First 2009 MSL Landing Site Workshop Options

[Stephen]

Abstracts are up on the Ames site:

http://marsoweb.nas.nasa.gov/landingsites/...op/program.html

Phil

Timothy Parker's paper on Friday June 2 is listed on the program as "Mapping the proposed MSL landing sites" but the title the paper itself bears is "Unravelling Terby's Turbulent Past" and seems to be more concerned with advocating the merits Terby Crater than with mapping. Is this an error?

======
Stephen

[RNeuhaus]

True, but a computer powerful enough to run that software at a respectable speed does have weight. The MER's only use a 20MHz processor.
A more powerful computer may be slightly larger, and it will require more power.

A good artificial intelligence for hazardous avoidance, needs a powerful CPU like a Pentium V or above in order to get a reasonable response time after some seconds. But, the MSL microprocessor might use ones IBM PowerPC RAD 750 (MER, MGS, ODY, Messenger, Spitzer, Pathfinder used RAD6000, Developed by IBM, similar to early members of PowerPC. he computer has a maximum clock rate of 33 MHz and a processing speed of about 35 MIPS. Its performance were inferior to Pentium IIs. A typical RTOS running on NASA's RAD6000 installations is VxWorks. Reported to have a unit cost somewhere between US$200,000 and US$300,000)

The RAD6000's successor is the RAD750 processor, based on Motorola/Freescale's PowerPC 750, and is used in NASA's latest Mars probe, the Mars Reconnaissance Orbiter and also probably to MSL. It is equivalent to IBM PowerPC 750 (32-bit). The CPU has 10.4 million CMOS transistors, nearly ten times more than its predecessor. It has a core clock of 133–166 MHz and can process at up to 300 MIPS, or greater with an extended L2 cache. The RAD750 system has a price tag in excess of US$200,000. The CPU power of RAD 750 is about to Pentium IIIs that is still very slow comparing to the present Pentium V runing its clock above than 3.0 Ghz.

Hence, MSL would have a high AI for hazardous avoidance maneuver as Richard would like. Then, the MSL will still rely much the drive command from the rover driver team. However, now, with the help of MRO with high resolution of surface pictures, will permit MSL choose smarter paths toward the desired target.

References, see RAD 6000 and RAD 750.

Rodolfo

[Phil Stooke]

Yes, I saw that thing with Tim's paper too and I think it is an error. I think it might be changed closer to the time of the workshop.

Phil

[mcaplinger]

...we are indeed back up to 15 to 22.5 km total drive distance.

Like I said back on page 3 of this thread -- http://www.unmannedspaceflight.com/index.p...indpost&p=51558 -- they are now saying that the mission total is at least 20 km.

[Guest_BruceMoomaw_*]

Yeah, the main news in that item was the specific reasons given why JPL finally rejected trying to develop improved autonomy for MSL's driving, as opposed to MER's.

[Stephen]

software has no weight!

That is not entirely true. The electronic ones and zeroes software is composed of may have no weight, but the hardware they reside in in order to do their job sure does, whether that be a CPU, RAM, flash memory, a hard drive, or whatever!

Furthermore, the more sophisticated a piece of software gets the more of those sorts of resources it will probably require in order to run efficiently (or at all). To give one example, a rover which needs to find its own way round will probably have some kind of map of its local surroundings sitting somewhere which tells it where it is, where it needs to go, where obstacles (and potential onstacles) are and what kind they are, what path it is currently following, what other path(s) it might be able to follow, and so forth.

The more autonomy you want to give a rover the more sophisticated, detailed, and extensive that map will have to be. Do you want it to be able to retrace its steps from potentially hazardous ground to safer ground? Then you're talking about a more detailed and sophisticated map which not only shows where the rover is now and where it hopes to go but also records where it has been and how it got from there to where it is now, the conditions (eg obstacles) it encountered along the way, etc.

Such a map is obviously going to be a larger and more complex than the one required by less capable software. It is also likely to grow as the rover covers more ground. (Which turn raises a whole heap of other issues I won't bother with here.) That in turn may require more RAM to hold the more sophisticated map, more space to hold the more sophisticated software which runs the more sophisticated map, and more computing power to run the software which runs the map.

