Status Report

NASA MEPAG: Analysis of arguments for and against removing the sample cache hardware from MSL

By SpaceRef Editor
November 21, 2008
Filed under , , ,
NASA MEPAG: Analysis of arguments for and against removing the sample cache hardware from MSL

To: Dr. Michael Meyer, Lead Scientist, Mars Exploration Program, NASA HQ
From: Scott McLennan, SUNY Stony Brook (chair)
Date: October 9, 2008
Subject: Analysis of arguments for and against removing the sample cache hardware from MSL

The primary benefit of the sample cache on MSL is the potential increase in the likelihood of Mars Sample Return. The primary quantifiable risk of retaining the cache is in mission operations. The immediate concern is the impact of the cache on implementation of the sample observation tray and possible enhancements to the sample delivery system that may be needed for the MSL analytical instruments to meet their measurement goals.


1. At the request of NASA HQ, a group of Mars scientists, drawn from the community to represent divergent opinions, debated the pros and cons of retaining the MSL cache; their deliberations are documented in this letter report.

2. The group agreed on the following:

  • If the cache is determined to be a hindrance to MSL achieving its target launch date, it should be removed.
  • If flown, the cache must not be allowed to adversely influence MSL mission operations, including traverse planning and end of mission scenario planning. Mission operations are subject to political influences, and the presence of a cache may place strong pressures to adjust MSL operations to further MSR at the cost of optimal scientific return from MSL.

3. The group had a range of opinion on the essential risk-benefit issues:

  • The ultimate benefit of the cache is not knowable in advance because it will depend on how it is used, the kind and condition of the samples, how far and where the rover goes, and the accessibility of the cache to a future MSR recovery operation, none of which will be known until after the fact. There are known liabilities that are likely to reduce the value of the cached sample: samples will be acquired only by scoop and would consist principally of cm-size regolith pebbles, and they will not be segregated (and thus will be mixed), and are exposed to the atmosphere (and sunlight) while in the cache. Thus, the committee talked about potential benefit. The committee had a wide range of opinion about this cache value, ranging from significant potential to either enable a future MSR or to contribute significantly to its design, to negligible potential benefit.
  • The MSL cache has potential negative consequences in two major areas, neither of which are easily quantified: 1). Attaching the cache and testing its access will take time and energy from an already tight schedule for finishing the observation tray and sample delivery system. The question is whether this impact is likely to be significant enough that the cache should be deleted now, and 2). Future impact on MSL’s operations, including use of time during the prime mission, and end-of-mission scenario planning.
  • The primary customer for the cache is a future MSR mission. It is generally accepted that the cached samples would be less than ideal for MSR but there are differences of opinion in the expected degree of interest by MSR. Scientists directly involved in the ND-SAG report for Mars Sample Return were concerned about the low diversity of cached samples, the loss of sample integrity and the potential for contamination in an open-basket cache (see Astrobiology 8(3), 489-535, 2008).

4. A wide range of positive and negative reasons for retaining the cache were expressed, reflecting many different prisms of experience. The discussion was unfettered and two diverging lines of thought emerged. 1) Any cache on Mars will be a positive step towards sample return and the cache box should be retained unless it is demonstrated that the MSL objectives will be compromised; and 2) The value of this cache is likely to be low enough that it does not justify the possibility that MSL capabilities or operations will be compromised by its installation, and it should be removed.


In September, 2008, NASA asked for a community-based analysis of the pros and cons of retaining the MSL cache. This analysis was a follow-up to a letter from the MSL project scientist and MSL PIs (dated Aug. 11, 2008), and the MEPAG letter from Jack Mustard dated Aug. 15, 2008. The requested analysis was conducted by a committee consisting of: Scott McLennan (chair), Jan Amend, Dave Blake, John Karcz, Mike Malin, Dave Des Marais, David Mittlefehldt, Jack Mustard, Clive Neal, Dimitri Papanastassiou, Lisa Pratt, Chip Shearer, and Dawn Sumner. Chris McKay and John Grotzinger presented the benefits and risks for retaining the sample cache. (Michael Meyer and David Beaty participated in an ex-officio capacity). This letter report summarizes their deliberations.

Arguments to retain the cache

  1. The MSL cache would represent the first concrete specific steps towards MSR other than viewgraphs. MSR in turn has been widely recognized as a very high priority by the NRC, MEPAG, and NASA. Despite the importance of MSR, and despite the fact that it has been consistently recommended for several decades, NASA has made little progress other than planning studies. Interest in MSR picked up with the announcement by HQ of MSR as the prime goal of the next decade, and the addition of the cache to MSL. The idea was that a small inconvenience to MSL is a small price to pay for a concrete step to MSR
  2. Acquiring samples and filling the cache will be an important demonstration of regolith/soil sample properties under martian conditions using martian material. It is critical to determine how surface materials respond to collection and transfer hardware.
  3. The cached samples may be useful scientifically. MSL will provide unprecedented measurements of martian surface samples and unparalleled roving capabilities. These attributes are likely to exceed those of a future MSR mission. The existence of the cache creates an option for a future MSR mission that may or may not be exercised.
  4. The cache hardware has already been delivered. Deleting it at this point has no meaningful direct savings in either money or mass. If the cache has a neutral impact on MSL’s short-term ability to make it to the launch pad in time, or on MSL’s long-term operations planning, the cache should be retained.

