I wonder if the stuck steering actuator might have helped to break the duracrust? Since the rover isn't moving perfectly straight, the slight sideways motion it experiences might provide the right kind of force to break through. Either that, or there was just too much loose sand underneath, and the weight was just finally enough to break through.

I think the picture posted by Doug http://www.unmannedspaceflight.com/doug_im...o448_tracks.jpg confirms that we didn't "suddenly run into powdery stuff". There seems to be no difference in the soil 2 feet or ten feet back.

I think JES may be on to something here. Does the bogie system have less compliance when moving backwards, and could this have caused the leading wheels to finally dig in rather than floating over the dunes? (This in contradition to my earlier opinion that this is a result of hardware and/or software and/or operator error, which I still hold open as a possibility.)

I'd argue that the stuck steering actuator isnt related, being at the 'back' of the line of three wheels on its side of the rover. they were driving backwards remember.

Not only that but it's only a few degrees of straight-forward anyway - not much more than a bit of excessive toe-out for a racing car

Doug

After looking at all the images I really think it just got high centered. The drivers will have to provide a bit more wheel clearance when cresting dunes from now on.
Maybe somebody could point out something I'm missing but I don't see anything that "suddenly" changed in the soil.
NASA changed the rover software to provide less stops from hazards, to get longer drives which I'm sure contributed to the current situation.

she'll get out and with a bit of tweaking have long drives again (I hope)

scott

This picture taken before getting stuck shows the rover riding on bottom and windward side of the dune. I am assuming that prevailing winds come from the right and the lighter colored deposits on the left side of the dunes is dust settlement on the leeward edge of the crests. Since the angle of approach is very shallow, and the majority of the rover’s weight is on wheels in the trough, there would have been very little pressure on the lead wheel as it crosses the next crest. IF the angle of approach to a new crest is perpendicular, there could be additional load on the lead wheels resulting from the downhill orientation of the rover. The additional pressure could force the wheel through the surface crust, causing the wheels to dig in. The other wheels would then dig in if they continue to turn.



Looking back at photos before entering the etched terrain I see plenty of blueberries suggesting a change in the soil composition. In addition, I see regular wave crests that suggest to me a continuous regular wind from a prevailing direction. I would expect that the windward side of the dunes would be lighter material that is dropped on the leeward side. This could result in a less dense, more recent surface cover at the edge before the crest. (Assuming I have the wind direction interpretted correctly).

I am interested in how the wheels operate. Is there anywhere I can obtain specific information on the rover design diagramming the suspension design, dimensions, load distribution and/or other capabilities and limitations?

As of May 04,05 is op still stuck?

Yes - Latest Squyres Update.

"We're still testing, and it's going to be a few more days, at least, before we're ready to try anything on Mars."

Looks like it'll be next week before they have a got at getting out.

It was my understanding that the rovers move to slowly to get inertial guidance. And even if they can, visual odometery will be more accurate. If they can detect slippage while driving, why would they allow it to dig in so far? That just wouldn't make sense.

Yes, it did do a "wheel wiggy", but that is a safing maneuver. No turn in place was attempted.

ed

I really liked this bit from the update:

"Jeff Biesadecki has led the charge from the engineering side, pulling some very long hours in the testbed driving the rovers into and out of the dirt, working out the best way to do it on Mars."

(My emphasis.)

Here is a colour topography picture of the area around Opportunity. The red area is about 70cm higher than the green. The dune where Oppy is stucked is 30cm high and 2,5m wide and slope 15°. The text is in Finnish.

Topography picture

here is the JPL release, with captions:

http://photojournal.jpl.nasa.gov/catalog/PIA07922

By this view, it really doesn't look like the single dune they're currently over is the reason they dug in. The whole terrain is rising in the southerly direction (little dunes on a big dune?) and they happened to come into a rising small feature at a tilt from starting to climb the rising large feature. Also interesting to see so clearly that they aren't really plowed into the ridge headon but rather straddling it.

"The rover team spent more than a week designing and conducting tests under simulated Mars conditions on Earth before choosing the best way for Opportunity to drive out of the dune. "

That sounds like they have their plan and will be executing it today?

I wouldnt say "today" or "this week" or "thursday" or anything to be honest. We'll see what we see, when we see it.

