1.4.1 Technology development
To succeed, MSL required several new technologies for landing and surface operations.15 Guided entry required alteration of existing ballistic-entry vehicle designs into a lifting body that could fly through the Martian atmosphere. The steerable spacecraft also needed a new guidance and control system. For sky crane, they needed more powerful descent engines that could support the massive weight of the rover and descent stage, as well as a special reel for the rope that would lower the rover slowly and gently to the surface.
For surface systems, the main challenges were longevity and sample handling. In particular, the motors (also known as actuators) for all the rover’s moving parts were required to survive testing 20 times longer than those developed for the Mars Exploration Rovers. The MSL mission decided to develop a new type of actuator that relied upon a dry rather than liquid lubricant. Liquid-lubricated motors must usually be heated on Mars in order for their lubricant to function, particularly at night. Dry-lubricated motors wouldn’t need heaters, so would be less complex and would save the mission a considerable amount of power and complexity. Heating costs both time and energy, and scheduling time enough for heating increases the burden of tactical planning.
For sample handling, they needed to develop the arm (by now, the design included only one) and a deck-mounted rock corer, rock crusher, and sample portioning and delivery system. There was also the problem of scaling up the Mars Exploration Rover mobility system – the wheels, rocker and bogie arms, and the differential that connected them – by a factor of more than two, without scaling up their mass by a factor of eight. Because the mobility system was also their landing gear, the engineers prototyped it early, developing a physical model that they could test across a wide range of landing scenarios.
The unusual entry, descent, and landing system proposal attracted the attention and concern of NASA officials and review panels. In early 2005, NASA Administrator Mike Griffin, speaking in the context of financial pressures on NASA in general, floated the idea of delaying MSL to 2011.16 Griffin personally visited JPL in June 2005. JPL director Charles Elachi reportedly lobbied Griffin hard not to delay the mission. Buoyed by their confidence in the wake of the successful landings of both Mars Exploration Rovers, the JPL team assured Griffin they would be ready for 2009.
1.4.2 Shifting design, early 2006
One particular decision made early on to reduce costs would wind up haunting the rest of the mission. They decided to give the spacecraft only one main computer, located in the rover, instead of having distinct boxes for cruise, landing, and surface operations. The shift toward fewer electronics boxes was enabled by improvements in commercial electronics technology.17 High-density field-programmable gate arrays (FPGAs), in particular, allowed the rover to have less hardware, but it made the spacecraft more dependent on software development. Software could, in theory, be developed by multiple teams in parallel. But with only one main computer, all that software had to be tested in series; adding more engineers wouldn’t make the test program go faster.
Late in 2005 NASA and JPL management decided to double the rover’s electronic brains, giving it a complete backup system in case any of its critical electronic components failed. There was space for a second set of computers inside the rover, but it added mass, significantly set back the avionics development schedule, and added complexity to the test program. There were now two computers to test, but both were inside the same piece of hardware, and through cross-strapping, both had interfaces with many of the same pieces of equipment, increasing the number and variety of tests that were needed exponentially.
On March 10, 2006, Mars Reconnaissance Orbiter successfully entered orbit. The first new NASA orbiter since 2001 Mars Odyssey brought many capabilities to Mars that would be essential to the success of MSL. Its science payload, which would begin regular operations in October, included a camera named HiRISE (for High-Resolution Imaging Science Experiment) that could map the surface in unprecedented detail. Perhaps more importantly for MSL, Mars Reconnaissance Orbiter carried an upgraded radio for communicating with surface missions and a 3-meter dish for high-rate communications with Earth. That dish gave Mars Reconnaissance Orbiter the ability to relay much more data to Earth than Odyssey did for the Mars Exploration Rovers (Figure 1.4). The successful arrival of Mars Reconnaissance Orbiter meant that the MSL mission could get rid of the enormous radio dish and related power requirements that had been part of the earliest design concepts.
Figure 1.4. Comparison of Mars Reconnaissance Orbiter with the two older NASA Mars orbiters. A 2010 photo of the MSL rover has been added in for scale. NASA/JPL-Caltech/Emily Lakdawalla.
By mid-2006, the rover concept was taking a shape that looks much more recognizable (Figure 1.5).18 Instead of two arms, there was just one, along with the sample preparation hardware bolted to the deck. The mast and arm were configured to support the instruments that had been selected in 2004. The high-gain antenna had shrunk. There was only one, rather than two, MMRTGs. (Because the environmental review process described in section 1.2.3 was not officially complete by this time, the publicly released version of the artist’s concept shown in Figure 1.5 had no MMRTG.)
Figure 1.5. Concept art of the MSL rover prepared for the Preliminary Design Review. From Vasavada ( 2006 ).
The EDL concept had matured more than the rover design (Figure 1.6). The active terrain hazard avoidance that had been scoped in the initial concept was gone, replaced with a much longer powered-descent phase. The final concept of the flight system – a Dagwood sandwich of a spacecraft consisting of cruise stage, aeroshell, descent stage, rover, and heat shield – changed very little after this time (Figure 1.7). This lander would require a larger, flatter landing ellipse than had been promised for the Mars Smart Lander concept.
