Lessons Learned From the Clementine Mission
Clementine operated quite differently from most large space science missions flown during the last two decades, but it is anticipated that many future NASA missions will function along the lines of Clementine. Thus several aspects in the implementation of the Clementine mission are relevant to the lessons that NASA might learn from the successes and failures of this mission. These characteristics are discussed below, and the strengths and weaknesses of the Clementine approach are addressed. Several of the lessons learned from the Clementine mission underscore the recommendations found in COMPLEX's report, The Role of Small Missions in Planetary and Lunar Exploration,1 as well as in previous documents prepared by other committees of the Space Studies Board.2 -4 COMPLEX's prime recommendations in these areas have included the following:
The budget, schedule, and risk envelope should be defined in advance and adhered to;
The approach to project management should be streamlined so as to eliminate excessive reviews and oversight; and
All data and information necessary for their interpretation should be promptly archived in NASA's Planetary Data System.
Clementine was launched less than 2 years after it was approved by the Department of Defense. The 22-month (or 28-month, using the criterion assumed in Table 3.1) development schedule imposed on this mission is acknowledged by the project management to have been significantly too ambitious, resulting in inadequate readiness for flight operations. Insufficient time was available to develop and fully test the flight software, to carry out enough end-to-end testing of the spacecraft and telecommunication links, and to recruit and train the full operations team. The deficiency manifested itself in an excessively stressful mission-operations phase that was clearly exhausting to the team and may well have contributed to the ultimate demise of Clementine after the lunar portion of the mission had been completed but before insertion onto the asteroid flyby trajectory.
TABLE 3.1 Comparison of Small Exploration Missions
Although the short schedule caused significant problems, the limited available time brought some benefits. These included a lower total price and accurate cost projections; a focused mission that was able to address problems of current interest with the latest technology; and maintenance of the enthusiasm of the staff who, for the most part, stayed on for the duration of the mission.
COMPLEX believes, based on comments from the Clementine operational and management teams, that a 6- to 8-month-longer schedule would have been much more realistic for the mission. Such a development time would inevitably call for a larger budget than that made available to the Clementine team. COMPLEX's discussions with personnel involved with Clementine suggested that this increment would have been a small fraction of the project's total cost.
Future Discovery missions planned by NASA's Office of Space Science assume a nominal 36-month development phase; NASA's other "smaller, faster, cheaper" science programs (e.g., Earth Probes, Earth System Science Pathfinders, Solar-Terrestrial Probes, Small Explorers, and MidEx) also demand short development times. The Clementine experience indicates that such a schedule can be adequate, but only if the Discovery projects enjoy the other advantages of Clementine, such as unchanging objectives, disciplined management, a proven launch vehicle, and resources made available according to the original plan. A 36-month schedule would not be adequate if the scientific instruments require substantial development. No such development was needed for Clementine since BMDO/DOD had earlier made substantial investments to bring the necessary technology to maturity.
The cost of the Clementine mission was, according to the figures supplied to COMPLEX (see Table 1.1), significantly lower (in inflation-adjusted dollars) than that of any previous planetary mission, and less than the projected expenses of three of the four Discovery missions selected to date (see Table 3.1). In judging Clementine's price, it must be recognized that the mission had both successes and failures. In the technology arena, Clementine was successful in achieving its goal of space-qualifying sensors and spacecraft subsystems, but the mission failed before autonomous acquisition and tracking of a moving object could be tested. Scientific goals were accomplished at the Moon; however, the asteroid science objectives were not achieved.
Even given the partial failures, the general perception exists that, for planetary missions, Clementine has set a new standard of performance within a constrained budget. This perception needs to be tempered by the fact that there are apples-vs.-oranges reasons that a DOD mission might come out ahead in a cursory comparison with a NASA mission. DOD's inherent advantages included:
A corporate culture that was able to dictate a schedule that was too tight. A more realistic schedule would necessarily add to the cost (as well as perhaps reduce the risk).
Previous investments in spacecraft technology by DOD, which meant that the additional cost for this mission could be relatively modest.
A degree of institutional strength in negotiating DOD's many contracts that has not been achieved by NASA in the past and may not be achievable with the kind of open process traditionally utilized by NASA. For example, government procurement rules prohibit the use of a single vendor, but this restriction compromises a project's ability to quickly develop and purchase specialized technological devices even if it is known that only a few contractors are experienced in a particular line of work.
Absence of certain overhead costs that are typically borne by NASA science missions (specifically, the Clementine project was charged only for its full-time team members and not for part-time support).
On the other hand, a NASA science mission, by its very nature, will incur cost penalties not applicable to DOD missions such as Clementine. These include:
The cost of the science team. NASA established and funded Clementine's science team to validate the data and plan for its archiving.
Data analysis expenses. In contrast to NASA's traditional policy, no data analysis was supported by the mission.
The development of an optimized science payload. Clementine's instruments were not optimized for scientific observations.
