[X] My Assistant

[karolp]

I am having difficulty finding decent comparison data, especially for the planned ExoMars's Pasteur rover as well as the planned MSL and ancient L's. I keep wondering: if we put them all at Oppy's current position, which one would be the first to reach Victoria's rim? I guess one should take into consideration not only the average speed (which I have never seen presented in a nice table for comparison) but also maneuverability and modes of navigation...

[djellison]

Well - I would put Exomars on a similar footage to that of MER in terms of daily range and driveability.

MSL, perhaps double the MER ability.

The old Lunokhods would do it in 10 minutes probably - as they were driven my humans sat on the earth....which is obviously not possible with Mars.

Doug

[Bob Shaw]

I am having difficulty finding decent comparison data, especially for the planned ExoMars's Pasteur rover as well as the planned MSL and ancient L's. I keep wondering: if we put them all at Oppy's current position, which one would be the first to reach Victoria's rim? I guess one should take into consideration not only the average speed (which I have never seen presented in a nice table for comparison) but also maneuverability and modes of navigation...

Karolp:

Well, Lunokhod 2 was the fastest rover to date - Pasteur, I have no idea about but being solar powered it must have similar limitations to the MERs, and MSL doesn't seem to be much faster in terms of top speed than the MERS, just able to drive longer. So, for a sprint, Lunokhod; but distance, MSL, and Pasteur, probably a bit better than MER.

And, in the far distance, there's Sojourner...

Wacky Races already!

Bob Shaw

[JRehling]

This is a fun basis of comparison, but remember that MERs did not except by Opportunity's luck land on top of any target of interest. MSL should head right for the prime target on the first try and will thus have a science return hard to compare with MER.

There is also a speed/risk tradeoff as we've seen with Opportunity. I suppose in principle the fast drives could have led to a mission-ending purgatory. Imagine if instead of a sand dune it had been a tire-wide crevice between two rocky plates: that might have been the end of the party.

I think speeds greatly exceeding any of the aforementioned will be possible with less risk that Opportunity faced as AI for this task develops. Watch the DARPA-sponsored robotic race on Earth -- the drive systems are totally different than anything we'll send to Mars, but software is as light as you want, and the winner last year averaged 31 km/hour. I think we'll see the bottleneck cease to be decision-making in the next 30 or so years, if we're still in the business of launching rovers anywhere then.

[Phil Stooke]

The longest Lunokhod drive was 3130 m on 18 February 1973. I don't know if that was spread over 24 hours or in a shorter work period.

Jrehling mentions MSL landing right on its target, as if a drive to get to it would not be necessary, but actually many proposed sites are GO To sites - safe landing followed by a drive to the interesting stuff. Some of those presentations at the recent workshop required drives of up to 20 km!

Phil (reporting from Southampton Public Library)

[dvandorn]

In re the drive-to concept:

The thing that bothers me is that, as I understand it, MSL's more sophisticated sensors will require dirt and rock samples to be ingested into the sensor, and that there will be a very limited number of processing cycles on these devices. Something like between 12 and 45 separate analyses, and after that, the device is inert and can never produce any further data.

If we had sensors like this on Spirit, what are the odds that your most important devices would be used up and no longer available by the time Spirit reached the Hills, thereby reducing the real interesting science quite a bit? As it is, Spirit's RAT was used up mostly on hard lava chunks out on the plains -- so, in a sense, we did run across this issue on the MER missions.

If you decide to wait to use your instruments until you reach your target location, there is the issue of landing a good, working scientific payload on Mars and willfully refusing to use it for a few months, while you drive to your "interesting" location -- all the while betting that nothing bad will happen to your rover on the way... and also betting that you don't run across an invaluable find en route, that you pass up because you're hoarding your processing cycles for after you reach Location X.

NASA probes to other planets have been designed to acquire enough data to declare "mission success" within the lifetime of the shortest-lived components on the spacecraft. Only after that is accomplished do they plan for delayed science vs. travel and/or engineering evaluations.

Whatever they do with MSL, I cannot imagine that they will design the mission to have to rely on 100+ sols of roving before the main science investigations can begin -- not with all the ways Mars can potentially kill a rover.

-the other Doug

[djellison]

I think you have to be sensible. You land - do science with whatever you're presented with, and then head off for your chosen target reserving the bulk of your expendable resources for that.

They SORT of did that with Spirit...Adirondak, Humphrey, Mazatsal, Route 66, and then no RAT work until Pot-of-Gold ( I think ), and then Clovis.

