The Space Reviewin association with SpaceNews

Dragon at ISS
Could the approach used to develop commercial capabilities to service the ISS, like SpaceX’s Dragon spacecraft (above), have applications elsewhere for NASA? (credit: SpaceX)

“COTS-like”: the future of space procurement

Bookmark and Share

The current Administration’s Review of Human Space Flight Plans Committee (aka the “Augustine Commission”) has generated a lot of debate, but one message has emerged loud and clear: NASA is going to have to find a way to do more with less, which means a greater use of commercial procurement practices. As USAF Colonel Peter Garretson observed in a recent article, “The most important and transformational program at NASA is not Constellation, but rather COTS [Commercial Orbital Transportation Services] and its innovative partnership and prize programs, which are focused on a meaningful and more important sustainable expansion of viable American capabilities.” (see “Elements of a 21st century space policy”, The Space Review, August 3, 2009) As NASA determines how best to engage the commercial space sector that is increasingly becoming an established part of the aerospace landscape, they should take a close look at their own Commercial Orbital Transportation Services (COTS) program, which the Obama campaign identified last year as a good model for government/industry collaboration.

The COTS model

Established in 2005, the COTS program is unprecedented in its structure and innovation. Initially conceived as a high-risk, high-return, low-cost experiment to test whether commercial industry could feasibly develop and demonstrate a “FedEx-like” cargo delivery service to ISS, the program competed and awarded a total of $500 million to two relatively new commercial companies to demonstrate this capability, with the potential for a “Phase 2” operational servicing contract (now known as Commercial Resupply Services, or CRS).

The key features that have made the COTS program unique and effective are:

One message has emerged loud and clear from the Augustine committee’s deliberations: NASA is going to have to find a way to do more with less, which means a greater use of commercial procurement practices.

Use of Other Transactional Authority (OTA): COTS is not implemented as a traditional government contract but as a “Space Act Agreement” (SAA) in accordance with the Commercial Space Act of 2003. SAAs generally are applicable to rapid prototyping and demonstration efforts, rather than services procured by the government. Reliance on SAAs relieves both the government and industry of Federal Acquisitions Regulations (FAR) with all of its associated overhead. Certain SAAs allow industry “partners” access to government expertise and facilities at government-internal rates. The agreements provide a mechanism for government employees to render assistance and advice by providing the accounting framework for charge numbers to be created. Conversely, the agreements allow commercial industry to operate by its own best practices, largely unburdened by the usual processes and overhead mandated under traditional government contracts.

Performance-based milestone payments: As part of the award negotiations, milestone definitions and payments are negotiated between the government and the industry partner. Payments are only made once the success criteria for a given milestone have been met to the government’s satisfaction. The result is low cost-risk to government; there are no sustaining funds or cost-plus contracts. Furthermore, if the industry partner fails to meet a milestone in the agreed timeframe, the government has the right to terminate the agreement for cause. Alternately the government can renegotiate the remaining milestones with the partner if it is deemed to be in the government’s interest. When one of the original COTS contractors failed to meet milestone obligations under its agreement, NASA re-competed and awarded the money to another provider, maintaining the objective of keeping two viable commercial providers. This process was a demonstration of the flexibility and robustness of the COTS paradigm. Even when a COTS partner “failed”, NASA essentially lost a limited, controllable amount of funding (though they likely gained valuable insight and lessons learned to apply to future projects.)

Government flexibility: The use of SAAs provides the government a large degree of latitude in selecting partners and making awards.

Requiring contractor “skin in the game”: With COTS, the government required commercial investment to supplement government funds. It was anticipated at the outset that the funding available from NASA would not be sufficient to entirely fund the development and demonstration of multiple commercial cargo delivery systems. It was expected that the partner would supplement NASA money with private capital. Commercial investment was incentivized by the probability (but not certainty) that NASA would be a significant customer for the capability developed. This results in “multiplication of investment” for the Government.

