Paolo (and, Fran, please correct me if I'm wrong), I think he's suggesting compressing the soil under the LF wheel by repeatedly pressing on it & thereby building a little base of compacted sand as new stuff trickles underneath the wheel between downstrokes.

This hearkens back to an earlier question: How compressible is this stuff (either side), anyhow? If it's sufficiently fine-grained, then an idea like this has some merit; flour gets pretty rigid under pressure. Unfortunately, it also appears to be relatively deep.

GIven the max static torque of the wrist and turrent, I'm not sure significant compacting would be possible anywhere you might want it to. You could line up the joints in such a way as to pass it down to the elbow and shoulder in some specific orientations I guess.

All this discussion about these "extreme" uses of the IDD raises another question.

I suppose that IF the IDD is used to e.g. lift the wheels, move objects, drag the rover..., its science instruments will finally become damaged / unusable. So, what's the best compromise? A stationary rover with an usable IDD or an unstuck one with a damaged IDD?

I would use the RAT as the 'blunt trauma' IDD instrument - I THINK that should keep the others REASONABLY safe.

Kind of compromise answer: once we have no more need of te IDD at the current position, we'd better become a rover again.

That is a very sensible comment. It would probably more productive for the moment to concentrate on ideas that do not involve the use of the IDD.

Paolo

Fran and nprev brought up trying to compress the sand beneath the wheels...

If this were in some way advantageous, could compression be achieved by vibrating the rover? I'm refering to the method used on Opportunity to try and shake dust free of the Mini-TES. You could vibrate to compress, then advance a couple centimeters and repeat. Perhaps progress could be made that way.

The idea of compressing the sand makes me uneasy though. You may gain some traction, but you could also sink her further into the sand.

Ah, yes. I meant the front side. -1 spatial sense for me.

I was thinking in something like the Brazil nut effect. Raising a little bit the LF wheel and dropping it many times so sand from the sides is dragged under it and compacted.

[Ignorant, flailing idea time]
Trying to simulate a 4x4 where all wheels are locked (coordinated)...
...would it be possible to vary the current to each wheel such that it is inversely proportional to the amount of slip?

Would that be useful?
(or has this already been done?)

We typically "lock the differential" so to speak since there is one motor per wheel. We can control each wheel independently but that has to be preprogrammed. We cannot control the wheel torque dynamically tho. For that we would need to sequence one command, downlink the telemetry, see the wheel current (proportional to torque) and change the parameters accordingly. That would be slow, but possible.

Paolo

Good and thanks for the pointing. That is a puzzle!
It is very difficult to differentiate the type of soil if it is already buried by ashes, and sands deposition. It seems like that on left track had passed a thin flow of water...

Better still, how NOT to get stuck. On the other hand, I suppose there are situations on Mars that even the best rover design cannot anticipate.

* Imagines some buyer at a large west coast swimming pool supply manufacturer who had the foresight to have Yahoo send him an alert any time the words 'stuck, 'sand', and 'Mars' turned up in the same news article yelling 'Eureka!' *

Starting a 360 degree pan?

01906 14:07:14 p2279.10. 1 0 0 0 0 0 12 pancam_Calypso_pan_col1_1x3_L257R2

This is probably a "duh" question, but still having trouble visualizing the "nearly high-centered aspect", even after looking at the excellent photos & models posted. Is she dug down far enough at the wheels that her belly is nearly resting on the surface near the crest of this little ridge? I've been thinking of a pile of rocks and/or sand underneath, but not sure if that's correct.

There is a little ledge, possibly delimiting the bright toned material. Near the ledge there are some rocks that generate some elevation on top of the ledge. The LF, RR, RL wheels are buried completely. The LM and RM wheels are buried about to the hubs. The RF is on top of the dark-toned material. Since most of the wheels are buried the bottom of the REM is quite close to the surface. Those small rocks near the ledge are the cause of the high centering.

Paolo

Thanks, Paolo; that helped!

I confess that I don't know what the REM is (rear electronics module?); is it part of the WEB? Anyhow, if the REM can take a weight load & rest on the rock underneath, here's yet another silly last-resort idea: Spin the R wheels and dig down on the right side till she bottoms out partially (keeping the left wheels seated, such as they are), then try another spin-in-place to the left using the rock as a pivot point. Hopefully, Spirit would slide off the rock (to the present aft direction) with her nose pointed more or less downslope.

Not a first-choice strategy by any means, and there are two major assumptions involved, but just throwin' it out there.