Arguments to remove the cache

  1. The MSL mission is a discovery-based mission that will assess the habitability of the landing site region. A concern is that the sample cache may affect end of mission planning, or perhaps even dictate an early end of the mission. All of the landing sites still under consideration for MSL have a relatively flat, safe landing area, along with surrounding terrain that is more complex geographically and geologically (and which is almost certainly not landable by MSR). It is important that MSL have the freedom to pursue its discoveries wherever they lead. If MSL were commanded to return to its safe landing area at the end of its mission it could be a significant blow to its ultimate scientific return. As several members of the panel testified, although advice of this sort from the Mars community is very effective in the long run, decisions from HQ/projects can have immediate consequences. A case in point is that the cache was added to MSL after CDR. There was no committee meeting and no involvement with the Mars community until after the fact.
  2. Caching a sample will require 2-5 sols, and could take longer for samples with issues such as too many chips or pebbles in the scoop. Assuming that at least 10 samples are collected as planned, this may represents 50 sols for MSL. As one instrument PI said: every sol spent feeding the cache is a sol not used for making other kinds of analyses that address formal MSL mission objectives. On MER, we have learned that there is always a choice of what to do – someone wins and someone loses in using the sol’s opportunity. So it is certain that sample caching will affect operations at least to the level of the time spent on the activity.
  3. The argument has been made that we would learn how to acquire, manipulate, and store samples with this opportunity on MSL. However, much of what we could learn about the challenges associated with handling rocks and fines will already be gained by providing samples to MSL’s analytical instruments, SAM and CHEMIN. Also, evaluations of MSR scenarios by the community have proposed very different sample acquisition and storage designs than the current cache and have emphasized the importance of sample integrity (e.g., see MEPAG ND-SAG report, Astrobiology 8(3), 489-535, 2008). In addition, the design of MSR’s sampling arm will most certainly differ from SA/SPaH. It was further argued that MSR sampling strategies and hardware might be understood with far less expense through a well controlled test bed and Earth-based field studies.
  4. The unexpected difficulties that Phoenix has had with sample handling have created great concern on the MSL team. This has caused the project to consider adding the CHIMRA anti-fouling tools. It was pointed out that given other constraints on space usage on the front of the rover, including the observation tray and the spare drill bits, both of which are of high importance to the scientific functionality of the rover, the effort to find a place for CHIMRA is non-trivial. The currently planned location for the cache is the optimal location. Trying to find alternate space is difficult because the turret at the end of the arm is very large (~30kg), and it will need to interact with the various functional elements (cache, pipe cleaner, funnel, etc) without knocking into the RMI (mast), REM’s boom, and other instruments. There will be some looseness in the arm joints so that they won’t always actuate the same way–this positional uncertainty will complicate the above concerns. As part of the system engineering of the front of the rover, there are lots of complicated interactions that need to be sorted out, all of which will require complete validation in order to confirm that they will work. Completing the system engineering and V&V work for the SA/SPaH system will be a major challenge for MSL. It is not possible to say that the problem cannot be solved within stated cost and time constraints without removing the cache. However, it is clear that doing so would significantly simplify this challenge.
  5. There is no reasonable scientific scenario in which MSR would bring back this sample cache. As noted by the MEPAG ND-SAG committee, the MSL cache will inevitably have low scientific quality in several key respects. One specific scientific limitation is that the value of the material in the cache will be significantly reduced unless we can link a cached rock to one analyzed on the martian surface, and MSL’s design will have trouble with this. In addition, physically rendezvousing with the MSL rover may require significant mobility by the MSR rover, which would increase the cost and mass of the latter. Plus, the existence of this cache may ultimately drive the program to go get a low science value cache. If so, it could negatively affect the value of MSR. If MSL finds something compelling for sample return, the design of a sample acquisition and packing capability of the MSR spacecraft would be substantially superior to the MSL caching capability.

Committee Discussion

Throughout the discussion, one consensus opinion of the committee was that if the presence of the cache negatively impacts the ability for MSL to achieve its launch target date, impedes its surface operations, puts at risk any of its primary goals, or impedes planning for the end-of mission, it should be removed. However, the major sticking point beyond this general view was how to evaluate what constituted significant impacts and how these could be quantified or objectively assessed. Proponents of removal of the cache argued that it had significant potential to put at serious risk both MSL development and mission surface operations that outweighed any potential benefit from samples that might be obtained. On the other hand, proponents for retention of the cache argued that there was no way to quantify this risk at the present and accordingly the cache should be retained at least until such risk was understood in some quantitative way.