Figure 1.6. EDL timeline (top) and model of the rover (bottom) prepared for the Preliminary Design Review. Compare to Figure 1.3 .
Figure 1.7. Components of the MSL flight system. NASA/JPL-Caltech.
JPL presented these concepts at the MSL Preliminary Design Review in June 2006. Manning “thought that the team was presenting designs that were rough, much less developed than they should have been at this stage.” Still, the review board voted to give MSL a passing grade, and NASA started writing checks. The biggest one was the first: NASA immediately announced a contract with Lockheed Martin for an Atlas V rocket for the fall 2009 launch, at a fixed price of $194.7 million.
1.4.3 First real cost estimate
Until the Preliminary Design Review, there had not been a solid cost estimate for MSL, because there had not yet been a design to price. The NASA advisory panel that had initially identified MSL as a mission worth pursuing had scoped it out as a medium-class mission, costing under $650 million. Internal NASA bookkeeping accounted it at $865 million in November 2003, a number that did not include the radioisotope power source, the launch vehicle, or the cost of the focused technology program needed to bring some key landing technologies to maturity. The Preliminary Design Review revealed that, despite years of effort to find ways to keep the mission’s cost down, MSL was going to be much more expensive than outsiders had anticipated. Including development costs, new technology, launch vehicle, operational costs, and reserves, project manager Richard Cook estimated a total cost of $1634 million.
1.4.4 Where to send the mission?
At the same time as the Preliminary Design Review, the landing site selection process began. On May 31, 2006, the Landing Site Selection Committee invited the world’s Mars scientists to the first in a series of community workshops. Anyone could propose landing sites, and explain how they would address the mission’s science objectives. To prepare for the first landing site selection meeting, deputy project scientist Ashwin Vasavada translated the science objectives into more specific terms that would help guide the choice of landing site (Box 1.5).
Box 1.5. Curiosity science objectives.
To assess the biological potential of at
least one target environment by determining the nature and inventory of organic carbon compounds, searching for the chemical building blocks of life, and identifying features that may record the actions of biologically relevant processes.
To characterize the geology of the landing region at all appropriate spatial scales by investigating the chemical, isotopic, and mineralogical composition of surface and near-surface materials, and interpreting the processes that have formed rocks and soils.
To investigate planetary processes of relevance to past habitability (including the role of water) by assessing the long timescale evolution of the atmosphere and determining the present state, distribution, and cycling of water and carbon dioxide.
To characterize the broad spectrum of surface radiation, including ultraviolet light, galactic cosmic radiation, solar proton events, and secondary neutrons.
The rover’s precision landing was aimed at an unprecedentedly broad swath of Mars. Previous landers were limited by their solar power systems to regions near the equator. But MSL, with its nuclear power source, could access latitudes as far as 60° away from the equator, including regions where there is modern-day ground ice within reach of a scoop or wheel scuff. The mission also planned for the ability to reach elevations up to 2500 meters above the Martian mean, opening up Mars’ southern highlands to exploration for the first time.19
MSL’s ellipse was smaller than that of previous missions, so landings could be squeezed into tighter spaces (Figure 1.8). At the beginning of the landing site selection process, MSL’s landing ellipse was taken to be 25 kilometers in the down-track direction, and 20 in the cross-track direction. Four main factors affected MSL’s landing precision: Navigational uncertainty. The navigation of MSL and other deep-spacecraft is highly precise, but there are limits. The spacecraft could miss the target by as much as 2 or 3 kilometers both down-track and cross-track.
Attitude knowledge. The mechanical alignment of MSL’s gyroscope relative to the star scanner on the cruise stage might be imperfect, potentially introducing error during guided entry of as much as 4 to 6 kilometers both down-track and cross-track.
Atmospheric and aerodynamic variability. The final 75 to 100 kilometers of MSL’s descent would be flown without guidance, in order to maximize altitude. Variability in the atmosphere, and resulting variability in the aerodynamics of the aeroshell, can cause the spacecraft to miss by 5 to 7 kilometers in the down-track direction.
Winds. Once MSL’s parachute opens, guidance is no longer possible, and winds can divert the spacecraft from its intended path, resulting in as many as 1 to 2 kilometers of positional error.20
Figure 1.8. Comparison of historic Mars landing ellipse dimensions with Gale crater. Gale is about 154 kilometers in diameter. The base image is Viking orbiter data. Map by Emily Lakdawalla.
More than 100 scientists attended the workshop. Scientists participated not only out of a desire to contribute to the landing site selection, but also because proposed locations would be short-listed for early scientific observations by Mars Reconnaissance Orbiter.
