ROVER WHEELS: Monitoring changes over time, NOTE: Read back through the thread to avoid repeating misconceptions Options

[Gerald]

1. Was there a design spec on having to navigate dunes? Is so what was it?

Yes.

[testguru]

Thanks for that link.

The video indicates that the rover can do about a 20 degree slope in soft sand if I am understanding properly. Was there a hard spec on the maximum angle that the rover could traverse in soft sand?
I watch this site every day and get a feel for the anxiety that everyone has felt about trying to drive over a ~ 3ft dune. Now that Curiosity has done that I assume the anxiety will lessen for the same size dune in the future?
Since it is pretty much impossible to perfectly simulate Mars on Earth I guess we will never know for sure what the limits are on dunes until we measure the result for the first time with the rover on Mars.
I also understand the reluctance to find those limits.

Pretty much anywhere we go on Mars is going to have sharp rocks and dunes.

On the next rover mission we should be able to place a larger payload on Mars using the same rocket due to the more favorable launch window, (pretty much the best perihelic opposition for the next 13 - 15 years).

I hope that the wheels can be beefed up or replaced with something better. Folks have been saying that the next mission will not be a carbon copy of this one or even close.

[Explorer1]

Depends on the landing site surface, which is far from decided. It could be like Meridiani for all we know.

[PaulH51]

Another nice set of MAHLI wheel images has been down-linked from sol 0537. One advantage with this set is the light dusting of sand on the inside of the wheels, that has helped to differentiate between small punctures and reflected light off the dimples which has sometimes been interpreted as punctures.

[RoverDriver]

...
Now that Curiosity has done that I assume the anxiety will lessen for the same size dune in the future?
...

One would think but Mars has proven to be a nasty place with surprises every now and then. We have learned some hard lessons so we never let our guard down.

Paolo

[Explorer1]

A few quick seconds in this update show the rock tests in the Mars yard:

http://www.youtube.com/watch?v=PiBbFC4Isr0

[atomoid]

no luck trying to find MAHLI pairs to create a stero view, but rolling along, heres a gif from SOL 546..

[Gerald]

Sol 549 MAHLI took 5 images for each wheel. So I think, there is a chance to get a complete coverage of the wheel surfaces with this series.

https://xkcd.com/1319/

The manual way to determine the parameters turned out to be rather time-consuming and difficult, since to get a good result, it seems, that I need to go to subpixel accuracy with at least six parameters.
So I've to go through these rethinking cycles of automization. That's time-consuming, too, but interesting, as well.

Here one of my tries to get an automated 2d-vectorization of features of this Sol 513 image, which one of the algorithms thinks, might be parts of a rover wheel:

(looks funny, almost like a comic, so I thought I should share it)

The algorithm is designed to be mostly independent of specific light conditions.

That's an intermediate step, when trying to match a simulated image of a 3d-model of a wheel with the raw image by kind of RANSAC. Still quite a way to go...

[elakdawalla]

I posted this in my blog yesterday, just looking at 2 of the wheels, the left middle and left front. I paired sol 513 and 546 images in similar positions to compare them. While I hate to suggest that any more pixels be devoted to wheels than already are, it does seem that a 5-position series would do better at completely characterizing them. Or maybe just a bit more rotation between adjacent pairs -- but then it would be harder to compare one sol to another.

[Gerald]

That's nice! Some of the pairs can almost be used as x-eyeds for comparison.

The exact same (relative) wheel and camera positions for corresponding images would be ideal for monitoring. But I doubt, that this is technically possible with reasonable effort.

For a longer-term sol-by-sol investigation I'm still hoping to be able to solve the automization. This could then be used for consecutive image processing, e.g. for marks, where changes occurred. It would also be easier to compare MAHLI with Mastcam images.

[elakdawalla]

There's a new hole in the right middle wheel, at position 8 near the center, between sols 546 and 554. It's easiest to see the before-and-after comparison in the bottom pair of images, where the holes are backlit.

[xflare]

Could the drive along the edge of that giant bank of dark sand dunes?

[atomoid]

finally some sloppy wheel crosseyes from SOL 559 & 560 MAHLIs

[Eric H.]

Here's a worrying photo from S564.

[eoincampbell]

I'm sure the team are keeping a close eye on the shard produced by this tearing ...