Remember - it's likely to be a many-sol procedure to get out of the dune - including pauses for a sol or two to look at and compare the data from short drive commands

Doug

Squyers' latest entry adds a lot of detail. Expect to see the rover do a lot of wheel spinning and soil tossing w/o moving much.

http://www.newscientist.com/article.ns?id=dn7356

NASA could take its first steps to rescue its Opportunity rover from a Martian sand dune on Monday.

If everything continues to goes well, ground controllers will begin developing the first escape commands to send to Opportunity on Monday. The first step will be to straighten rover’s wheels, but getting the rover clear of the dune may take days, possibly even weeks.

Also from that article:

"It is 30 centimetres high, slightly taller than the dunes it had previously climbed, which were between 20 and 25 centimetres. The current dune is also steeper, with a 15° incline". (My italics)

I would bet that this had a lot to do with getting stuck.

Chris

What can be done in the future to avoid running into these dunes? Can the rover planners make a guided blind drive route for Oppy to follow to avoid the worst dunes ahead by looking at navcam/pancam images?

Are we even sure it was the dumes that got us stuck?

What else? The rover was on a 90 meter blind drive and after 40 meter it just got stuck in this dune and kept spinning it wheels enough to cover the remaining distance of 50 meter. At the end of the drive it was about to perform a turn in place, but noticed it couldn't turn so it halted.

Thats something I find pretty scary........I suppose it could have been even worse, if the rover hadn't experienced major slippage before reaching it's current spot, the wheels could have continued to spin with the rover reamining stationary, dug in to the dune and completely disappeared.

We can see from the images, that there wasn't a sudden change of texture of the dirt/sand,talcum powder. Also, this dune just doesn't look that more severe than others. Maybe, we were just lucky in the past.

I did notice that on Mar 22, Sunspot noticed something strange and brought it up in post #89 of this thread. Look at the wheel tracks.

http://qt.exploratorium.edu/mars/opportuni...Y2P1969L0M1.JPG

http://www.unmannedspaceflight.com/index.p...topic=776&st=75

I'm sure she'll get unstuck, but looking at the route maps posted, this difficulty was predictable.

A recent Oppy route map clearly shows a change in terrain just past the Viking/Voyager craters: the dunes become more prominent. But a hundred or so meters past that point, the dunes become more subdued, which shows another change in terrain and might suggest a more sandy surface. These changes have been subtle, but nonetheless significant.

I'll predict slow and cautious travel to Erebus, and from Erebus to Victoria.

--Bill

It's been moved somewhat:

http://www.greuti.ch/oppy/oppy_hazcam_sol455-458k.gif


P.S. I dont know whether it's a correct English (German = es hat sich etwas bewegt)

I'm pretty sure they have not budged the rover. Those images just show dirt settling a bit. After all, the pancam mast is looking about and I'm sure that transmits -some- vibration down to the wheels.

The wheel hasn't moved or rotated at all as far as I can see.

Some dust settling though as mentioned above

That looks like the auto nav software trying to find a good path.

But I did see other images that made me think that it might have gotten in and out of trouble on its own. We know that happened durning egress from eagle crater. They made a blind drive toward the crater rim, but didn't make it. It slipped and dug in. REALY DEEP. Still in blind drive mode it made a ninty degree turn and drove away.

All the other images were too far back to see what happened.

Keep an eye out on NASA TV's Video File for some great footage of the trials at the sand-box at JPL

It looked in those still images like they had the rover in a 'pit' - but there's video footage of it climbing thru and out of a fake dune - it was also just a mobility test rover - no cameras on it (to compensate for the mars-earth weight difference I presume)

It would SEEM - based on the testing - that the easiest way 'out' is 'thru' not 'back'. Steve was right though - a LOT of wheel turning without much happening, then a very small turn in place to use the excavated dirt as a 'ramp' to drive on - and it climbed very well

Doug

If we do get to the point where nothing seems to work, the "rocking" approach should be considered. A simplified rocking would not require feedback and it does not require "free" or unpowered wheel rotation. The key concept of rocking is that a forward motion followed by an immediate reverse motion (assuming you want to go backwards) can result in a gravity assist from coming down a small ramp. This provides reverse momentum through the point where otherwise the wheels would have been stationary. Thus you get the rearward force that is applied through wheel traction, plus the gravitational momentum, at the point where otherwise you would only have wheel traction.