Contractor-retained intellectual property: In return for this commercial investment, ownership of the system and its design is retained by the commercial partner, except in the case of default. This allows uninhibited commercial sale and application of capabilities developed. Under typical FAR-based government contracts for the development of goods, the government pays in full and owns the result.

Only high-level objectives mandated: The government specified only broad technical objectives and performance goals, not detailed requirements or implementation restrictions. This is critical since other, non-government applications are assumed in order to stimulate commercial investment in the program. This “what, not how” approach is also generally a good model for cost optimization since it allows commercial competitors to perform their own trades of performance and cost (both NRE and recurring) and arrive at their proposed solution, which the government can then assess and award based on stated (and unstated) preferences.

Hardware demonstration, not paper: COTS was designed to be a demonstration program, not a design study or PowerPoint exercise. Since the purpose of the demonstration program is to follow with a competed services contract that pays per kilogram on-station, it is essential to minimize any bridging development required for an operational system. Maximal feed-forward and applicability is essential.

Insight, not oversight: The government has insight but not oversight and control in the traditional sense. Commercial enterprise is allowed to operate commercially in order to reap maximal benefits. The commercial partner must provide sufficient insight to allow the government to assess milestone completion, progress toward remaining milestones, and verification of critical safety requirements.

The above characteristics are certainly not all unique amongst previous Government programs. Although there are some limited similarities with programs such as OSP, X-33, X-34 and EELV, none of those programs shared the combination of factors outlined above, nor the demonstrable success and value for money that has been shown to date by COTS.

Benefits of the COTS-like paradigm

The advantages of the COTS approach are numerous. The government gets access to a capability that it needs, for substantially lower non-recurring development cost. In fact, in some cases it is possible for the government to end up with multiple different, redundant commercial solutions for less investment than for a single government-funded development. Furthermore, recurring and life-cycle costs will also be lower than for a government-run operation due to the efficiencies of commercial implementation and competition, especially if non-government customers and applications also emerge.

The most important benefit of this approach is that the availability of the newly-demonstrated commercial capability allows the government, in this case NASA, to off-load work and de-scope requirements from the formidable task before it.

It is also worth noting that commercial applications for any emergent capability, like commercial space transportation for cargo or crew, tend to lag behind the initial demonstration. That is, most entrepreneurs and investors would balk at a business plan that requires commercial space access until after it is clearly demonstrated to be feasible and profitable—at which point venture capital is likely to clamber to make use of, and profit from, this new frontier. So if non-government markets can be identified with some certainty even during the proposal phase, then realization of the capability will only result in more customers with associated commercial economies. This is often cited as an important and appropriate role for government-funded research and development: to be an early investor and customer to help commercial enterprise “over the hump” by proving feasibility and commercial viability, but not retaining ownership or operational control. This develops and expands the domestic industrial base and capabilities, while also benefitting the government.

The most important benefit of this approach is that the availability of the newly-demonstrated commercial capability allows the government, in this case NASA, to off-load work and de-scope requirements from the formidable task before it. This has already been done with the removal of cargo transportation requirements to ISS from Orion spacecraft, as commercial and other foreign cargo vehicles approach operational status. Project Management 101 teaches that cost, schedule, and requirements are the three pillars of any project and can be traded against each other to a large degree. By off-loading requirements, cost can be reduced. Alternately, more can be achieved within existing budgets by focusing resources (people, facilities & money) on the remaining tasks.

Essential ingredients

Not all government developments are appropriate candidates for implementation in the manner outlined above. For example, where there is significant technical risk in a development, commercial industry may be reticent to invest its own funds without a high degree of confidence that a solution is plausible, well-bounded in time and cost, and extensible beyond the immediate government application. In these high-risk developments the government will need to promise full reimbursement of development costs, as under traditional cost-plus government contracts.