The proper calibration of the science payload and the data it returns. Clementine's data calibration is being paid for by NASA's Office of Space Science.
Provisions for making the data available through the Planetary Data System.
The integrated effect of all these factors on the final budget is difficult to assess without more detailed analysis by those with the appropriate expertise. COMPLEX is confident, however, that even with an adjustment of 50 to 100% (note that in the latter case, the total cost
would be only a little more than Discovery's cost cap of $150 million (FY 1992 dollars), which excludes the cost of the launch vehicle and operations), Clementine still provides a challenging cost standard for space science missions. It appears that this challenge is being addressed by NASA with the Discovery and MidEx missions now under development, as well as by NASA's planned New Millennium missions.
Clementine operated within a rigid cost-cap. This was a two-edged sword: the presence of the cost cap was valuable in forcing economies and prompt decisions, but the rigidity of the cap also meant that insufficient funds were available shortly before launch, when monies could have been valuably spent. The obvious (and already well understood) lesson here is that contingency funds, and their proper disbursement, are vital for the effective management of space missions.
The management approach adopted by Clementine was, according to many of the presentations given to COMPLEX, typical of that used in DOD space programs since the late 1950s. Early NASA programs adopted a similar style, but many of these practices have been eschewed in recent decades. In essence, a "skunkworks," a small hand-picked team, was given complete responsibility and accountability for the success of the project from design and construction through to launch and operation. BMDO program management served to provide resources according to the agreed schedule and to facilitate procurements and interfaces with other federal agencies; oversight was limited to essential and judicious reviews. In other words, the approach was very similar to that recommended by COMPLEX as the appropriate model for small missions.5
Given the obvious high quality of the team and its commitment to success, the Clementine management approach can serve as a model (for NASA and for the other governmental institutions that control the flow of resources) for the management of future low-cost planetary missions.
Clementine was a technology-demonstration mission that was partially successful in meeting its primary goals. Its failure to completely meet all of its secondary science goals does not appear to have been due to any deficiency in the advanced technology used. Moreover, the significant deep-space capabilities demonstrated within the payload capacity of a relatively small launch vehicle were, in fact, enabled by the lightweight/high-performance technologies used. So, again, a new standard was established in terms of the speed with which space qualification was achieved for new technology in sensors, electronics, structure, and propulsion. NASA should seek more insight than has been acquired by COMPLEX into the process by which this qualification was achieved so rapidly. Indeed, many of the technological aspects of the Clementine mission, e.g., its use of prescreened, commercially available electronic components rather than special-purpose,
radiation-hardened devices, could bear closer examination by groups more appropriately constituted than COMPLEX.
The Clementine technology itself can be of substantial benefit to future missions, both in reducing future costs and in improving science yield. Table 3.2 lists the heritage and possible future applications in the space sciences and elsewhere for some of the advanced technology components that Clementine helped to space-qualify for civilian flight.
TABLE 3.2 Clementine's Advanced Technology Components, Their Heritage, and Examples of Future Applications
A note of caution must be provided to this discussion of Clementine technology. The Clementine mission was carried out in one of the more benign of space environments—lunar orbit, which substantially resembles Earth orbit. For planetary missions that might transit great distances inward or outward from Earth's orbit, the thermal environment becomes extreme, solar electrical power generation may be problematic, and communication delays of tens of minutes or hours create the need for substantial autonomy. Clementine's transit time to the Moon was short and the mission lifetime was abbreviated. By contrast, a more typical planetary mission may take years simply to reach its target, and an Earth-orbiting observatory might have an operational lifetime of a decade or more. Hence, some question must remain as to whether the new technology employed by Clementine would survive an extended flight in a more hostile environment, either much closer to or farther from the Sun than Earth.
Managers' conservatism with respect to the use of new technology in deep-space missions derives legitimately from the extremely challenging environment faced by their spacecraft as well as from the traditional reluctance of NASA to take risks. Such conservatism may be costing NASA's space exploration program a substantial price because the newest technologies can greatly enhance performance at the same time that cost and mass are reduced. Reductions in mass can translate into major savings in launch costs. Since COMPLEX does not claim expertise in spacecraft design and operation, it cannot attempt to resolve the issues associated with new technology and risk. Nevertheless, COMPLEX is encouraged by the general success of Clementine and, accordingly, anticipates that the balance will turn toward the use of the latest technologies with their many potential benefits.
In COMPLEX's view, the Clementine mission, though technological in nature, has been useful for space science. Programs like it, which emphasize technology, should be initiated. The testing of new technologies, fresh management approaches, and innovative spacecraft-operation schemes as the primary objectives in a low-cost, short-duration program is an important component of a healthy space program. Two essential features of such missions must be that they do first-class science and that the scientific community become involved in them as early as possible.
NASA's New Millennium program could, potentially, continue the role of technological innovation begun by Clementine. To do so, however, the organization of New Millennium will have to differ considerably from various past NASA programs that were organized with the intended goal of developing valuable space technologies; occasional external oversight of such a program might be useful.
