The Design and Engineering of Curiosity

by Emily Lakdawalla

  Two things limit the lifetime of the active scanning capability of the DAN experiment. The pulsing neutron generator has a warranted lifetime of 10 million pulses, so Curiosity can expect to perform about 1000 typical active observations over its lifetime. It is not likely that Curiosity will hit this limit, though, because DAN’s neutron generator also has a clock time limit of about 3 years after launch. Helium generated by the neutron-generating pulses and the decay of its tritium target eventually ruins the vacuum inside the ion accelerator. Similar hardware on Earth has lasted anywhere from 2 to 6 years, with a median lifetime of 3 years, before failing. If the neutron generator fails, DAN will be less capable of estimating the depth of any subsurface layering that might be present, but it can still be used in passive mode to watch for subsurface variations.

  8.3.3 Using DAN

  DAN operates most of the time in a passive mode, its two detectors counting up neutrons any time the rover’s computer is awake. The counts are typically binned every 20 seconds, producing a nearly continuous record of the number of neutrons hitting the detectors, including along rover traverses.

  Because of the neutron generator’s limited lifetime, the DAN team desired active scans to be performed frequently during rover traverses. Initially, most drives of any length included two to four mid-drive DAN active scans. But the 15-minute length of each DAN observation traded directly against drive distance, so the strategy of frequent DAN observations was abandoned after sol 403. Instead, there is most commonly one DAN active observation per drive sol. The observation may be performed after the end of a drive sol, between drive sols (its most typical location during restricted-sol periods), or before the next sol’s drive. DAN cannot be used in active mode at the same time as some other rover activities because of the neutrons it generates. Examples include ChemCam observations, CheMin analyses, and driving or arm motion.

  There were nearly 500 DAN active experiments performed over the course of the mission up to sol 1417. DAN has operated throughout the mission with no significant gaps in coverage; nearly every rover stop is documented with a DAN active measurement. DAN active measurements have fed back into tactical planning. DAN measurements of abundant thermal neutrons on sol 991, combined with unusual ChemCam measurements of rocks in the same area, led to the drilling of the high-silica target Buckskin below Marias pass on sol 1060. DAN had the opportunity to experiment on silica-rich materials at the Greenhorn and Lubango sites on sols 1144 and 1329.

  8.3.4 Anomalies

  The three-year expected lifetime of DAN’s neutron generator ran out at the end of 2014, but DAN continues to operate normally. In October 2016, it showed the first signs of degradation. To the relief (and surprise) of the DAN team, the instrument returned to nominal operations afterward. Still, the generator can be expected to fail any day. The DAN and rover operations teams have collaborated on a workaround to allow DAN to continue operations as the neutron generator ages, and they monitor its health regularly.

  8.4 REMS: ROVER ENVIRONMENTAL MONITORING STATION

  One of the signal accomplishments of NASA’s Mars program has been the continuous monitoring of Martian weather and climate since the beginning of Mars Global Surveyor’s science mission in 1998. The data set permitted scientists to develop general circulation models for Mars’ atmosphere like those that have been developed for Earth. But moving from global to smaller-scale models depends on surface meteorological data that has been historically scarce. REMS is designed to gather the wind, temperature, moisture, and pressure data necessary to constrain small-scale weather models for Mars. It also studies the ultraviolet radiation that penetrates Mars’ atmosphere, which can have harmful effects on organic molecules.7

  The principal investigator of the REMS experiment is Javier Gómez-Elvira of the Centro de Astrobiología (CSIC-INTA) in Spain. Most of the REMS components were provided by the Centro de Astrobiología, except for the pressure sensor located within the rover body and humidity sensor on Boom 2. They were provided by the Finnish Meteorological Institute, who also contributed the pressure sensors on the Viking and Phoenix landers.