It is possible that one rocking motion would be sufficient. But repetitive rocking takes a hole shaped like the wheel and elongates it, building ramps on both the forward and reverse ends.

Special timing and feedback is used to optimize the rocking motion so that no digging occurs at the endpoints of the motion, but a little analysis and test can probably give enough information to pre-program this to simply minimize digging at the ends. You could even do each rocking cycle separated by a day to see how the ramp-building is coming along.

Scott

It is interesting to note that, according to this site http://www.ast.cam.ac.uk/~ipswich/Miscella...flight_news.htm , "On several occasions, Lunokhod 2 sank in loose rock up to the hubs of its wheels."

Scott

The immediate problem I see with that though is that rocking requires that the motors have some speed behind them. These wheels don't seem like they are meant for anything closely resembling speed.

It was really funny to see the speed up part of this movie when the rover drove into the dune
http://www.exn.ca/news/video/exn2005/05/09...09-sandtrap.asx

Imagine if that were the normal pace

Could repeated forward and backwards movements also further compress the current wheel pit, setting the rover deeper into the soil?

Dependent on the material of course, and I suppose that if it were really fluffy (which this stuff may be), you could pound it down even more -- but I think the problem we saw with initially getting stuck had to do with digging more than compressing. And the compressing has limits. Hmmmm...

Depending on the soils, some digging at the ends of the hole (on the created ramps) may be advantageous as long as it results in some linear motion of the rover, and doesn't dig much deeper than the vertical progress that has already been made.

Even with low speed motion, as long as you can get some motion forward, followed by some motion back, you should be able to continually elongate the hole.

Somebody ought to do some research and write a paper on rocking motions to get out of loose materials with different particle sizes etc. Fun stuff. Seems like this "art" known only to hard core 4x4 drivers could be reduced to something resembling a science.

Scott

The wheels have been straightened

http://qt.exploratorium.edu/mars/opportuni...DJP1214L0M1.JPG

http://qt.exploratorium.edu/mars/opportuni...DJP1314L0M1.JPG

Pedal to the metal

seriously know, what is the next step? move forward a little bit?

Look at these two Pancam pictures

http://qt.exploratorium.edu/mars/opportuni...DIP2131R1M1.JPG

http://qt.exploratorium.edu/mars/opportuni...DJP2131R1M1.JPG

The last one is after the front wheel has been straightened.

To me the bottom of that trough looks pretty firm

The sides of the trough also look surprising firm (and vertical) in the second pic, too, as if they had been sliced with a knife, with no sign of sand or dust falling back in that seemed to be suggested in the first pic.

(Actually, when I look back at pics like this rear hazcam one:

http://marsrovers.nasa.gov/gallery/all/1/r...DIP1314R0M1.JPG

from the look of that wheel on the right there hasn't been much sign of infilling there either.)

Would that be the action of the wheel compacting the sides which did that or a characteristic of the dune material itself?

Well - Thursday was touted as a possible day of action - and so that would tie in with yesterdays straight-wheel move - I guess we'll find out what the first step is over the next 24-48 hrs.

I've got a crate of mars bars that says people at Mark Carey's forum, the Habitable Zone and elsewhere WILL - despite the repeated warnings from Steve - brand the exit attempt a complete failure after one sol

Doug

Of course. But isn't it possible for Spirit to help out Oppy? The drive attempt will be a failure because the wheels are stuck in either mud, fossils or eyesockets of skulls. But most likely it was all a trap designed by intelligent martians. It's quite obvious when you look at the geometric layout of the dunes. Look:

But, nonetheless, we'll all keep our tentacles crossed...

--Bill

Hey! you've got to show the whole picture, don't retouch it like those NASA guys usually do...

http://img.photobucket.com/albums/v352/ust...conspiracyb.jpg

A heads up for y'all. I just returned from a talk by Dr. Albert Haldemann, a Deputy Project Scientist on the MER team. Most of what he presented would be old news for this group, but he did mention that their sandbox testing for Oppy consistently shows that a straight reverse out of the dune is their best option . They think it might involve a lot of wheel spinning to back all the way out, and they will take it in stages.

