Use of Autopilot in Competitions - is there a performance benefit gained? (Again ;-)

Moderators: Uros, Tom, OXO

6266
Posts: 925
Joined: Tue Aug 25, 2020 7:07 am

Re: Use of Autopilot in Competitions - is there a performance benefit gained? (Again ;-)

Post by 6266 » Fri Nov 11, 2022 11:52 am

Last thinking about that:

The sinking of the polar at the manually flown speed (0,7017) multplied with the relation of the autopilot flight (1,037) gives a sinking of 0,727

That is 1,2 cm/s less than manually flown (0,739) and matches really good to my post above with the difference of 1 cm/s ...
Visit https://www.baleit.no

Vintage Series 24, The Journey
25 landscapes with free available textures, newest: Maine

Discord https://discord.gg/bAXJaRT6ZC

User avatar
wickid
Posts: 2429
Joined: Mon Dec 04, 2006 7:32 pm
Location: Venlo, NL
Contact:

Re: Use of Autopilot in Competitions - is there a performance benefit gained? (Again ;-)

Post by wickid » Fri Nov 11, 2022 1:40 pm

1.6% is probably within the margin of error.

To put it into perspective:

1 cm/s is 36 meters per hour. Or about 12 seconds time loss with 3 m/s thermals.
PH-1504, KOE

User avatar
Bre901
Posts: 1790
Joined: Tue Nov 22, 2016 8:57 pm
Location: Annecy (France)
Contact:

Re: Use of Autopilot in Competitions - is there a performance benefit gained? (Again ;-)

Post by Bre901 » Fri Nov 11, 2022 1:56 pm

wickid wrote:
Fri Nov 11, 2022 1:40 pm
...Or about 12 seconds time loss with 3 m/s thermals.
... which is less than the turn too much in the thermal which will make the difference
CN: MPT — CondorUTill webpage: https://condorutill.fr/

6266
Posts: 925
Joined: Tue Aug 25, 2020 7:07 am

Re: Use of Autopilot in Competitions - is there a performance benefit gained? (Again ;-)

Post by 6266 » Fri Nov 11, 2022 2:11 pm

viewtopic.php?f=29&t=21337&start=15#p180217

12 seconds, podium or not. For me it's the same, I'm only interested in good processes
podium.png
You do not have the required permissions to view the files attached to this post.
Visit https://www.baleit.no

Vintage Series 24, The Journey
25 landscapes with free available textures, newest: Maine

Discord https://discord.gg/bAXJaRT6ZC

cbrunschen
Posts: 37
Joined: Mon Dec 09, 2019 2:48 pm
Location: London, UK
Contact:

Re: Use of Autopilot in Competitions - is there a performance benefit gained? (Again ;-)

Post by cbrunschen » Fri Nov 11, 2022 2:38 pm

So the autopilot does offer a measurable, and potentially critical, performance advantage.

I'd prefer to compete on things other than who can best use the autopilot.

So for a task-setter to be able to either completely disable, or apply penalties for, the use of the autopilot continues to seem like a good idea.

To prevent the use of the autopilot for performance gains but still allow it for necessary brief interruptions, being able to specify a grace period seems like it strikes a reasonable balance.

Or alternatively, being able to specify not just a linear number of penalty points , but a geometrically increasing penalty could be useful; making a small rule-break inexpensive, but a bigger one significantly more expensive. And that might be useful for any of the penalties, really – making prolonged or repeated infringements increasingly expensive.

User avatar
wickid
Posts: 2429
Joined: Mon Dec 04, 2006 7:32 pm
Location: Venlo, NL
Contact:

Re: Use of Autopilot in Competitions - is there a performance benefit gained? (Again ;-)

Post by wickid » Fri Nov 11, 2022 4:04 pm

cbrunschen wrote:
Fri Nov 11, 2022 2:38 pm
So the autopilot does offer a measurable, and potentially critical, performance advantage.
Well no. In one test (2 flights with the AP flight at different speed to the manual flight) 6266 found a 1,6% advantage. He mentions flying different headings with wind present. That affects the L/D as well. If you fly with more headwind component you will have a lower L/D. So if he accidentally flew more seconds with headwind while in manual flight versus with autopilot that could affect his results. The difference he found is so small I would be be careful with drawing the conclusion that it is purely down to the autopilot.

