from -500fpm to -1000fpm? It’ll probably overshoot to -1200fpm before stabilising.

Case 2
Going from -500fpm to -1000fpm? It’ll reduce first to sth like -300 fpm before increasing to the set value.

Voted!

I get quite annoyed at the current autopilot system, especially since I strive to make my flights (at least on the expert server) as realistic as possible. I’ve managed to work around most of these problems, although some are honestly impossible to! But I’ve never seen or been in an aircraft that dumps its passengers all to one side of the plane just to capture the localizer or G/S!

The heading bug resetting to your current heading whenever armed is also really annoying. It makes it almost impossible to prepare a non-LNAV turn, such as when directed by ATC, smoothly, as the aircraft starts to turn as soon as you move the heading.

That being said, I’m still very grateful for what we have in game, I just hope this can get sorted sooner rather than later, as it would benefit the game immensely!

5 Likes

I have been pretty annoyed by the coupling of the altitude and V/S modes lately. I was wiping something off of my screen one time and accidentally disengaged either the V/S or the altitude mode and my airplane was falling off the sky. Oh well…

3 Likes

Does anyone else hate how violently the APPR mode captures glideslope. it pitches up violent then down.

6 Likes

haha yes. It’s also more violent in some aircraft than others.

1 Like

I uses those as a indicator that there has been a change in modes.

I’ve found that when you disengage LNAV in order to comply with ATC heading instructions (for example), if you slowly change and time the heading (by + or - 1), the plane changes direction smoother.

Though it might take a bit longer to change from lets say 180 to 240, the bank angle seems more realistic. Or just manual fly

1 Like

Yeah, I’ve found that manually flying and adjusting the heading bug while doing so, so that I can line up with the correct heading easily, is a good way to overcome that particular hurdle.

I think IF AP design is perfect for mobile users, but the way it works is far from good. Its very aggressive with the yoke and throttle.

It would be very nice if it had a “climb button” So that the VS will dynamicaly change based on the throttle you gave it, so it maintains speed and climbs, without controlling the throttle. Kind of like the real 738.

If you dont like this, you can always still climb like you want without using the climb button.

1 Like