Accuracy trend in Falcon 9 booster landings

[rashomon]

Given that there used to be a landing bingo on this site, is anyone tracking Falcon 9 booster landing accuracy, say distance from bullseye center, over time? My impression from watching landings is that it's improving. It certainly seems significant given Super Heavy Booster return accuracy required for the tower catch. Granted that SH will likely have more thruster authority, etc, but it would seem to be something that SpaceX would want to use F9 to learn as much about as possible.

[xvel]

You're assuming they're aiming for the center, that's probably not true.

[wannamoonbase]

I'm sure the data is not public, but I'd like to know the position accuracy of the ASDS compared to the F9 landing.

Land landings have a stable landing spot, so we could look at those, but that's a much smaller sample.

[meekGee]

...and maybe you're plotting the improvement in drone ship positioning accuracy...

[ulm_atms]

You're assuming they're aiming for the center, that's probably not true.

Um....they got to aim for something or the computer can't do the math to land....and the center gives the greatest margin on both X and Y axis for...issues.  I can't see them using anything other then center personally.  If you can explain why they wouldn't use center...we are all ears.

[Bob Niland]

You're assuming they're aiming for the center, that's probably not true.

Yep.; perfect being the enemy of sufficient+margin, and all that.

We would need to read Elon's mind to see if the goal is:
☑ sufficient for F9 ops, or
☐ refine algorithms for SS practice (and perhaps risk F9 ops doing it)

[xvel]

You're assuming they're aiming for the center, that's probably not true.

Um....they got to aim for something or the computer can't do the math to land....and the center gives the greatest margin on both X and Y axis for...issues.  I can't see them using anything other then center personally.  If you can explain why they wouldn't use center...we are all ears.

A rocket when it lands is flying through the air, the air is moving relative to the ground (sometimes called wind) and this movement is not entirely deterministic, so from the rocket's perspective this seemingly stationary "center point" to aim at is not stationary at all, aiming at it would mean that the rocket has to correct all the time, so the RCS fuel would be used up in no time.

[alugobi]

Not really a Starship topic.

[testguy]

You're assuming they're aiming for the center, that's probably not true.

Um....they got to aim for something or the computer can't do the math to land....and the center gives the greatest margin on both X and Y axis for...issues.  I can't see them using anything other then center personally.  If you can explain why they wouldn't use center...we are all ears.

I believe we learned a few months ago that the stage is programmed to land at a coordinate.  The ASDS is programmed to position at the same coordinate.  Therefore what we are probably talking about is both the position accuracy of the ASDS as well as the separate position accuracy of the stage.  There is not a transmitter on the drone telling the stage, “here I am”.  Sea state should be a large driver.  Their is no sea state to take into consideration for Starship and Starship Heavy.

[envy887]

You're assuming they're aiming for the center, that's probably not true.

Um....they got to aim for something or the computer can't do the math to land....and the center gives the greatest margin on both X and Y axis for...issues.  I can't see them using anything other then center personally.  If you can explain why they wouldn't use center...we are all ears.

The pad is way bigger than they need margin for, they could move it 5 or 10 meters with no effect. And repeated landings in exactly the same spot will damage the concrete.

[Vettedrmr]

I thought that they applied an ablative coating to LZ-1/2 to protect the concrete.  ASDS decks are also protected via water, plus the decks are metal and act as heat sinks.

I've not seen anything stating that the landing control software randomizes an error from center to generate an offset to aim for.

[alugobi]

Concrete is easy.  Consider how many times they fixed the landing pad at BC and how quick it was to do.

[envy887]

I thought that they applied an ablative coating to LZ-1/2 to protect the concrete.  ASDS decks are also protected via water, plus the decks are metal and act as heat sinks.

I've not seen anything stating that the landing control software randomizes an error from center to generate an offset to aim for.

Why would it be random and in flight? They could just pick GPS coordinates a few meters away and tell the booster to land there.

[beelsebob]

I'm sure that they will have some level of acceptable margin included in the software .  Hysteresis is important to stop systems chasing their tail trying to extremely precisely position itself.

That said, that doesn't mean that we can't track the accuracy.  The standard deviation circle around the centre of the pad, and how it evolves would certainly be interesting to track.

[rashomon]

This is the February 3 Starlink mission landing, pretty impressive as the droneship is a somewhat moving target. I, too, would love to see a plot of distance from center versus mission number or time.

[Jim]

the droneship is a somewhat moving target.

No, it isn't

[kevinof]

This is the February 3 Starlink mission landing, pretty impressive as the droneship is a somewhat moving target. I, too, would love to see a plot of distance from center versus mission number or time.

Droneship sits stationary over a lat/long and waits for the booster to do its party piece.

[Polaroid]

the droneship is a somewhat moving target.

No, it isn't

Since we have seen droneships roll during landings, what makes you say that?