What the autopilot does is maintain wings level. So it is constantly steering to maintain wings level. Every movement causes drag. You could argue that a human pilot who just lets the plane do its thing and only corrects gross errors creates less drag.

I think the test is better if you use turbulence, but no wind. And make sure the test is flown at exactly the same trimmed speed. Right from release.
PH-1504, KOE

User avatar
Paul_UK
Posts: 828
Joined: Sat Aug 29, 2009 1:05 pm
Location: Southampton, UK

Re: Use of Autopilot in Competitions - is there a performance benefit gained? (Again ;-)

Post by Paul_UK » Fri Nov 11, 2022 4:56 pm

wickid wrote:
Fri Nov 11, 2022 4:04 pm
cbrunschen wrote:
Fri Nov 11, 2022 2:38 pm
So the autopilot does offer a measurable, and potentially critical, performance advantage.
Well no. In one test (2 flights with the AP flight at different speed to the manual flight) 6266 found a 1,6% advantage. He mentions flying different headings with wind present. That affects the L/D as well. If you fly with more headwind component you will have a lower L/D. So if he accidentally flew more seconds with headwind while in manual flight versus with autopilot that could affect his results. The difference he found is so small I would be be careful with drawing the conclusion that it is purely down to the autopilot.

What the autopilot does is maintain wings level. So it is constantly steering to maintain wings level. Every movement causes drag. You could argue that a human pilot who just lets the plane do its thing and only corrects gross errors creates less drag.

I think the test is better if you use turbulence, but no wind. And make sure the test is flown at exactly the same trimmed speed. Right from release.
I agree. The only way to test initially is with zero weather. No wind, no thermals, no turbulence etc etc.

It would also be interesting however to understand if using autopilot does provide more benefit in more adverse weather conditions. That one is a little trickier to conduct tests for.

Peter, you'd really need to refly the same track manually numerous times to increase your sample size to be sure the 1.6% is constant. I'm not going to suggest you do that though as I know how much time that'd use up!
Image
2018 - Silver Badge complete - Forgot I could claim for the 100km diploma flying an out and return :oops:
2019 - Get up to speed and aim for a 300km task?! maybe.....

6266
Posts: 925
Joined: Tue Aug 25, 2020 7:07 am

Re: Use of Autopilot in Competitions - is there a performance benefit gained? (Again ;-)

Post by 6266 » Sat Nov 12, 2022 8:02 am

Yes, it was only one flight, I mentioned that this is not scientific
wickid wrote:
Fri Nov 11, 2022 4:04 pm
He mentions flying different headings with wind present. That affects the L/D as well. If you fly with more headwind component you will have a lower L/D. So if he accidentally flew more seconds with headwind while in manual flight versus with autopilot that could affect his results.
In red the manually flight, blue the autopilot. The autopilot flight is heading more against the wind
flights.png
wickid wrote:
Fri Nov 11, 2022 4:04 pm
What the autopilot does is maintain wings level. So it is constantly steering to maintain wings level.
I never flew autopilot before, is it really only the wings level? With autopilot the speed was more constant than I was able to fly manually, so I guess the autopilot controls the speed too
wickid wrote:
Fri Nov 11, 2022 4:04 pm
I think the test is better if you use turbulence, but no wind.
Agree, I thought to make a weather that most impacts the flight but of course the wind impacts the L/D too
wickid wrote:
Fri Nov 11, 2022 4:04 pm
And make sure the test is flown at exactly the same trimmed speed. Right from release.
LOL, until you wrote this I was happy with my only 2 % difference in speed. For sure I'm either a top pilot or a robot. If I could fly like the autopilot, where will be the advantage for the autopilot?
Paul_UK wrote:
Fri Nov 11, 2022 4:56 pm
Peter, you'd really need to refly the same track manually numerous times to increase your sample size to be sure the 1.6% is constant. I'm not going to suggest you do that though as I know how much time that'd use up!
The test including flights, calculation and writing needed nearly two hours. And now one hour more ... There were some things I never did before, so I was really interested to do it. Flying online offline (hosting a flight on the same pc like flying, very easy to do, but never done) and using the autopilot. But there will be no more test flights done by me. If someone is really interested, I showed the way how it can be done.


