The Design and Engineering of Curiosity

by Emily Lakdawalla

  Figure 3.11. View from the Kimberley, a mosaic of many left Mastcam images taken on sol 590. In the foreground are layered rocks, the lowest of the three distinct units exposed at the Kimberley. Two more units make up the lower and upper slopes of Mount Remarkable, the mound at middle right. Curiosity drilled near the toe of that mound at Windjana on sol 621. In the distance on the left are the lower foothills of Mount Sharp, Curiosity’s eventual destination. In the distance on the right is the rim of Gale crater. NASA/JPL-Caltech/MSSS.

  The pause at the Kimberley allowed the wheel damage Tiger Team to complete their work. See section 4.​6.​4 for details on the wheel investigation and results. Wheel imaging at every drive proved that damage was progressing only slowly and at a rate predicted from Earth experiments. So sol 636 was the last time that the engineers sequenced single sets of wheel images before every drive; after that, they continued doing full wheel imaging approximately every 500 meters.

  Instrument teams began preparing for extended-mission operations. CheMin, for instance, tested whether they could re-use sample cells. On sol 640 the rover delivered a Windjana sample to a CheMin cell that had previously held Cumberland material. They detected no cross-contamination of the Windjana sample by Cumberland and cleared future deliveries of samples to previously used cells.

  On May 30, 2014 (corresponding to sol 645) the team selected a new future traverse that diverted the rover to the south around a large region of potentially wheel-damaging caprock called the Zabriskie plateau. The new route had the advantage of leading the rover to rocks that represented the base of Mount Sharp earlier than originally planned, before crossing over the dune field. They would need to cross a short stretch of the Zabriskie plateau in order to reach those rocks. The rover climbed onto the plateau on sol 691.

  They started using surge sols again in order to make the most of unrestricted drive sols. They tested a new “sidewalk mode” of MARDI imaging on sol 651 (see section 7.​3.​2) and used it for science purposes on the drive onto the plateau. As the rover approached the dune field, it encountered more sand ripples and some valleys filled with rippled sand. The rover bogged down in sand twice, once at Sourdough on sol 672 and again at Hidden Valley on sol 711. The team backed out of Hidden Valley and modified the path slightly to avoid valleys containing obvious sand ripples, sticking to places where the sand seemed to be a thin coating over rock. On the way out of Hidden Valley, they noticed a bit of Mount-Sharp-related rock that appeared suitable for drilling, but a drill attempt at Bonanza King on sol 733 resulted in the fracturing of the rock, halting drilling. The team elected to abort the sample attempt and drive onward to a better-looking outcrop.

  3.5.4 Mission to Mount Sharp

  Curiosity arrived at basal Mount Sharp rocks at a site called Pahrump Hills on sol 751, just after the first extended mission began. The Zabriskie plateau had ended, and Curiosity left the rock units of the Bradbury plains behind. The bright-colored Pahrump Hills outcrop contained material that the mission had referred to as the “paintbrush unit” when mapped from orbit, but it was renamed the Murray formation as Curiosity approached. It consisted of very finely laminated mudstone. Across a distance of 150 meters, the outcrop rose 15 meters of elevation, a convenient vertical slice through hundreds of rock layers. Curiosity drilled into the lowest-elevation spot on the outcrop, at Confidence Hills, on sol 759.

  The rover proceeded to walk the Pahrump Hills outcrop three times. On the first circuit (sols 780 through 799) the focus was remote sensing. On the second circuit (sols 800 through 862) there was more contact science work with APXS and MAHLI to characterize the rock (Figure 3.12). Finally, on the third trip, Curiosity drilled at two locations, Mojave near the base of the outcrop on sol 882, and Telegraph Peak near the top on sol 908.

