Printable Version of Topic

Posted by: Glevesque Jul 21 2017, 10:48 AM

Hello everyone !

I would like to know more about the mileage grid on the ground. I have difficulty finding precise data from various unofficial sources.

Here is my list:

______________________________________________________________________________
Bonjour a tous !

J'aimerais avoir des précisions sur la grille de kilométrage parcourus au sol. J'ai des difficulter a trouver des données bien précises à différentes sources non officiel.

Voici ma liste :
______________________________________________________________________________

1 km Sol 335, 17 juillet
2 km Sol 363, 14 août
3 km Sol 403, 24 sept
4 km Sol 431, 22 oct
5 km Sol 521, 23 janv
6 km Sol 568, 12 mars
7 km Sol 641, 26 mai
8 km Sol 668, 23 juin
9 km Sol 709, 4 août
10 km Sol 957, 16 avril
11 km Sol 1296
12 km Sol 1439
13 km
14 km Sol 1448
15 km
16 km Sol 1669

Posted by: pospa Jul 21 2017, 11:31 AM

How about this one : http://www.curiositymsl.com/tracking/drivelog.html

Posted by: Glevesque Jul 21 2017, 11:55 AM

Thank you, but I think the site does not take actual distances into account. It includes the general rolling without descriminating the landslides, the enclosures and others. I am referring to the actual mileage on the distance traveled on the Martian surface.

________________________________________________________________________________

Merci, mais je crois que le site ne tient pas compte des distances réel. Il inclut le roullement général sans descriminé les glissements, les enlisements et autres. Je fais référence au kilométrage réel sur les distance parcourus à la surface martienne.
________________________________________________________________________________

Posted by: RoverDriver Jul 21 2017, 11:00 PM

The data on that web site appears to be matching the telemetry pretty well. If VO (visual odometer) is enabled the odometer does take into account slippage (if that's what you meant by landslide). VO is being used all the time nowadays but at the beginning of the mission it was left off or on auto *VO is automatically turned on by the rover under certain conditions). The odometer reports the overall distance traveled by the center of the vehicle which is obviously different from the odometer for each wheel (a turn-in-place counts as zero meters).

Paolo

Posted by: nogal Jul 23 2017, 11:18 PM

Hello Glevesque,

Here is the data that I use for my MSL route map kmz (Google Earth) file. I hope it helps.

--------------------------------------
Salut Glevesque,

Voici les donnees que j'utilize dans mon fichier kmz de la route de MSL (Google Earth). Je souhaite qu'ils te seront utiles.
---------------------------------------
Fernando

 MSL_KML.txt ( 1.75K ) : 793


Km Sol Earth Date Information sources
1 335 16-07-2013 Analyst's Notebook entry for Sol 336
2 365 16-08-2013 Analyst's Notebook entry for Sol 365
3 406 27-09-2013 Analyst's Notebook entry for Sol 404 and 406
4 436 28-10-2013 Analyst's Notebook entry for Sol 438
5 540 11-02-2014 Analyst's Notebook entry for Sol 542
6 574 18-03-2014 Analyst's Notebook entry for Sol 575
7 655 10-06-2014 Analyst's Notebook entry for Sol 655
8 670 24-06-2014 Analyst's Notebook entry for Sol 670
9 735 31-08-2014 Analyst's Notebook entry for Sols 735-751
https://www.nasa.gov/press/2014/september/nasa-s-mars-curiosity-rover-arrives-at-martian-mountain/#.VBRvNWRdVVt
10 957 16-04-2015 Analyst's Notebook entry for this day gives total distance as 10231m,however
NASA/JPL press report below states the 10km mark was passed on this date
https://www.jpl.nasa.gov/news/news.php?feature=4556
11 1094 04-09-2015 Analyst's Notebook entry for Sol 1094
12 1248 09-02-2016 https://astrogeology.usgs.gov/news/astrogeology/sol-1249-twelve-kilometers-and-counting
However Analyst's Notebook entry for sol 1248 mentions "12305 Odometry"
13 1376 19-06-2016 https://mars.nasa.gov/multimedia/images/curiosity-rovers-location-for-sol-1376
14 1448 01-09-2016 https://mars.nasa.gov/multimedia/images/curiosity-rovers-location-for-sol-1448
15 1526 21-11-2016 https://mars.nasa.gov/multimedia/images/curiosity-rovers-location-for-sol-1526
16 1666 13-04-2017 https://mars.nasa.gov/multimedia/images/2017/curiosity-rovers-location-for-sol-1666
17 1754 13-07-2017 https://mars.nasa.gov/msl/mission/mars-rover-curiosity-mission-updates/?mu=sol-1755-getting-ready-to-disappear-behind-the-sun
...................................................................