What I think the biggest advantage of the autopilot is the relaxing. While the autopilot flight I could read and answer some emails on another pc. The manually flight needed my concentration, and I only had to control the speed and the heading (the flight was over the see, no points on the landscape for orientation).

In my dreamworld there will be no rules for using helpers, because all will have a code of honor and use only the helpers they really need. So no top pilot will use the autopilot (only for the bio brake), but another one needing it, can use it without penalties or discussion about it.

That this dreamworld doesn't exist in Condor either I learned in a group allowing the Q for that all have fun and can finish. A top pilot of the group winning a lot of races there couldn't accept to outland a few meters before the finish and used Q getting 200 points penalty. I finished later (of course without Q, I would never use it) and was in the list below that top pilot. If I remember right it was my last flight with that group.

My suggestion to solve the problem of this thread is to give a message in the chat each time a pilot activates the autopilot. It exists today with using thermal helpers, I think. I believe the flying group will then manage the problem better than every penalty or prohibition
You do not have the required permissions to view the files attached to this post.
Visit https://www.baleit.no

Vintage Series 24, The Journey
25 landscapes with free available textures, newest: Maine

Discord https://discord.gg/bAXJaRT6ZC

User avatar
wickid
Posts: 2429
Joined: Mon Dec 04, 2006 7:32 pm
Location: Venlo, NL
Contact:

Re: Use of Autopilot in Competitions - is there a performance benefit gained? (Again ;-)

Post by wickid » Sat Nov 12, 2022 8:28 am

6266 wrote:
Sat Nov 12, 2022 8:02 am
wickid wrote:
Fri Nov 11, 2022 4:04 pm
He mentions flying different headings with wind present. That affects the L/D as well. If you fly with more headwind component you will have a lower L/D. So if he accidentally flew more seconds with headwind while in manual flight versus with autopilot that could affect his results.
In red the manually flight, blue the autopilot. The autopilot flight is heading more against the wind
flights.png
[/quote]

So there is wind and you didnt start at the same position? Difficult to compare the flights in that case.
wickid wrote:
Fri Nov 11, 2022 4:04 pm
What the autopilot does is maintain wings level. So it is constantly steering to maintain wings level.
It keeps the stick centered at the trimmed speed, just like a normal joystick would. It doesn't really actively steer in pitch. If you enter a thermal your angle of attack will increase because the "relative wind" is suddenly coming more from below. This will cause the plane to accellerate to get back to the trimmed AoA. Vice versa when entering sink. So it doesn't keep speed accurately, just keeps stick centered. I think it does dampen the natural long period pughoid oscilation that is a natrual pitch oscillation that occurs in almost all planes.

In roll it actively steers to wings level. So if you get an upset on one side it returns to wings level, but doesn't steer back to the original heading. Also if your rudder isn't perfectly centered (small calibration offset) it will do a slow skidding turn keeping wings level.
PH-1504, KOE

6266
Posts: 925
Joined: Tue Aug 25, 2020 7:07 am

Re: Use of Autopilot in Competitions - is there a performance benefit gained? (Again ;-)

Post by 6266 » Sat Nov 12, 2022 9:34 am

wickid wrote:
Sat Nov 12, 2022 8:28 am
So there is wind and you didnt start at the same position? Difficult to compare the flights in that case.
I used the same flight plan and started airborn in multiplayer (because of the autopilot), I don't know how to have more control about the exact start position. It's for sure a good Condor feature in an airborn multiplayer flight, not to start all planes at the same position, but not my mistake
Visit https://www.baleit.no

Vintage Series 24, The Journey
25 landscapes with free available textures, newest: Maine

Discord https://discord.gg/bAXJaRT6ZC

Fingers
Posts: 35
Joined: Wed Sep 01, 2021 11:34 am

Re: Use of Autopilot in Competitions - is there a performance benefit gained? (Again ;-)

Post by Fingers » Sat Nov 12, 2022 1:16 pm

Personally I don't use auto pilot at all. I understand the possible need for it and think it would be very easy for the Devs to limit the total use in a flight to 5 mins for eg. Or if a task setter wishes to completely disable it then why not.