The end result may well be an incredibly sophisticated and accomplished software package, but it may or may not be able to run on your existing rover. To accommodate it may require you to add more RAM, a faster CPU or multiple CPUs. Such changes in turn may require hardware changes of other sorts. For example, the more powerful CPU needed to run the more sophisticated software may also require a different motherboard; or it might produce more heat than the old processor, which in turn may require more sophisticated ways of dissipitating that heat. Or it may draw more power, which in turn may impact on the rover's power system.

In short, software may well have no weight as you say, but that does not mean it is weight-less.

======
Stephen

[centsworth_II]

...they went back to the MER level of driving and target-approach autonomy... a 3-sol cycle to approach any particular sampling target at one of its detailed study locations.

As I recall, Steve Squyres' retrospective wish list for the MERs includes six-wheel steering that would avoid repeated, time consuming back and forth jogs to reposition in front of a target or obstacle. Any chance of the MSL having six-wheel steering? Or is the thought of having six steering actuators subject to failure too scary?

[Cugel]

As I recall, Steve Squyres' retrospective wish list for the MERs includes six-wheel steering that would avoid repeated, time consuming back and forth jogs to reposition in front of a target or obstacle. Any chance of the MSL having six-wheel steering? Or is the thought of having six steering actuators subject to failure too scary?

Instead of adding two steering actuators it seems to make more sense to me to simply remove the two non-steering middle wheels. In other words, make it a single rocker-bogie suspension instead of a double. Make the (4) wheels a bit bigger and you end up with what Steve wanted at less weight and better equipped for crossing ripples. If you place the wheels far apart, it doesn't cost you much stability either.

[Guest_BruceMoomaw_*]

No. They tested different designs right up the wazoo (both in computer simulations and in real situations), and the 6-wheel bogie consistently came out the best overall. This is one long-awaited space engineering decision -- the best design for a rover -- that now seems absolutely firm.

[mcaplinger]

No. They tested different designs right up the wazoo (both in computer simulations and in real situations), and the 6-wheel bogie consistently came out the best overall. This is one long-awaited space engineering decision -- the best design for a rover -- that now seems absolutely firm.

For a particular rock size and frequency, maybe. For the actual terrain at the MER landing sites, six wheels is serious overkill for rocks and the system is clearly quite challenged on dunes and drifts.

[Guest_BruceMoomaw_*]

If you have room enough in your aeroshell to include much bigger wheels, a 4-wheel system MIGHT be better. I'll have to review my documents on this, if I can find them (and the time).

In any case, there is some consideration being given to switching over to big inflatable wheels on Mars rovers, which could change everything. Make them big enough and you can simply roll directly over all small obstacles, thereby tremendously simplifying your navigation needs and accelerating your daily progress. The question is whether they can be made durable enough, given all the trouble with the Pathfinder and MER airbags.

[lyford]

Careful, Bruce, you're making me nostalgic....

[mcaplinger]

If you have room enough in your aeroshell to include much bigger wheels, a 4-wheel system MIGHT be better.

Most places on Mars, it's quite easy to just drive around what few rocks there are. A modestly larger wheel might help, but inflatable wheels are hardly necessary. A four-wheel system would have no trouble at all at either MER site.

[RNeuhaus]

Most places on Mars, it's quite easy to just drive around what few rocks there are. A modestly larger wheel might help, but inflatable wheels are hardly necessary. A four-wheel system would have no trouble at all at either MER site.

Absolutely yes. A inflatable wheel must not be of a gas (better) but by mechanical expansion force soon after exploding a control pyro. The wheel must be totally of metal base. A bigger wheel, the rover will advance easier over any cracks, rug surface. Needs electrical motors more power than the smaller ones but the ridding is smoother and makes less effort to motors to overcome any small stones and sand due to its greater buoyance on the ripples or soft ground sand.

But, I think the MSL team has already sticked its design with 6 wheels. Isn't ?

Rodolfo

[MaxSt]

Most places on Mars, it's quite easy to just drive around what few rocks there are. A modestly larger wheel might help, but inflatable wheels are hardly necessary. A four-wheel system would have no trouble at all at either MER site.

Pathfinder's site was quite rocky...