In addition to this prerequisite, essential ingredients for COTS-like projects include:

  • Clear government need, as part of an established priority or funded program with reasonable likelihood of ongoing political support, such as human space flight (regardless of the architecture or objectives). This is critical to stimulating commercial investment.
  • Cost efficiency must trump a specific implementation or detailed design solution. This is essential for the solicitation to specify only broad design goals and not detailed requirements.
  • There must be multiple viable commercial competitors to complete for the award. Viability means “commitment and credibility”, and includes technical competence and financial/business solidity.
  • There must be a reasonable likelihood of commercial applications beyond the immediate government need. This further motivates commercial investment and drives down operating and life-cycle costs. As discussed above, additional entrepreneurial applications will likely lag demonstration.

Candidate developments for future COTS-like procurements

Clearly commercial cargo transportation to ISS, as implemented under CRS, fits the above criteria. Given existing NASA plans and objectives, some potential candidates for COTS-like development and demonstration programs include:

Commercial crew transportation to and from ISS: This is a logical and incremental extension of cargo transportation capabilities, especially if the cargo system includes down-cargo capability (i.e. return of payload to Earth). Many functions and requirements for crew transportation are levied on the cargo vehicles by virtue of the fact that they must approach (and berth with) the ISS. Safety concerns for ISS crew, and prudent stewardship of the $100-billion ISS itself, mandate that factors of safety, fault tolerance, air circulation, touch temperatures, sharp edges, and many other “human rating” requirements be imposed on the cargo transfer vehicles. Accommodating crew involves up-rating of some subsystems, adding crew monitoring and over-rides, and a launch escape system in case of booster failure during ascent.

Docking adapter: Almost all space exploration architectures ever proposed include this as a critical element, permitting on-orbit rendezvous, assembly and transfer of crew. Astoundingly, the US has not retained this capability since the end of the Apollo program, opting instead to adopt Russian systems on ISS and shuttle. The Low Impact Docking System (LIDS) has been under development by NASA for over two decades with varying levels of support and is now on the critical path for Constellation. A COTS-like commercial crew transportation program could significantly reduce the looming six-year gap in US human space flight capability if initiated soon, while LIDS is unlikely to be available before other parts of this system are ready for demonstration. Furthermore, LIDS would be a critical dependency on the government in the midst of an otherwise commercial system, representing an unacceptable risk to commercial competitors. A commercial solution, compatible on the back-end with defined ISS interfaces, is essential for commercial crew. This would also benefit proposed non-government customers, such as Bigelow Aerospace, that require a docking solution. These factors make a commercial docking system an excellent candidate for COTS-like development.

Rather than de-scoping national human space exploration objectives, this can be an opportunity for constructive change.

Lunar reconnaissance landers (~100 kilogram-class): The step from a fully autonomous cargo transfer vehicle to a lander is not as large as one may initially believe, especially when combined with launch vehicle capabilities. This basically requires some new structure and mechanisms—no more complex than launch vehicle or cargo spacecraft designs—and a descent engine with deep-throttle capability. Guidance and sensor requirements are no more onerous than those for berthing with ISS. The size (mass) of the lander is primarily determined by launch vehicle capability and costs. Over a dozen teams are currently competing for the Google Lunar X PRIZE to land 10 kilogram-class landers on the Moon. An astonishing amount of utility can be packaged on a lander of this scale. The 100 kilogram-class offers greater capabilities that could prove immensely useful to the Constellation program in performing precursor science, landing site selection and reconnaissance, infrastructure such as navigation and communications aids, and in-situ resource utilization (ISRU) demonstrations to support human exploration.

Lunar logistics landers (~1000 kilogram-class landers): Same as above, but with enough capacity to allow pre-positioning of supplies, tools, rovers, and even ascent/return vehicles in support of human exploration. This class of lander can greatly enhance capabilities for initial sortie missions, decrease risk and increase safety, and can be launched on existing EELV-class boosters.