Apart from the fact your instantly turning a SIM into an arcade game by pressing P. There is an advantage because your able to disconnect your brain, giving you a rest.

I can't watch the You Tuber's who permanently have it engaged unless they are turning. Toe curling.

JLN
Posts: 114
Joined: Wed Jun 17, 2020 5:46 pm

Re: Use of Autopilot in Competitions - is there a performance benefit gained? (Again ;-)

Post by JLN » Sun Nov 13, 2022 4:31 pm

Untill yesterday I would have said that I had never used autopilot. And I do not understand using it - except for say 5 minutes by hour for any reason.
I used it yesterday, not for a long time but many times. I did a competition flight without PDA in VR . So if I wanted to consult my printed maps I had to take off my headset, my glasses (I wear glasses in my headset) find my map, put it north up (very important...), look it to find the detail I had just seen, and make the reverse to go back in my cockpit...
As I am afraid of leaving my glider alone, I put the autopilot. and clearly did not get any advantage !
That is also why I ask for a pda with a map (no gps and glider figure in it obviously) with instead of pages making it move north sud, east and west and zoom for "no pda" tasks.
Regards
PS I am also upset when my performance in a flight is weaker than one of a miracle survivor, the worst being a 800 pts one. Ok to the miracle but all users before non users.

6266
Posts: 925
Joined: Tue Aug 25, 2020 7:07 am

Re: Use of Autopilot in Competitions - is there a performance benefit gained? (Again ;-)

Post by 6266 » Sun Nov 13, 2022 10:37 pm

That's offtopic to the thread, but ...

I usually fly without pda, have some paper tables on the desktop and a map. I don't have VR but got an idea from a guy who has nothing to do with flight sims. You can integrate other programs in VR, some does it with xcsoar. Why not integrate a program showing the live video of a webcam? The webcam shows a place on your desktop where you can set in the map or tables or whatever else. You move the papers manually under the camera and you can see it in VR
Visit https://www.baleit.no

Vintage Series 24, The Journey
25 landscapes with free available textures, newest: Maine

Discord https://discord.gg/bAXJaRT6ZC

Filgueiras
Posts: 96
Joined: Thu May 23, 2019 9:24 pm

Re: Use of Autopilot in Competitions - is there a performance benefit gained? (Again ;-)

Post by Filgueiras » Mon Nov 14, 2022 11:30 pm

I do use autopilot. A lot.

I'm a low-rank condor pilot and, in RL a student glider pilot ( who has evaded school in pandemics :mrgreen: ). I1m also private pilot, small aircrafts.

When flying thermals - my usual drill - I turn AP on when going to the next lift. And I click it off just before reaching it,

With AP on, I make small heading adjustments with the rudder - what I know is uneffective, but comfortable. Turns and major adjustments, w/ AP off.

I made some tests, as Bre901 suggested. In total, 4 flights, 2 w/AP, 2 without. Unfortunatelly, I used my home lanscape, witch may not be popular, Bebedouro.

Anyway, here are added an image of the flights together, and the igc files,
TESTE AP.zip
Teste AP.JPG
You do not have the required permissions to view the files attached to this post.
Image

6266
Posts: 925
Joined: Tue Aug 25, 2020 7:07 am

Re: Use of Autopilot in Competitions - is there a performance benefit gained? (Again ;-)

Post by 6266 » Tue Nov 15, 2022 6:39 am

Here the results of your igc files
AP.png
You do not have the required permissions to view the files attached to this post.
Visit https://www.baleit.no

Vintage Series 24, The Journey
25 landscapes with free available textures, newest: Maine

Discord https://discord.gg/bAXJaRT6ZC

Post Reply