The outcome of the workshop was a list of 33 potential landing sites, of which 11 were voted to be top-rank possibilities. All of the favored 11 were below Martian mean elevation, and all were within 30° of the equator.21 The fact that no high-elevation landing sites were among the top-ranked locations presented an opportunity of some relief to the engineers developing the entry, descent, and landing systems. In July 2006, NASA’s Mars Exploration Program made this relief official, setting a new requirement for the MSL landing to lie below 1000 meters’ elevation.
Shortly after the workshop, on November 2, 2006, NASA lost contact with Mars Global Surveyor, and the mission was declared over on November 22. The remaining orbiters at Mars – 2001 Mars Odyssey, Mars Express, and Mars Reconnaissance Orbiter – got to work performing detailed observations of the proposed sites. The committee scheduled a second workshop for October 2007, after a year of Mars Reconnaissance Orbiter’s primary science mission.
1.4.5 Plans for planetary protection
We go to Mars in part because we are interested in searching for past or present life there, so it’s crucially important that we avoid forward-contaminating the planet with Earth microbes. Although Mars is more clement than the rest of the planets beyond Earth, it is not a particularly kind environment to Earth microbes. It’s cold, the atmospheric pressure is very low, it’s incredibly arid, and it’s bathed in ionizing radiation. But Earth life is tenacious. There are some places on Mars warm enough for microbial activity. There are Earth microbes that can survive very low pressures. Some scientists argue that Mars may possess limited present-day liquid water (though likely very salty); and a mere 1 millimeter of Martian soil is enough to shield microbes from damaging ionizing radiation, so life could theoretically be hiding in briny aquifers buried beneath the surface.22
Preventing forward contamination (and also preventing the backward contamination of Earth with alien microbes) is the role of NASA’s Office of Planetary Protection. One of the central concepts in planetary protection is that of “special regions.” A special region is defined by the International Council for Science’s Committee on Space Research as “a region within which terrestrial organisms are likely to propagate” or “a region which is interpreted to have a high potential for the existence of extant Martian life forms.”23 In practice, there are no places on Mars that have yet been identified as having high potential for extant life, so it’s the first definition that applies to planetary protection.
Because MSL is not a life detection mission, it was not planned to target a special region. However, a catastrophe on landing day could actually create a “spacecraft-induced special region.” If MSL crash landed at a place with modern near-surface ground ice, the hot plutonium power source would melt that ice to water and keep it nice and warm. Any microbes that hitched a ride could potentially propagate in such an oasis.
Because MSL had the potential to create such a spacecraft-induced special region, in August 2005, NASA’s Planetary Protection office classified the mission as “Category IVc”. This categorization required either complete sterilization of the spacecraft, or a landing site restriction to regions where water ice is no shallower than 1 meter from the surface, plus sterilization of the parts of the rover that were expected to penetrate below the surface: the wheels and drill bits.24 The 1-meter number was based on estimates of how deeply large hardware fragments could be buried upon impact.
The price tag for system-level sterilization was expected to be between $60 and $170 million – more than the project could afford.25 The project elected to go with the cheaper option, accepting the restriction of avoiding near-surface ground ice, which Odyssey mostly mapped poleward of 45° north and south latitude. The project would sterilize the drill bits and wheels, and closely evaluate landing ellipses late in the landing site selection process to ensure that there were no hints of near-surface ground ice.26 In the end, 89% of the spacecraft’s surface area, and 61% of its volume, were subjected to heat sterilization, making it the cleanest NASA spacecraft launched since the Viking landers.27
1.5 THE COST OF COMPLEXITY (2007–2008)
It didn’t take long for the development effort to run into trouble. The main problem was schedule. Everything seemed to take longer than it should, but the 2009 launch date was fixed and immovable, and complexity cropped up everywhere. Engineers took shortcuts, moving quickly to final designs without time to test early ones, risking that problems would crop up later on, which they did.28 Avionics development continued to be slowed by its complexity. A suite of sensors called the Mars Science Laboratory Entry Descent and Landing Instrument (MEDLI) – essentially another science instrument – was added to the heat shield late in 2006; the project would benefit future Mars landing efforts by supplementing models of entry with actual data, but created another interface to incorporate. The sample acquisition and handling mechanism design was completely scrapped and design effort restarted in 2
007. On top of all of this, the mission suffered two particularly huge setbacks in 2007, involving the design of the motors and the heat shield. Solving these problems required massive redesigns, imposed major costs, and resulted in huge schedule delays.
1.5.1 Sample handling restart
Mars Smart Lander had been initially scoped with two arms, one for coring and one for sample processing. At the time of the Preliminary Design Review, Mars Science Laboratory had had a single arm with a driller/corer and scoop, and a deck-mounted piece of sample preparation and handling hardware to crush, sieve, and portion the sampled rock. While intact cores would provide insight into near-surface layering or weathering, neither of the two analytical laboratory instruments, SAM or CheMin, needed intact cores. In fact, techniques available to crush the cores for the instruments had difficulty in achieving the fine particle sizes the instruments required.
The Design and Engineering of Curiosity Page 4