Doug

[JRehling]

If you decide to wait to use your instruments until you reach your target location, there is the issue of landing a good, working scientific payload on Mars and willfully refusing to use it for a few months, while you drive to your "interesting" location -- all the while betting that nothing bad will happen to your rover on the way...

The issue of waiting til you see what's behind Door Number Two is a common phenomenon, and in cases where simple mathematical models can be made, optimizations have been derived. In the real world, there is an endless cascade of unknown variables, but the general principles of wisdom should be accessible to people smart enough to run a mission like this. I would construct crude models of landscape and value (in imaginary science dollars) of targets and run some simulations and give the planners some simulations of an entire mission sequence, and let them figure out which strategies lead to joy and which to regret. There are no guarantees, of course.

While the MERs were fairly exploratory (see the middle letter in the acronym), we could hope that MSL has a known target and can confidently drive past unexpected phenomena without destructive experimentation en route to answering its stated goal-questions. Hope is a key word, of course.

[Phil Stooke]

I'm sure the old Apollo idea of a contingency sample will be applied here. If there are limited analysis opportunities, you still do an early one (maybe on whatever looks most interesting in the first pan) before heading for the target, just to make sure you get something. And occasional samples on the way might be OK if really compelling.

Anyway, a GO TO site might not be selected in the end for this very reason. The presentations recently put up on the Ames site make for very interesting reading.

Phil

[edstrick]

Beyond that... ANY sample will yield new science when run through new instruments, and a perfectly typical sample of mars-dirt (especially if it's the globally transported dust) will be easier to compare with other mission's results than local "weird" stuff.

It also gives a first real-use look at instrument behavior and performance...when they have really important samples after a "go-to" traverse, they'll be ready to do it "right".

[dvandorn]

I will make a point about the Lunakhods... yes, they put in impressive numbers for distance traveled by an unmanned rover.

But they did precious little science while setting those records.

As I recall (and someone please correct me if I'm wrong), the Lunakhods carried no sensors for in-situ analysis of rocks or soils. Not even any "sky science" sensors to detect radiation or magnetic environment details.

The only active scientific experiment I can recall, on Lunakhod 2, was a soil penetrometer. Which gives you more in terms of engineering information -- about bearing strength, cohesiveness, etc., of lunar soil -- than it gives you real scientific data. (Yes, they also had passive laser reflectors, one of which has never been acquired and the other of which is in an unknown location on the lunar surface, making its use somewhat less than useful... in point of fact, the return from the laser reflectors may have been greater had they been attached to stationary landers and not to rovers.)

In other words, the Lunakhods were primarily engineering demonstrations, with almost all scientific return removed to ensure the success of the engineering tasks. (And they were propaganda tools, to show that the USSR was more capable of roving the Moon than Apollo was, though that's a very debatable point.)

So, I'd be a little wary of comparing Lunakhod performances to those of the MERs, or the projected performance of MSL or ExoMars. If all you want to do is demonstrate speed (and, specifically, a speed or range of exploration greater than that of your political rival's system), without giving more than a passing nod to science, it's a little bit easier to accomplish.

Let's hope no one decides they need to prove their bravery or masculinity or any other such nonsense by making a Mars rover that's fast, but not much else...

-the other Doug

[tedstryk]

Actually, the Lunokhods had other instruments, including an X-ray spectrometer, an X-ray telescope, cosmic ray detectors, and energetic particle detectors. The problem is finding original data sets, although it is certainly inferior to Apollo data. The main reason they went so far is that they could be driven by controllers via nearly live video feed, allowing human drivers to steer.

[Guest_DonPMitchell_*]

There are papers by Grant Kocharov about the Lunokhod rock analysis. They both contained x-ray fluorescence spectrometers and returned data about Al, Si, K, Ca, Fe abundances along their routes.

[dvandorn]

Wonderful! I've made several, probably somewhat desultory, internet searches for Lunakhod science returns over the past several years and have found next to nothing (including very little, as is obvious from my comments, about the actual existence of geoscience sensors on either craft)... it's good to know that something is out there somewhere.

Is there anything in those results that fits poorly with Apollo-based lunar models? Or have they even been analyzed as part of the model-creation-and-maintenance process?

-the other Doug

[Guest_DonPMitchell_*]

I've seen his papers cited, and some are on the ADS site. I've read about the work in Yuri Surkov's book on the exploration of terrestrial planets by spacecraft. I recommend that book for information about planetary geology, both American and Soviet missions.