  Figure 3.12. Left Mastcam panorama of the “work volume” in front of the rover after a drive to an outcrop named Chinle on sol 826, during the second circuit of Pahrump Hills. These kinds of mosaics are used to plan contact science with MAHLI and APXS. Curiosity surveyed the outcrop to study trends in sedimentology and composition up the outcrop. NASA/JPL-Caltech/MSSS.

  Unfortunately, as the rover prepared for the second circuit on sol 801, the autofocus laser on the ChemCam instrument failed. By sequencing many observations at slightly different focal depths the ChemCam team could continue to gather science data, but less efficiently, and large arrays of shot points were no longer possible. With prodigious effort the team developed a new autofocus capability using the ChemCam imager, but not until sol 983, so autofocus was not available for the entire Pahrump Hills campaign (see section 9.​2.​3).

  Another persistent problem began at the Telegraph Peak site: the drill experienced a soft short in the percussion mechanism on sol 911 that has recurred intermittently on a number of occasions (see section 5.​3.​4.​2).

  Curiosity departed Pahrump Hills on sol 949 to drive around the northern edge of the Bagnold dune field through sand-floored valleys. It proceeded across the Murray formation, among ridges of a new capping rock, initially called the “washboard unit” and later named the Stimson formation. The scenery was especially fine during this traverse, because the Stimson erodes into steep buttes.

  The ChemCam team recovered autofocus capability as the rover was preparing to travel from primarily Murray valley bottoms to primarily Stimson higher plains, at a site called Logan Pass. When Logan Pass proved too sandy for safe travel, the rover turned around and headed to a new location, Marias Pass, where the newly capable ChemCam serendipitously found some silica-rich rock. After standing down for the mission’s second solar conjunction from sols 1005 through 1026, the rover drilled into this silica-rich rock at Buckskin on sol 1060.

  Curiosity turned south, aiming directly toward the dunes. The path took it repeatedly across the boundary between the Stimson and Murray formations. The team observed interesting cracks with bright alteration haloes around them all over the Stimson unit. They drilled into both unaltered and altered Stimson unit at Big Sky (unaltered) and Greenhorn (altered) on sols 1119 and 1137.

  On sol 1174, the rover arrived at the northern edge of the Bagnold dune field. After some initial reconnaissance, the rover scooped samples at Gobabeb beginning sol 1224. Unfortunately, the sample processing activities were cut short by an anomaly on sol 1231. One of the four motors in the CHIMRA sample handling mechanism, the primary thwack actuator, stalled. Investigation proved that it still worked, but out of caution the engineers modified sample processing activities to rely less on the affected motor (see section 5.​4.​6.​3 for more detail).

  Curiosity left the first dune site behind on sol 1248, now traveling west to skirt around the extreme northern edge of the dunes. On sol 1281 the rover again ascended a steep boundary between the Murray and Stimson formations to reach a highland called the Naukluft plateau, where the wind had eroded the sandstone into fantastic shapes. Since the Naukluft plateau would be the last time Curiosity drove on the Stimson unit, the team decided to drill again into an alteration halo (at Lubango, on sol 1321) and into unaltered Stimson (at Okoruso, on sol 1332) (Figure 3.13). On sol 1353, the rover descended from the plateau, placing wheels on the Murray formation again. The team drilled at Oudam, the lowest-elevation site of a long future traverse across the Murray formation.

  Figure 3.13. Curiosity MAHLI self-portrait at the Okoruso drill site. In the foreground is Okoruso. In the middle ground, just below the REMS boom, a bright spot marks the location of the Lubango drill site. Lubango was in an altered halo, Okoruso in unaltered Stimson rock. NASA/JPL-Caltech/MSSS.

  On sol 1369, they turned south to finally cross the dunes among the Murray buttes. The rover would remain on Murray formation rocks for many kilometers, with well-exposed bedrock everywhere, rapidly ascending in elevation. It gave the science team the opportunity to systematically read the layers of the rock to see how the environment changed over time. To do that, they changed their approach to selecting drill sites. The team began to drill every time t
he rover gained 25 meters of elevation. Three such regular drill intervals followed. Marimba, on sol 1422, was just north of the buttes; Quela, on sol 1464, was just to their south; and Sebina, on sol 1495, was another 25 meters above. Conveniently, the new regular-interval style of traverse roughly coincided with the start of the second extended mission on September 1, 2016, corresponding to sol 1448.