Let's look on the positive side: we're digging one heck of a trench The ever-ambitious Spirit rover wanted to become the Spirit digger, exploring the hitherto unknown nether-regions of Mars.

You almost got it. ROVER electronics module, but I meant the WEB (warm electronics box).



Yes it can! It is pretty robust and eve in case of puncture there's about 1" of space before hitting anything.

Paolo

Grazie tante, Paolo.
Right now you're keeping a lot of us sane.

The Calypso pan... with reference to, I'm guessing, the Odyssey?

Odysseus (of Ithaca!) spends ten years on a Mediterranean island, 'sorrowing bitterly in the house of
the nymph Calypso, who is keeping him prisoner, and he cannot
reach his home for he has no ships nor sailors to take him over
the sea.' He gets out, but it takes a while.

http://www.gutenberg.org/dirs/etext99/dyssy10.txt

I guess the last time Spirit got stuck ( Jammerbugt between sols 833 - 843 ) it sat deeper in the sand. However, let's not forget it took almost 5 weeks to get Oppy out of Purgatory Dune ( sols 446 - 484 )

Opportunity was the one that got stuck in Jammerbugt.
Actually, not even stuck. By that time they had a technique to get out from the experience learned at Purgatory.
Past events demonstrate that the Rover drivers know what their vehicles are capable of.

Did I hear someone calling?
So...what's the deal? Where do you guys want me to focus my kinetic powers?...

The Spirit case is even more complicated than her sister due to its inoperative right front. This causes twice obstacles: 1) No traction on both sides and 2) Yes drag in two ways but going forward causes less drag than going backward only if the RF is not able to lift a little from surface.

All at all, after long time of long dragging and if Spirit manages to reach on the other side of firmer soil, she will be back on the business.

Here a Panorma from Spirit on Sol 1904:

Click to view attachment

This is probably overly simplistic, but how about turning the wheels VERY slowly for extend periods? It has worked for me after getting my 65 Ford Galaxy 500 stuck (without a shovel) in deep snow, and in sand.

They already do that

Good time for a related amplifying base question: Are the MERs wheels constrained to drive at a constant speed? I assume that the answer is yes.

During each command yes, they ramp up quickly to the steady speed that is required for the motion and just prior to the end they ramp down. Just like any servo motor. But during an arc, say to the left, the left side wheels move slower than the right side, simulating a mechanical differential if you want.

Paolo

Really? That's interesting. I was thinking that they were open-loop constant speed motors for sake of design simplicity, not servos.

So, Beauford's idea is possible, then? You can control wheel rotation speed with precision?

Some wheel movement thisol?

Comparing the image with those from sol 1899 it appears the rear bogie has move up slightly.
sol 1899: http://207.7.139.5/mars/spirit/rear_hazcam...DNP1354L0M1.JPG
sol 1908: http://207.7.139.5/mars/spirit/rear_hazcam...DOP1314L0M1.JPG

Reversing the rear wheels in an attempt to put more weight on the middle wheels?

Does the consistency of the soil change any with temperature? Any chance things would be different in the winter now that we have nice sparkly solar panels?

Simple test for the ISIL once there's a suitable buried testbed rover.... a pull test with a force meter to establish if the sort of force required to shift our stubborn girl is anywhere near the IDD's abilities.

Something was commanded (site/drive numbers have changed) but I can't see any movement.

To return to nprevs suggestion about rocking, this is the preferred way of getting yourself loose when stuck in snow here in Sweden: You gun the car for a moment, then unclutch and let it fall back. However it will move a little bit past the equilibrum position, so you catch it on the return and gun it again and so on. If you time it right the oscillation will gradually build until the car gets loose. I was taught the technique for snow, but it works just as well for desert sand and laterite clay (I've tried both). Surprisingly the technique does not seem to be generally known in other parts of the World.
I don't know if such short bursts of power are practical for MER, and the timing as I said, must be precise. Possibly it could be tried in the test-bed.

Also I would like to comment on the possible use of the IDD for pushing. It might make a difference. I remember a case when two people pushing (I was one of them) was enough to get a ten-ton truck unstuck.

She could also wave the arm and become a Tesla's oscillator.

Too many automatic transmissions here in the states, although I was familiar with the technique with my old econobox that also didn't have power steering. It was the best possible car for detecting slippery road conditions-- both the clutch and the steering wheel let you know if you were losing traction.

I don't think the rovers have a clutch of either variety, though.