  8.4.1 Introduction: Gale weather

  Gale is in a complicated weather regime, where global, regional, and local weather patterns are all important, and the patterns change on daily and seasonal timescales.8 There are three main global wind patterns. In the “daily thermal tide,” the Sun heats air on half of Mars, making the atmosphere expand and so decreasing its surface pressure, while cooling air on the night side has higher pressure. Air flows from the high to low pressure areas, generating a current that wants to flow perpendicular to the terminator, toward the sunlit side and away from the night side of Mars. In the “hemispheric dichotomy slope flow,” Mars’ global topographic dichotomy also drives winds, which flow upslope (north to south) during the day and downslope (south to north) overnight. Finally, there is a seasonal pattern driven by the heating and cooling of Mars’ poles: air rises at the hot pole and sinks at the cold pole near the solstices, and rises from the equator and sinks at the poles near the equinoxes. All of these effects combine to make complex patterns of winds that change diurnally and seasonally.

  Where is Gale in all of this? It’s close to the equator and has high elevations to the south but also has Elysium Mons to the north. At some times of year (like the southern autumnal equinox at Ls=0), all the different circulations add up to produce practically no predicted wind in the Gale region. By contrast, near southern summer solstice at Ls=270, Gale is predicted to experience strong winds blowing from north to south all day and night.

  However, Curiosity may feel few of these global currents. Gale is a deep hole in the ground, surrounded by high walls and containing a tall mountain in its center, and generates its own weather. During the day, wind blows upslope, toward the crater walls and the mountain peak. At night, winds flow downslope. Curiosity landed near the deepest part of the crater, and over the course of the mission it has moved from a wind regime controlled by the crater walls to one controlled by the mountain.

  Martian calendar information relevant to the REMS experiment is summarized in Table 8.1. For more information on the Martian calendar, see section 3.​2.Table 8.1. Mars seasonal events and correspondence with Earth dates and Curiosity sols.

  Mars year

  Spring equinox (Ls = 0°)

  Summer solstice (Ls = 90°)

  Autumnal equinox (Ls = 180°)

  Winter solstice (Ls = 270°)

  31

  Sep 13 2011

  Mar 30 2012

  Sep 29 2012 / sol 53

  Feb 23 2013/sol 196

  32

  Jul 31 2013 / sol 350

  Feb 15 2014 / sol 543

  Aug 17 2014 / sol 722

  Jan 11 2015/sol 865

  33

  Jun 18 2015 / sol 1018

  Jan 03 2016 / sol 1212

  Jul 04 2016 / sol 1390

  Nov 28 2016/sol 1533

  34

  May 05 2017 / sol 1687

  Nov 20 2017 / sol 2059

  May 22 2018

  Oct 16 2018

  8.4.2 How REMS works

  REMS consists of four units: two booms mounted on the mast, an ultraviolet sensor mounted on the deck, and an electronics box located inside the rover (Figure 8.6). The electronics box contains a pressure sensor. Both Boom 1 and Boom 2 have wind and air temperature sensors. Boom 1 has a ground temperature sensor, and Boom 2 has a humidity sensor. REMS is a relatively autonomous instrument, gathering data 10 times per second for the first 5 minutes of every hour, all the time. The science team also inserts several longer periods (an hour or more) of continuous 10-times-per-second data collection into every sol of activity, rotating the periods around the Martian clock to cover all times of day over a period of several sols. MAHLI is routinely used to monitor the condition of the ultraviolet sensor, and occasionally to view the booms.

  Figure 8.6. REMS components. Top left: Boom 1 as seen by MAHLI (0526MH0003430000201119C00). Top right: Boo
m 2 as seen by MAHLI (1572MH0006620020601062C00). Boards 2 and 3 of the Boom 1 wind sensor (orange) failed upon landing. All three boards of the boom 2 wind sensor (yellow) failed around sol 1500. Bottom left: Mastcam mosaic of the ultraviolet sensor, imaged in the sol 1197 Mastcam self-portrait. The individual labels identify the different photodiodes. Bottom right: REMS instrument control unit located in the body of the rover, which also contains the pressure sensor. NASA/JPL-Caltech/MSSS/Emily Lakdawalla.