Lunar communications services: Of all the functions that could feasibly be performed by commercial enterprise this is possibly the most obvious. Both terrestrial and space-based commercial communications infrastructure has attained such maturity as to be taken for granted by the majority of people today. Extending this capability to the lunar surface and lunar orbit is an incremental development that could fully support NASA’s need under the Constellation program.

Earth observation data: As NASA’s Science Missions Directorate struggles to deliver Earth-orbiting missions and space-based observation data within existing science budgets (strained by reallocation to the development of exploration systems), one could ask the question: why should NASA contract these spacecraft and operate them on-orbit? Why not instead specify the type and quality of data required and offer to buy it by the gigabyte? There are numerous commercial companies already designing, building, and operating all types of spacecraft, in all types of Earth orbits from LEO to GEO. And with fully commercial launch services becoming available, NASA does not even need to provide the booster, just a contract for the data. In this way several similar, redundant sources of data from orbit could be made available to NASA for less than the cost of a single mission, and with no cost risk to NASA. Commercial applications for this data are also likely to emerge as has been seen with radar, imaging, multi- and hyper-spectral data sets over the past few years.

F-1-class engine development: At least three domestic commercial companies are well placed to compete for the development and supply of a large booster engine in the class of the Saturn 5 first stage F-1 engine (1.5 million pounds-force). Between six and nine of these engines would enable an all-liquid heavy-lift booster of Ares 5 lift capability, as required for almost any exploration architecture. Liquid rockets are intrinsically safer and more flexible than those using solid motors, and are especially safer than hybrid solid-liquid architectures. This is particularly true if engine-out capability is included, as on the Saturn 5. If NASA were to specify only propellants, broad technical performance goals, and the technical interfaces, it could attain at least two compatible suppliers of engines for its heavy-lift booster. The application of this class of engine to other, smaller commercial launch vehicles is also conceivable, perhaps as a successor to the EELV-class boosters.


It is generally accepted that NASA will not receive sufficient funding over the coming decade to achieve the exploration goals before it. But rather than de-scoping national human space exploration objectives, this can be an opportunity for constructive change. By off-loading requirements commercially, NASA could achieve more of its exploration goals within existing funding limits, while the commercial industry picks up the lower-risk pieces of the architecture. This allows NASA to focus its resources (personnel, facilities, and money) on the “edge of the envelope” developments while simultaneously stimulating the American aerospace sector and industrial base.

All that this approach requires is a will to embrace industry as a full partner in enabling NASA and this nation’s plans in space

What if NASA were to step back from its exploration plans as currently conceived, survey the whole architecture, and ask: which pieces of this could conceivably be broken out and handed to commercial enterprise? Once candidate segments were identified in accordance with the criteria outlined above, NASA could then take a significant portion of the funding that it was planning to spend developing that piece of the architecture and hold a fixed-price COTS-like competition to develop and demonstrate that capability. In order to provide assurance to industry that NASA will be a customer in the short-term it should alert stakeholders that it does not intend to spend any additional money on this piece of the architecture. This will allow American commercial enterprise to step up to the plate and provide this capability, while also allowing NASA to focus its (now augmented) remaining resources on the pieces that are not good candidates for commercial development, specifically those with higher technical risk.

This COTS-like approach would herald a new era of cooperation between NASA and industry, where responsibilities, costs, and risks are more evenly shared, to the benefit of both. Together, more can be achieved with the same level of government funding. The targeted stimulation of the commercial space sector will not only enhance national space exploration capabilities, but will ultimately result in a larger, healthier, truly commercial industrial base that can grow to support far more skilled workers than under the existing government-supported paradigm.

Note that this solution is independent of both the specific architecture and of its goals—it is a new approach to government procurement. It will work whether NASA decides to continue to pursue Ares 1 and Ares 5, side-mount shuttle derivatives, in-space fuel depots, or any of the other architectures that are under consideration. It is also independent of whether NASA aims for the lunar surface, asteroids, or Mars. All that this approach requires is a will to embrace industry as a full partner in enabling NASA and this nation’s plans in space.


ISPCS 2015