Posted by: Gerald Jul 24 2017, 04:00 PM

Note the glitch on Sol 967 of unresolved cause, which suggests a drive of 346.56 m. The map is displaced since then.

Posted by: PaulH51 Jul 24 2017, 08:49 PM

Joe used to periodically (manually) adjust the drive distances / path, to bring his map in line with the actual path. It appears to been quite some time since he carried out maintenance on the page as other functions such as 'Map It' that relied on his mapping function can no longer be used. I still use his page for other activities such as the camera pointing info and image sorting.

Posted by: Glevesque Jul 25 2017, 12:24 PM

thank you very much nogal !

Posted by: elakdawalla Jul 25 2017, 05:04 PM

There is also a lot of information available in the PLACES database:
https://pds-imaging.jpl.nasa.gov/data/msl/MSLPLC_1XXX/

Unfortunately, while it includes positions, it does not include drive odometry. You can get some of that from Vasavada et al. papers:
http://onlinelibrary.wiley.com/doi/10.1002/2014JE004622/abstract

Posted by: nogal Jul 30 2017, 05:05 PM

One more thing I'd like to add: The PLACES database - on which most of Curiosity's route map for my GE file is based - includes for each location entry the (x,y,z) distances to the landing location.
Thus it would seem that computing the 3D distance between two successive locations would give that segment's length, and that adding all of the segments would allow computing the total traveled distance.
Here is a table of the sols in which each km was crossed (up to km 15, the latest covered by PDS release 14) and a comparison with the previously posted table. As can be seen, there are some differences, sometimes large, and I don't know why. I'd be indebted if anyone can provide an explanation, by posting it here or by PM. Cheers

... Sol..... Sol
Sol Computed External
1 335 335
2 369 365
3 406 406
4 437 436
5 546 540
6 589 574
7 657 655
8 672 670
9 743 735
10 956 957
11 1080 1094
12 1196 1248
13 1357 1376
14 1438 1448
15 1514 1526

Posted by: RoverDriver Jul 30 2017, 06:35 PM

It depends on what you mean by traveled distance. Places only reports the coordinates of each stopping location and connecting each stop with a straight line would not take into account the swerving and turns we typically do while driving the rovers. This is especially true in rough terrain. The odometer that JPL typically reports does take that into account.

Paolo

Posted by: Glevesque Jul 31 2017, 02:03 PM

J'aimerais également comprendre d'ou vient la différence de calcule sur la distance journalière parcourus entre ses deux sites :

Exemple pour le Sol 406 :

72,94 mètres pour http://www.curiositymsl.com/tracking/drivelog.html

66,8 mètres pour http://www.midnightplanets.com/web/MSL/sol/00406.html (cumulatif)
_______________________________________________________________________


I would also like to understand where the difference in the daily distance traveled between its two sites is calculated:

Example for Sol 406:

72,94 mètres for http://www.curiositymsl.com/tracking/drivelog.html

66,8 mètres for http://www.midnightplanets.com/web/MSL/sol/00406.html (cumulatif)

Posted by: nogal Jul 31 2017, 05:22 PM

Many thanks for your reply, Paolo. Appreciated.
Wouldn't including the twists and turns - which, to me, seems the correct thing to do - make the odometry grow faster and, thus, crossing the km mark earlier? This seems to be the case up to km 10 (comparing the two columns) but not afterwards.
As a curiosity: the average segment size is 0.9024 m with a standard deviation of 0.6396
I'm sticking to the JPL announcements
Fernando

Posted by: RoverDriver Jul 31 2017, 06:19 PM

Yes, it should but in cases where we had a very high value of slip that might have not been accounted for by the on-board software, or because of accumulated errors in position estimate, the rover might think it has moved more than it actually did so periodically Tim Parker (our resident Phil Stooke) re-localizes the vehicle position, and that's what it is poked in PLACES. While MSL has a better definition of rover odometry than MER it is still not uniquely defined. As the saying goes "a man with a watch knows what time it is, a man that has two is never sure!"