  8.4.2.1 REMS booms

  The two booms are located 1.6 meters above the surface, sticking out from the mast, on a part of the mast that does not rotate, so their orientations are fixed. Boom 2 projects directly forward, while Boom 1 is oriented 120° clockwise from it, pointing back and to the right side of the rover (see Figure 7.​1 for their locations). The 120° separation was intended to make sure that one boom would always experience wind that did not lie in the wake of the mast.

  The wind sensors use hot film anemometry. By recording the amount of power it takes to keep tiny titanium resistors at a constant temperature, the science team can figure out how effective the wind is at cooling the resistors, and use that to determine how fast the wind has to be moving past the resistors. Because hot film anemometry depends upon the temperature of the instrument, it’s best to have the resistors separated from the hot rover as much as possible. It’s also best to have the resistors located as far as possible from anything that could obstruct the flow of the wind. The mast does disturb the motion of the wind, so there were two booms to measure the wind blowing at different points with respect to the mast. To keep the resistors disconnected from the boom structure, they were mounted on little pedestals and attached to their electronics boards with extremely thin wires. Unfortunately, two of the three boards on the wind sensor on Boom 1 were damaged during landing, so the wind experiment has never operated as designed.

  The air temperature sensor consists of two thermistors on each boom, one at the middle and one at the tip of the boom. The one at the tip is intended to record air temperature; the one at the middle helps to calibrate out any heating of the thermistor by heat conducted through the boom itself. The electronic circuits that run the two booms, located at the base of each boom, must be kept above –70°C. When a sensor detects that the temperature falls below that, it turns on its heater.

  The ground temperature sensor uses three thermopiles to measure the infrared brightness temperature of the ground. It detects the temperature from an area to the right of the rover. Mastcam “clast survey” images, taken at the end of drives, cover the area of the REMS ground temperature sensor field of view (see section 7.​2.​2.​3). The MMRTG and its 2000 watts of waste heat are an important source of error in the REMS temperature measurements. There was concern that it might heat the air and ground around the rover, including some of the ground within the view of the ground temperature sensor. However, the data show little evidence of the MMRTG’s heat affecting air or ground temperature measurements.9

  The humidity sensor employs a polymer film whose electrical properties change as temperature and humidity change. The polymer film constantly responds to changes in the environment, but humidity can only be read once the sensor receives power. Supplying power to the sensor warms it, so the most accurate humidity measurements are the ones made immediately after it has been powered on.10 Mars has little water in its air, but at night the temperature drops low enough that relative humidity can reach as high as 70%. In places where the ground gets exceptionally cold at night (dusty places with low thermal inertia), it’s possible that the near-ground air gets cold enough for water frost to precipitate in a very thin layer on rock surfaces.11 Curiosity has periodically looked for ground frost in the early morning after winter nights, but has so far not observed it.

  8.4.2.2 REMS Ultraviolet sensor

  The ultraviolet sensor has six photodiodes sensitive to different ultraviolet wavelengths. Five of them (named UVA, UVB, UVC, UVD, and UVE and labeled in Figure 8.6) look at narrower slices of the ultraviolet spectrum, and the sixth (UVABC or UV total dose) looks at the full ultraviolet range (Figure 8.7). By measuring the amount of light falling on the photodiodes and correcting for the solar elevation angle and amount of dust that has accumulated on the diodes, the REMS team can derive the aerosol optical depth, a measure of how much sunlight has been blocked from reaching the ground by particulate matter in the atmosphere. Mastcam photos of the Sun can make the same measurement more precisely than REMS, but the REMS UV sensor measurements are far more frequent (multiple times per day, as opposed to the Mastcam cadence of roughly once per week). The sensors are covered about 10% of the time by shadows (mostly from the rover mast), but they spend most of their daytime unobstructed, and data taken in shadow are easy to remove from the data set.

  Figure 8.7. Spectral response of the REMS UV photodiodes. “UVABC” is also known as “UV total dose.” UVA, UVB, and UVE have provided the highest-quality data. From Smith et al. ( 2016 ).