After yesterday's wheel straightening they may well wait for the weekend to pass before they send up commands for a "couple of meters" of wheel turning. They don't expect to get nearly that much actual movement. It took about 60 meters of wheel spinning to dig them in, he said, and it could take that much and more to get out.

The slow approach is not all caution, according to Heldemann. They are curious about the change in soil characteristics, if there has been one, and they want to see how this soil behaves as they churn it up a bit. In response to a question he mentioned that a current "best guess" as to why they went deep into this dune is a combination of locally larger dunes (actually "drifts", he corrected), the configuration of this particular drift... getting aligned with the rover, and the gentle rise in the terrain. Apparently they are driving slightly uphill.

He didn't seem the least concerned that they wouldn't get out, and saw this as a good learning experience while the really cool geology was happening over at Gusev.

Oh...and apparently the JPL people like the "blue sunset" picture as much as we do.

BTW: He mentioned that they are looking for a few more folks for the science team. "Get your proposals in!" is how he put it.

Damn, I can't wait the whole weekend before Oppy makes her next move. Was really excited to see new pictures at exploratorium, but it was day old pictures

Why wait until after the weekend? Haven't they done all the testing they need to do? Why not send the command now and spend the weekend looking at the result??

He may well have been making a small joke about the wait. It would make a lot of sense to do as you suggested and I would not be surprised to see some eveidence of real wheel spinning before the weekend. But ... he did say it. We'll see if he was serious or not. He was definitely serious about the direction, which is straight back, and the pace, which will be bit-by-bit. I also got the sense that one reason it takes so long is that the team is smaller and not running 24/7 these days.

...and that bugs me. Here we have two robots currently operating on Mars, both having survived the most extraordinary flights and landings. Both rovers are going to seize up and die for some reason in the future (could be tomorrow, via flakey solder joints, for all we know) and the team isn't funded/sized enough to run them for 24 hours a day while they're still healthy.

Surely to maximise the near-billion dollars spent on this project, Spirit and Opportunity should be worked all the time? The (Earth-based) costs at this point can't begin to match the development, launch and preceding operational costs incurred so far.

Grr!

Andy

Yeah, but they have to maximize the support for future landers. They are doing great with current funds and i dont see what good would increased budget do. As we have seen, the rovers lifetime is probably limited to how much miles it racks up, not how much time it spends idling. Atleast in 2 years span that is. Probably biggest time critical component is the battery, it will wear out from the constant natural charcing/recharging cycles at some point. It would be interesting if there was public access to the rovers health stats.

http://marsrovers.jpl.nasa.gov/mission/sta...tml#opportunity

OPPORTUNITY UPDATE: Opportunity Mission Manager Report for Sols 456-463 - sol 456-463, May 13, 2005

Opportunity has started moving its wheels again after a couple weeks of holding still while taking some amazing images. While waiting for the rover team to finish tests for planning the best strategy for driving out of a sand trap, Opportunity has been busy taking a comprehensive color panorama of the area. On sol 461 (May 11), Opportunity straightened its wheels. After checking data and images confirming the success of that move, the team planned commands for beginning to rotate the wheels on sol 463 (May 13). The rover is healthy and ready to go.

Sol 461:
Today Opportunity was allowed to position its wheels to their egress direction. The plan is for the rover to leave the sand trap with an arc, moving forward and slightly to the left. The wheels were placed in that position. Before this sol's steering move, the wheels were in position from a turn in place that was Opportunity's last attempted move on sol 446.

From the same report:

Sol 463 (May 13):
After confirming the new position of the wheels, the team proceeded with plans for Opportunity to rotate its wheels about two and a half times on sol 463. Results from that move will be evaluated before rotating them some more.

So they have send the command, just waiting to see the result

There have been very few Lithium-ion batteries operated in space, so they could very well be a life-limiting feature. Although as yet, I have seen no information to indicate any degredation thus far in the mission.

For those technologists interested in the Lithium-Ion batteries produced for the rovers, here are some links. The comparison between Sojourner and MER was especially interesting.

http://www.findarticles.com/p/articles/mi_...212/ai_n9152840

http://www.electrochem.org/publications/in...-Pages22-23.pdf