Regarding the average segment length, I actually never measured it. This also would require a definition of what "straight line segment" actually means. I typically use the Pavlidis algorithm to define a straight line segment and set the tolerance to 10% (if the width of a set of points is within 10% of the length of the path, I call is a straight segment). If I have some spare time, I will run some numbers.

Paolo

Posted by: Phil Stooke Jul 31 2017, 11:56 PM

"our resident Phil Stooke"

Inside every Phil Stooke is a Tim Parker trying to get out (I have TARDIS-like qualities)

I am studiously avoiding this discussion (oops, until now) because it's so difficult to sort out exactly what these distances mean. I don't think we can definitively answer the simple-sounding question 'how far has the rover moved', not to everybody's satisfaction. Any distance flags I put on my maps are only approximations.

Phil

Posted by: tanjent Aug 1 2017, 01:53 AM

Even if the slippage could be precisely accounted for, how could we settle on a definition of "distance traveled"?
It's like trying to compute the exact length of the coastline of France.
Any path with self-similar irregularities down to an infinitely small scale is not going to have a unique length unless you first specify the radius of measurement.

Ok, maybe a wheeled vehicle has a minimum turning radius, but in three dimensions the terrain will still cause infinitely small bumps that need to be factored in if you want the "exact" distance, and many of these will have occurred in between measurements. It will always be necessary to simplify the true path to a series of segments, and the distance between stopping points seems like the most expedient segment to use for most purposes.

Posted by: RoverDriver Aug 2 2017, 03:57 AM

I was not trying to be pedantic but really it all depends what you are trying to measure. For me it is mostly to measure the wear and tear on the vehicle, so to me discarding all twists and turns is definitely not the way to go. Only rarely we drive in a single straight line, most of the time we have two, three, even five segments with large turn in place (every 90 deg turn in place corresponds to about 5m of wheel motion!). Since we have been trying to reduce wear and tear on the wheels drive paths have been quite tortuous and I would not be surprised to find out that just connecting the end of drive coordinates would underestimate by more than 30%.

Paolo

PS: I finally ran some numbers. Missionwide, the difference between reported odometry (*) and straight line distance between start and end of drive on average is 35.5%.
(*) reported odometer: odometer as reported by the vehicle.

Posted by: Glevesque Aug 2 2017, 01:00 PM

I apologize in advance for having very poorly expressed myself in my previous positions, and I greatly thank all those who participated in this discussion, as well as providing valuable assistance that gave a better understanding of the challenges of Driving the rover on Mars. Indeed, I understand the dilemma of driving the rover facing the premature wear of its wheels. And this is of utmost importance to potentiate a longer mission of Curiosity on Mars. In fact, my question was just to know where I could find reliable data to reproduce on a mileage map traveled according to the trajectory, the answer that my nogal given is greatly satisfactory on this. If I asked this question, it is because the three sources that serve as my reference have different data on daily distances traveled, and I wanted to understand where this difference comes from.

I thank you all.
______________________________________________________________________

Je m'excuse d'avance de m'avoir très mal exprimé dans mes postes précédant, et je remercie grandement tout ceux qui ont participé à cette discussion, tout en y apportant une aide précieuse qui a donnée une meilleurs compréhension sur les défis de la conduite du rover sur Mars. Effectivement, je comprend bien le dilemme sur la conduite du rover faisant face à l’usure prématurée de ses roues. Et cela est d'une très grande importance pour potentialiser une plus longue mission de Curiosity sur Mars. En fait, ma question était juste de savoir ou je pouvais trouver des données fiables pour reproduire sur une carte de kilométrage parcourus en fonction de la trajectoire, la réponse que ma donné nogal est grandement satisfaisante à ce propos. Si j'ai posé cette question, c'est que les trois sources qui me sert de référence ont des données différentes sur les distances journalière parcourus, et je voulais comprendre d’où viens cette différence.

Je vous remercie tous.
______________________________________________________________________

Exemple with Sol 967 :

1- Midnight Planets : http://www.midnightplanets.com/web/MSL/sol/00967.html
2- Curiosity Drive Log : http://www.curiositymsl.com/tracking/drivelog.html
3- Curiosity Rover's Location/JPL-Nasa : https://mars.jpl.nasa.gov/multimedia/images/?ImageID=7148

Posted by: Glevesque Aug 2 2017, 07:15 PM

I would also like your help to trace the daily cumulative distances (per Sol) that were greater than one hundred meters. Here is the list that I was able to come out on this subject:

PS: I did not find the distance record assuming the Sol 976 of 262.41 meters (May 2015)
----------------------------------------------------------------------------------------------------