A broader issue raised by the Clementine mission concerns the degree to which DOD's space-technology activities, such as Clementine and its possible successors, should be coordinated with similar NASA technology endeavors such as New Millennium. NASA's technology programs have been roundly criticized for many years.6 Competition from other organizations—whether governmental laboratories (e.g., the Naval Research Laboratory or Phillips Laboratory), quasigovernmental facilities (e.g., the Applied Physics Laboratory), or commercial ventures—could be constructive, if properly managed. A full assessment of the potentially important benefits to space science that could be achieved by a technological alliance between NASA and DOD deserves attention at the highest levels in both organizations and close scrutiny by a group more appropriately constituted than COMPLEX.
The mission operations phase of the Clementine project appears to have been as much a triumph of human dedication and motivation as of deliberate organization. The inadequate schedule, referred to above, ensured that the spacecraft was launched without all the software having been written and tested. The inflexible budget imposed on the project also meant that the complete operations team was not recruited and trained in time to be ready for launch. Thus the team was under enormous pressure from the outset. This pressure was heightened by the operation team's self-imposed determination to completely achieve the science goals related to mapping the entire lunar surface (which required, for example, some ingenuity to fill inevitable gaps in coverage by reprogramming the observation sequences). Clementine's operations team was able to accommodate orbit-by-orbit variations in the configuration or operational mode of instruments based on the science team's examination of data already in hand. The team also agreed to conduct last-minute additions to the scientific program, such as the bistatic radar experiment, and was remarkably responsive to the additional requests of the science team. Further, the spacecraft performance was marred by numerous computer crashes. It is no surprise that the team was exhausted by the end of the lunar mapping phase.
The extraordinarily large volume and completeness of the lunar data set acquired in the face of the problems cited above indicate that the Clementine project's approach to organizing and staffing the operations phase of the mission enjoyed many essential strengths. The organization of the operations team was carefully designed by the project management to provide minimal sufficiency and, within the severe limits of schedule and budget, intense training and rehearsal. It would be very useful for some group, more appropriately constituted than COMPLEX, to undertake a study into the process by which the mission operations software was developed. Given Clementine's small budget, it seems likely that relatively little software was custom-designed; there are likely to be valuable lessons to be learned by Discovery program participants here.
To observers of recent space-science programs, a curious aspect of Clementine's failure to achieve one of its primary objectives (namely autonomous tracking of a cold target) was that BMDO suffered little public embarrassment over this loss. This was in sharp contrast to NASA's disgrace when Mars Observer was lost just 6 months prior to Clementine. The message may be that technology-demonstration missions are expected to be difficult undertakings, whereas even challenging space science missions, such as Galileo, are assumed by the public to be fail-safe because of their high cost.
The volume of data returned by Clementine was very large, about 10% of the amount returned by the $700 million Magellan mission from Venus. Thus the task of data management was a daunting one, given the limited budget available to the project. The project evidently succeeded in collecting the data, storing them in an orderly way, and making them available to the science team during real-time operations. Thus, to first order, the Clementine project has demonstrated that, despite tight budget constraints, large volumes of data can be managed. Full reduction and organization of the data set for inclusion in NASA's Planetary Data System were not planned for, nor budgeted for, by the Clementine project. This responsibility devolved upon the scientists and was stated in NASA's Announcement of Opportunity7 to be the primary role for members of the science team. Nonetheless, funds to allow calibration of the returned data were not provided and are now coming from other sources in NASA's Office of Space Sciences.
An important concern about the ultimate value of the Clementine measurements remains the degree to which the data can be calibrated for quantitative analysis. This concern arises more from the way in which the sensors were selected and developed for flight than from inherent limitations in the Clementine data management process.
1. Space Studies Board, National Research Council, The Role of Small Missions in Planetary and Lunar Exploration, National Academy Press, Washington, D.C., 1995.
2. Space Science Board, National Research Council, A Strategy for the Explorer Program for Solar and Space Physics, National Academy Press, Washington, D.C., 1984.
3. Space Science Board, National Research Council, The Explorer Program for Astronomy and Astrophysics, National Academy Press, Washington, D.C., 1986.
4. Space Science Board, National Research Council, Strategy for Earth Explorers in Global Earth Sciences, National Academy Press, Washington, D.C., 1988.
5. Space Studies Board, National Research Council, The Role of Small Missions in Planetary and Lunar Exploration, National Academy Press, Washington, D.C., 1995, page 28.
6. See, for example, Space Studies Board and Aeronautics and Space Engineering Board, National Research Council, Improving NASA's Technology for Space Science, National Academy Press, Washington, D.C., 1993.
7. Office of Space Science and Applications, NASA, "Science Team for the Clementine Mission Deep Space Program Science Experiment (DSPSE)," NRA-93-OSSA-2, NASA, Washington, D.C., January 15, 1993.
Last update 7/19/00 at 9:49 am
Site managed by Anne Simmons, Space Studies Board