  Unfortunately, the regular intervals of drilling came to a halt on sol 1536, when Curiosity attempted to drill at a site named Precipice. A problem had developed in the drill that was unrelated to the intermittent drill percussion problem that first appeared on sol 911. Now the drill feed mechanism would not advance reliably. As of sol 1800, the rover has not drilled since. It has collected one more sample, a scoop of sand at a site named Ogunquit Beach at the southern edge of the Bagnold dunes on sol 1651, but did not deliver the sample to CheMin prior to sol 1800 because of concern about the drill.

  The rover continued to advance southward to approach the interesting-chemistry rocks first seen from orbit, doing science on the way with its other instruments while engineers investigated the problem with the drill. As of this writing, the rover had climbed onto Vera Rubin Ridge, formerly known as Hematite Ridge. Engineers began testing on Mars a new mode of drilling without using the drill feed on sol 1848.

  As of sol 1800, Curiosity had attempted sampling in 20 locations, of which 17 resulted in the successful acquisition of sample and subsequent delivery to SAM and CheMin. Sample processing related to these 16 drill sites and 2 of the sand scooping sites is summarized in Figure 3.14 and Table 3.3.

  Even if the mission were to end tomorrow, scientists would be working on interpreting Curiosity’s data for decades. Of course, the mission hopes for much longer survival than that.

  Figure 3.14. Seventeen Curiosity sample sites on Mars. Each is a MAHLI focus stack taken from a standoff distance of 5 centimeters, except for Rocknest (a photo of 150-micrometer sample on the observation tray from a standoff distance of 5 centimeters) and Sebina (a zoom in on a single image from a 25-centimeter standoff distance). As of sol 1800, there was no close-up photo of material sampled from Ogunquit Beach. NASA/JPL-Caltech/MSSS.

  Table 3.3. Summary of Curiosity drill and scoop campaign activities through sol 1800. As of sol 1800, Ogunquit Beach sample remained inside CHIMRA. For explanations of the activities listed in the leftmost column, read Chapter 5 .

  Activity

  Rocknest

  John Klein

  Cumberland

  Windjana

  Bonanza King

  Confidence Hills

  Mojave

  Telegraph Peak

  Buckskin

  Big Sky

  Greenhorn

  Gobabeb

  Lubango

  Okoruso

  Oudam

  Marimba

  Quela

  Sebina

  Precipice

  Ogunquit Beach

  Mini-Drill(*=unsuccessful)