  Dust is an obvious concern to an upward-pointed light sensor, therefore each of the photodiodes is surrounded by a ring-shaped magnet that prevents Martian dust from falling in its center, working to keep the photodiodes clean. MAHLI takes a photo of the sensor roughly every two months in order to monitor dust deposition on it (see Figure 8.8). Although a lot of dust has accumulated on the magnets over time, the windows over the sensors have remained relatively clean over the course of the mission, and have become cleaner when the rover has paused in windy areas, particularly during the Pahrump Hills investigation from sol 800–900 and while driving through the Murray buttes around sol 1400 and following. Empirically, the REMS team has found the UVA, UVB, and UVE sensors to provide the best estimates of optical depth.12

  Figure 8.8. All MAHLI images of the REMS ultraviolet sensor to sol 1500. More images were taken on sols 1552, 1614, and 1675. NASA/JPL-Caltech/MSSS.

  8.4.2.3 REMS Pressure sensor

  The pressure sensor is located inside the REMS electronics box, itself inside the belly of the rover. Although it is thermally isolated from the environment, the rover warm electronics box has vents that allow its interior to be filled with Martian atmosphere at ambient pressure. The pressure sensor has two transducers, one of which is designed to be more stable, and the other of which is designed to be more responsive to rapid pressure changes. Either provides good measurements, though, so the two transducers provide redundancy in the experiment design. Each has two electrodes, with the distance between the electrodes changing as a result of changes in pressure. That changes the capacitance of the transducer, providing a sensitive measure of pressure changes. The pressure sensors provide better readings after warming up, so measurements toward the end of the 5-minute window of each hour are considered more accurate than those recorded in the beginning.13

  8.4.3 REMS on Mars

  The loss of two of the wind sensor boards on Boom 1 (identified with orange tags in Figure 8.6) during landing was catastrophic to the wind experiment. REMS recorded data from the forward-facing wind sensor on Boom 2 since landing, but because the rover has been facing primarily south throughout the surface mission, and the prevailing wind has come mostly from the north, virtually all of the wind measurements have been contaminated by rover hardware being in the way. The first board on Boom 2 failed on sol 1485, and the other two on sols 1491 and 1504.14

  Apart from the issues with the wind sensor, the REMS instrument package has been performing reliably, sol after sol, recording measurements of Mars’ weather. Figure 8.9 summarizes some of the REMS data for most of the mission, sols 0–1514.

  Figure 8.9. REMS air temperature and pressure minima and maxima for sols 0–1514 give an illustration of the continuity and density of the data set. The air temperature repeats reliably (within a narrow band of variation) year over year. Seasonal pressure variations have been consistent, with a notable secular decrease in pressure with time, caused by Curiosity’s increasing elevation. Seasons (gray text) refer to the southern hemisphere,
where Curiosity landed. Note that sol numbers for each season go a bit into the future, beyond the data set. Graphs courtesy Javier Gómez-Elvira.

  REFERENCES

  Hassler D et al (2012) The Radiation Assessment Detector (RAD) investigation. Space Sci Rev 170:503–558, DOI: 10.1007/s11214-012-9913-1

  Hassler D et al (2013) Mars’ surface radiation environment measured with the Mars Science Laboratory’s Curiosity rover. Science, DOI:10.1126/science.1244797

  Gómez-Elvira J et al (2012) REMS: The environmental sensor suite for the Mars Science Laboratory rover. Space Sci Rev 170:583–640, DOI: 10.1007/s11214-012-9921-1

  IKI Laboratory for Space Gamma Spectroscopy (2011) Russian neutron detector DAN for NASA’s Mars Science Laboratory landing rover. http://​l503.​iki.​rssi.​ru/​DAN-en.​html.​ Accessed 21 May 2014.

  Martínez G et al (2016) Likely frost events at Gale crater: Analysis from MSL/REMS measurements. Icarus 280:93–102, DOI: 10.1016/j.icarus.2015.12.004