I learned to drive on an old standard transmission truck in the mountains of Montana & lived in Alaska for a few years, so have had to rock out of a few slippery situations...

Problem is that doing this old-school is very dependent on real-time feedback to the driver; we don't have that luxury with the MERs. If this is tried, I'm hoping that the major effect is to precess the fore end of Spirit downslope (to the left) & let gravity help with the final extraction.

I did that rocking back-and-forth thing quite often, but I don't know what the clutch needs to be used for? I just had the van in first gear and gave it gas everytime I started on the forward motion. Basically it is like pushing a swing, you give it an extra push everytime you reach the highest point on the backswing. It is also how Robert Langdon gets out of an iffy situation involving a big book cabinet and an armed glass window in the new Angels and Demons movie

That technique really needs timing, so I think it must be very difficult to perform on Mars.

Well, I don't know about timing, but it should be possible to use the IMU's accelerometers to measure any rocking and apply throttle appropriately. You push in the clutch normally on the backwards roll. Unfortunately, I don't think the wheels turn anywhere near fast enough to initiate the rocking to begin with. What's more likely is that she'll have to "swim" out of the sand by sheer brute force, using the treads as paddles to move sand from the from in front to behind the wheels. The problem is that it is a race between getting out, and the wheels sinking below the top of the hubs or high-centering on the bottom. One experiment to try is figuring out which part of the sol gives us the best traction, similar to Nomadd22's idea of waiting until winter. I'll bet that the consistency of the sand changes with different times of the day/night, as the relative humidity increases the sand/dust grains will stick together better. Plus, the area beneath Spirit is probably the coldest place on the Columbia Hills, so it'll collect any moisture around, a la freezer burn. So sitting and thinking for a bit is a good thing.

Yeah, like I said, wasn't really envisioning rocking in the sense of getting out of a snowdrift. Getting the nose turned downslope via precession is about the best we can hope for IMO.

Still, though, it would be interesting (and useful!) if the MERs could execute a complex autonomous subroutine as you described, Hendric. The required response-to-feedback time is pretty short.

This was a short duration command on the LM wheel to diagnose the wheel stall we had on 1899. I haven't looked at the data yet, so I am not sure of the results.

Paolo

Unfortunately, the interval between commands is one second. That limits the response time. In addition, sensors are queried every 1/8 sec. Way too slow for any real-time control of the vehicle. Some serious coding would be needed, tested on the ground, uploaded on the vehicle and verified on Mars before we would be even attempt to use it.

Paolo

Alan: Some wheel movement thisol?

According to their filenames, both of those images indicate that they were taken on Sol 1897. So this was preseumably some of the early moves after getting stuck. The images just came down on the later Sols.

FWIW here's how it looks to a complete outsider in these matters:

Elegant idea (using resonance as an amplifier) + proven effectiveness on Earth + outwith present capability + possibility of developing software to do it + potential benefit to other rovers, present and future + serendipitous availability of 0.4g sandbox = highly worthwhile project.

The rocking technique ( I've used it as well ) involves free-wheeling back to the bottom of the little trench you dig, then throttling back up. MER can't free wheel. Essentially, it would just be driving backwards and forwards in little steps. To me- that sounds like an excellent means of digging the wheels in further, without making any progress.

Hopefully we'll get some better visualisations soon, but I thought I'd play with a Rover model, some flour (for the soft material) and topped with some cinnamon (for contrast) to simulate for me what's going on. I created a ridge structure under the 'soils' to create a spot where loose material is stacked up against some kind of ridge, as suggested by some of the images.
Click to view attachmentClick to view attachment
Click to view attachmentClick to view attachment

I've tried to get the wheels in at roughly the right depth and haven't tried putting rocks underneath.
Just looking at it though, there's not much room between ground and WEB.
The rocks themselves though don't look that big. I guess it's more a matter of how far the RL/R and ML/R wheels have dug in.
What we really need is a look underneath Spirit.

I thought it too but I am still uneasy since this is a like a chess game. By taking the advantage of the left side using the gravity help but after, toward the Braun hill (south), there are small slope that Spirit must overcome that, I think it would be very tough. In order to evaluate the north evacuation way, it is necessary to study about how is the soil surface toward the north.

That is pretty awesome! If only that model had working rocker-bogie system the state of the wheels would have been represented with even more fidelity. If only I could use flour in the testbed but I was told it is a fire hazard. And cinnamon too! The ISIL would smell so good!

Paolo