  Scoops: 61, 66, 69, 74, 93

  174, 176, 180

  –

  615

  726*

  756

  867*, 881

  –

  1059

  1116

  1134

  Scoops: 1224, 1228, 1231

  –

  –

  –

  1420

  –

  –

  1536*

  Scoop: 1651

  Full drill

  182

  279

  621

  –

  759

  882

  908

  1060

  1119

  1137

  1320

  1332

  1361

  1422

  1464

  1495

  –

  O-tray dropoff

  70, 76, 77, 78, 95

  –

  284, 289

  –

  –

  –

  –

  –

  –

  –

  –

  –

  –

  –

  –

  –

  –

  –

  –

  –

  Pre- and post-sieve sample volume inspection

  79

  193, 194

  279, 464

  623

  –

  762

  884

  922

  1061

  1121

  1139

  1224, 1228

  1323

  1334

  1362

  1425

  1465, 1466

  1495, 1496

  –

  1651

  Drill wall survey with MAHLI

  –

  270

  279, 283, 292

  628

  –

  768

  883

  910

  1064

  1123

  1142

  –

  1324

  1337

  1364

  1426

  1466

  1496

  –

  –

  Pre-sieve (coarse fraction) dump

  64, 67, 73, 81, 128

  229

  486

  704

  –

  765

  889

  930

  1064

  1123

  1142

  1226, 1228, 1251

  1324

  1337

  1364

  1426

  1466

  1496

  –

  –

  Post-sieve (fine fraction) dump

  781

  894

  954

  1089

  1132

  1198

  1226, 1228, 1251

  1327

  1358

  1418

  1457

  1491

  1533

  –

  –

  MAHLI self-portrait

  84, 85 (stereo)

  177, 270

  –

  613, 627

  –

  –

  868, 882, 884

  –

  1065

  1126

  –

  1228, 1241

  –

  1338

  –

  –

  1463, 1466

  –

  –

  –

  CheMin delivery

  71, 77, 94

  195

  282

  623, 640

  –

  765

  884

  922

  1061

  1121

  1139

  1226

  1323

  1334

  1362

  1425

  1466

  1496

  –

  –

  SAM delivery

  93, 96, 98, 116

  196, 199, 224, 227

  281, 286, 290, 353, 367, 381, 414, 463, 464

  624, 653, 694

  –

  773

  887, 891, 892

  928, 954

  1075

  1129

  1147, 1178

  1224, 1230

  –

  –

  1382

  1443, 1456

  1484

  –

  –

  1651

  REFERENCES

  Allison M (1997) Accurate analytic representations of solar time and seasons on Mars with applications to the Pathfinder/Surveyor missions. Geophys. Res. Lett. 24(16):1967–1970, DOI: 10.1029/97GL01950

  Bass D, Wales D, and Shalin V (2005) Choosing Mars time: Analysis of the Mars Exploration Rover experience. Paper presented at IEEE Aerospace Conference, 5–12 March 2005, Bi
g Sky, MT, USA, DOI: 10.1109/AERO.2005.1559722

  Cantor B, James P, and Calvin W (2010) MARCI and MOC observations of the atmosphere and surface cap in the north polar region of Mars. Icarus 208:61–81, DOI: 10.1016/j.icarus.2010.01.032

  Chattopadhyay D et al (2014) The Mars Science Laboratory supratactical process. Paper presented at SpaceOps 2014 Conference, 5–9 May 2014, Pasadena, USA

  JPL (2013) Rover Team Working to Diagnose Electrical Issue http://​mars.​nasa.​gov/​msl/​news/​whatsnew/​index.​cfm?​FuseAction=​ShowNews&​NewsID=​1559 Status report dated 20 Nov 2013, accessed 15 Aug 2016

  Reichardt T (2015) The man who named the Martian day. http://​www.​airspacemag.​com/​daily-planet/​man-who-named-martian-day-180957350/​. Accessed 2 Mar 2016

  Footnotes

  1Bass et al (2005)

  2Reichardt (2015)

  3Allison (1997)

  4Allison (1997)

  5Local Mean Solar Time is defined for a fixed longitude on the surface. For Curiosity, that longitude was defined before landing to be 137.42°E. Curiosity’s mission time does not shift with Curiosity’s changing longitude. For every 246 meters that Curiosity drives west of the initial landing position, the Sun rises 1 second later than it does at the longitude of the landing site. Curiosity actually landed at 137.441635°E, which meant that there was about a 4-second difference between Curiosity mission time and the Local Mean Solar Time – not enough of a difference to make it worth it to adjust the software, especially because this tiny difference is swamped by the variations in sun rise/set times caused by the difference between True and Mean Solar Time.

  6Chattopadhyay et al (2014)

  7This section is based upon my years of reporting for planetary.​org on the ongoing adventures of the Curiosity mission. That reporting is based upon mission images, press releases and team blog entries on the JPL and United States Geologic Survey websites, roughly monthly interviews of Ashwin Vasavada, and occasional conversations with numerous other team members