J'aimerais également votre aide pour retracer les distances cumulatifs journalière (par Sol) qui ont été supérieurs au cent mètres. Voici la liste que j'ai pu resortir a ce sujet :

PS : Je n'ai pas retrouver le supposer record de distance au Sol 976 de 262.41 meters (mai 2015)
-----------------------------------------------------------------------------------------------------

Curiosity Rover's Location (Nasa/JPL)
Analyst's Notebook entry for Sols
---------------------------------------------------
Sol - Dist(m) - Earth Date - Information sources
385 - 128.81 - 06/11/2013 - https://mars.nasa.gov/multimedia/images/?ImageID=5549
419 - 117.84 - 11/10/2013 - https://mars.jpl.nasa.gov/multimedia/images/?ImageID=5649
657 - 112.63 - 12/06/2014 - https://mars.jpl.nasa.gov/multimedia/images/?ImageID=6319
661 - 109.35 - 16/06/2014 - https://mars.jpl.nasa.gov/multimedia/images/?ImageID=6342
662 - 123.88 - 17/06/2014 - https://mars.jpl.nasa.gov/multimedia/images/?ImageID=6344
665 - 121.60 - 20/06/2014 - https://mars.nasa.gov/multimedia/images/?ImageID=6350
668 - 104,85 - 23/06/2014 - https://mars.jpl.nasa.gov/multimedia/images/?ImageID=6366
671 - 113.58 - 26/06/2014 - https://mars.jpl.nasa.gov/multimedia/images/?ImageID=6372

Source : https://an.rsl.wustl.edu/msl/mslbrowser/tab.aspx?t=MP

Posted by: RoverDriver Aug 2 2017, 10:43 PM

Sol 00976 was 74.33 meters.

Here's the list of 100+ meters drives.

00548 100.06 meters
00547 100.26 meters
00340 100.27 meters
00960 102.35 meters
00569 102.95 meters
00454 103.33 meters
00644 104.24 meters
00668 105.24 meters
00670 107.15 meters
00371 110.15 meters
00751 114.32 meters
00671 116.79 meters
00657 120.96 meters
00419 125.79 meters
00662 132.95 meters
00661 137.57 meters
00385 141.49 meters
00665 143.64 meters

The distance record still is held by Oppy with 219 meters on Sol 410.

Paolo

Posted by: Glevesque Aug 4 2017, 04:55 AM

A big thank you for sharing his information.
------------------------------------------------

Un grand merci pour le partage de ses informations.

Posted by: nogal Aug 17 2017, 06:11 PM

Hello again,

The discussion is this thread prompted me to thoroughly review the km and mile signposts in "The Martian Way" kml file, especially at the light of new (to me) information in the articles by http://onlinelibrary.wiley.com/doi/10.1002/2014JE004622/full and http://onlinelibrary.wiley.com/doi/10.1002/2016JE005200/full. which provide sub-meter accuracy for the total traversed distance for sols 0-494 and 535-634, respectively. The former allows very precise location determination of kms 0 through 4 and miles 1 and 2; the latter extends this to kms 5 and 6 and miles 3 and 4.

Additionally, starting with sol 1282, the description on NASA/JPL Location and Traverse maps ("Where is Curiosity") includes the total traversed distance to an accuracy of 10m. In this way good estimates for kms 13 to 17 and miles 8 to 10 can be made.

Km 10 was abundantly mentioned. Thus I "only" needed to "guesstimate" the locations of kms 7, 8, 9,11, and 12, and miles 5, 6, and 7. This proved to take a lot longer than I antecipated, which explains why I have not made any recent updates to Curiosity's GE path. Phil Stooke's maps have several of the guessed kms, which is good corroboration - thank you Phil. Despite doing my very best, errors may exist. I have after all, "guesstimated" several locations. I would be very thankful for any information that could improve the accuracy of those locations.

In short, here is the revised list of kms and sols:
1 335
2 365
3 404
4 436
5 540
6 587
7 655
8 670
9 743
10 957
11 1098
12 1248
13 1376
14 1448
15 1526
16 1666
17 1754

And here is the complete list, with information sources. To use it, dowload it and drop the ".txt" from its name.
 MSL_kms.csv.txt ( 129.33K ) : 645


Fernando

Edit: Something I've learned: in the Analyst's Notebook odometry or total odometry is not equivalent to traversed distance. The former really does exceed the latter.