The discussion of whether or not the MSL cache should be retained mostly focused on three issues, described in greater detail below (listed roughly in the order they were first raised during the discussion).

1. Impact on MSL Surface Operations

The potential impacts on MSL mission surface operations are summarized above and were re-iterated by several committee members during the discussion. Experience from MER suggests that the cache will affect surface operations in some material way. Although the Mars science community could recommend that this not happen, certain key decisions will be made elsewhere. For example, the community had little impact on the decision to include the cache in the first place and, in general, community involvement mostly influences long term programmatic planning rather than issues related to mission operations. Michael Meyer was asked if any guarantee could be given that the cache would not affect operations and his answer was that such a guarantee could not be given.

Cache operations are expected to consume 20-50 sols of time that would otherwise be spent working toward the L1 science objectives. Subtracting some non-commandable sols due to Earth-Mars phasing (but making the generous assumption that every other sol is productive), this is likely to represent 5-10% of the primary mission. While the value of each MSL sample analysis vs. cache sample is impossible to know, the time impact against the MSL’s nominal science operations is an objective fact. It was suggested that caching of samples should not occur until the prime mission is over. This would alleviate the idea that time would be spent caching rather than analyzing. However, if the cache is a part of the mission, this does not seem plausible. If the cache is on the rover, its use would be persistently advocated.

2. Cache Sample Quality

Questions were raised about the scientific value of the samples that might be stored in the cache, and whether or not they would be attractive to a future MSR mission. The concern is that the cache samples would not be of sufficient quality to address our scientific goals related to aqueous history and habitability. These concerns were recently raised in both the Steele et al. (2008) report and the MEPAG ND-SAG (2008) report, and they are summarized in Mustard’s letter to NASA. Concerns centered on issues of sample degradation, contamination between samples and between samples and spacecraft, unrepresentative nature of cm-size surface fragments, planetary protection, and the impact on MSR sampling priorities and strategies. On the other hand, proponents of the cache pointed out that the cache could increase sample diversity (the opposite view was also expressed) and may provide new insights and findings that are unanticipated, as was the case with lunar samples.

It was argued that a possible value of the cache, independent of whether or not it was ever collected, is that changes in sample characteristics could be monitored over time under Mars’ conditions. This could have both scientific value and could inform us about MSR sampling strategies. This led to a discussion of whether the cache could be monitored during the mission, and by instruments adequate to the task of monitoring any change. There are two issues: 1) Only the outer samples of the cache, which would consist mostly of those added at the end, could be examined over time, and 2) The only instrument that might be brought to bear is MAHLI. In addition, there were mixed views about the scientific value of such observations.

3. Threat of the Cache to MSL Launch Date

The impact of cache removal on the launch schedule was discussed at some length by the committee. This is not a clean question–the cache is a part of the SA/SPaH system, and although the SA/SPaH engineering team is facing some significant challenges, it is not possible to determine the degree to which these challenges are attributable to the cache; there are too many interconnected parts. Many components of the SA/SPaH system are absolutely essential to MSL’s ability to achieve its scientific objectives. It was suggested that perhaps a better way to think about this is that if the SA/SPaH engineering team is under serious schedule pressure, what is the risk that MSL will end up without all of the tools needed to complete its scientific objectives? Removing the cache would simplify the engineering trade space that the SA/SPaH team needs to reconcile, and this weighed heavily on the minds of the advocates for removing the cache. The advocates for retaining the cache argued that we do not know that deleting the cache will make a meaningful difference to the chances of SA/SPaH success.

In follow-up to this concern, Richard Cook, the MSL project manager, was contacted and invited to describe his perception of the engineering implications involved in the cache question and achieving the launch date. His response (Oct. 8, 2008): “MSL is engaged in an effort to improve our ‘tolerance/ to material handling problems. The highest priority aspect of this effort is to add additional tools that can help mitigate the risk of clogging the sample processing equipment. The specific designs of these additional tools are still in work, but one major concern is likely to be the physical accommodation of these tools. The size and complexity of the Sample Arm turret makes it very difficult to access rover-mounted tools (particularly if they are to be used to physically interact with portions of the sample processing hardware). One of the key features of the current cache location is that it is the prime real estate where the turret can be readily positioned (which is why the cache was placed there to begin with). The SA/SPaH team has taken a brief look at whether other sites exist where these additional tools could be placed. Although the answer is still premature, it appears that there are no other spots which would be accessible without significant additional effort. A burden will be placed on the SA/SPaH team to accommodate these additional tools, and not having access to the cache location will make that burden somewhere between moderate and major.”

SpaceRef staff editor.