Readiness for Open Systems: How Prepared Are the Pentagon and the Defense Industry to Coordinate?

Available Downloads

CSIS Briefs

The Issue

Driven by advances in commercial technology and increasing investments by competitors, the Department of Defense (DOD) is attempting a transition as to how it acquires new technology by following in the footsteps of commercial open systems like IBM personal computers or the Android smartphone. This method of acquisition would rely on open standards and interfaces to increase competition and make integrating innovative technology easier. However, success will disrupt present business models and requires navigating a difficult coordination problem. To better understand how to make this transition a success, this paper presents a framework for evaluating the DOD’s readiness for Modular Open Systems Approaches (MOSA).

Why MOSA? Why Now?

MOSA is an attempt to change the way the DOD buys weapon systems by making it easier to incorporate technology from a range of sources. Modularity refers to the segmenting of systems into tightly integrated systems or components that are loosely coupled with one another. Standardization in weapons goes back to the interchangeable parts of the early industrial age. Today software is increasingly important and often a module itself. For modern complex weapon systems, segmenting into modules is commonplace. Still, the interfaces between modules, the overall system, and the modules themselves are often proprietary with the intellectual property (IP) and data rights owned by the manufacturer or integrator. Openness means that key interfaces instead use an architecture that is freely available and ideally in widespread use. This increases competition and, through standardization, reduces the transaction costs of developing and bringing together modules.

The DOD has been pursuing some form of MOSA for decades, but efforts have intensified because DOD innovation has often been outpaced by the commercial sector. This period has also seen proofs of concept of incremental improvement enabled by software-intensive adaptable systems. The transition of the Aegis Combat System to an open architecture is an example of employing commercial off-the-shelf infrastructure and achieving faster upgrade cycles for hardware and even more so for software.i By emphasizing speed, innovation, and competition, MOSA addresses the acquisition reform priorities not just of present efforts to speed acquisition but also the prior attempts at cost controls. In economic terms, MOSA does not immediately create efficiencies that compensate for growth for the defense budget, as many expenses will be front-loaded. However, continued economic growth by the People’s Republic of China, combined with greater competition for labor within the United States, suggests that investing in innovation at reduced cost is a more sustainable path to superiority than relative defense budget increases.

The transition of the Aegis Combat System to an open architecture is an example of employing commercial off-the-shelf infrastructure and achieving faster upgrade cycles for hardware and even more so for software.

Regardless of urgency, the long history of attempts raises the question of whether MOSA can deliver on its promises. The commercial sector provides the greatest reason for faith: open systems are a concept routinely used in day-to-day life, from the rise of IBM personal computers to Universal Serial Bus (USB) standards to Android smartphones.ii These high-profile examples involve information communication technology, but vehicles also have a history of open systems. In the automotive sector, the GENIVI Alliance and AUTOSAR achieved greater standardization and reuse for automotive software.iii In aerospace, open integrated modular avionics have grown up in parallel in the military and commercial sectors and are used by both the Airbus 380 and Boeing 787, with the former citing “cut[ting] in half the part numbers of processor units” and the latter reducing weight by “2,000 pounds.”iv

Commercial open systems show that there are real world examples for MOSA to learn from and emulate, but also raise the question of why these approaches are not just directly and immediately adaptable to defense. In some cases, they are; ethernet is an example of one commercial standard directly employed by the government. But more often, military specifications clash with commercial approaches and may require relaxing requirements, developing government-specific standards, or finding ways to adapt commercial standards to meet cybersecurity concerns. Unlike the commercial sector, leading defense firms do not necessarily have a straightforward prospect of increasing the overall market size for their products, even if MOSA delivers on its goals. Similarly, their customers can unilaterally change the requirements for future products. As a result, industry investments to build architectures and adapt their systems and components to comply with them depends on confidence in the government’s MOSA approach.

How should the transition to MOSA be measured? Despite high-profile standards efforts, in 2013, the Government Accountability Office (GAO) critiqued slow progress toward MOSA.v Since then, the profile of standards efforts has notably risen, including the software-oriented Future Airborne Capability Environment and the Air Force’s Open Mission Systems. However, the GAO’s suggestion of wider use of MOSA metrics has made little progress and, despite the existence of tools for measuring openness across programs, metrics are largely absent in public Early steps such as reporting on MOSA goals in Selected Acquisition Reports cannot yet be assessed due to a slow pace of release. The National Defense Industry Association, a collaborative organization for the defense companies, agrees that metrics are critical but must be closely tailored to each programs goals, suggesting that progress may come slowly.vii

This paper focuses on the topic of readiness rather than the progress of individual projects for two reasons: the need to get MOSA right at the early stages of the project and the coordination problem inherent in an open-source industrial transition. The technical opportunities of designing an open architecture and the business leverage for the government are greatest at the start of a project.viii The DOD is attempting a larger technological refresh and faces decisions over preserving, upgrading, or replacing legacy systems. In addition, the middle tier of acquisition has sought to allow for faster-moving program starts, although this has not yet translated to a series of production program starts.ix Readiness matters because the opportunity to give these range of programs a solid start with MOSA is easily missed and hard to recapture. The larger industrial transition challenge will be explored more below, but in short, MOSA upends some traditional business models and requires a degree of partnership between government and industry in developing and implementing standards. Readiness matters because the question of whether or not transition will be rewarded is important to industries’ decisions on whether to invest in MOSA or to sit it out (or maliciously comply).x

How to Think about Readiness

When considering a multifaceted and multisided challenge such as MOSA adoption, an overarching framework can help in laying out the big picture and showing where attention must be focused. This paper proposes that MOSA readiness has seven dimensions that can be grouped into three broad categories. Each of the categories has both government and industry aspects and most involve both technical and business considerations. A lack of readiness in any of the dimensions could imperil success even if progress is made in other areas. Moreover, readiness is rarely evenly distributed throughout the DOD enterprise or industry and identifying areas of strength can encourage wider adoption.

The first category, Communication and Commitment, involves sharing knowledge between both government and industry expertise, addressing uncertainty, receiving feedback from industry and users, and establishing confidence in follow-through. There are two dimensions under this category:

  1. Government and broad industry communication about achievable first iterations and future roadmaps
  2. Sustained government commitment to MOSA objectives and a credible MOSA funding model

The first dimension relates to the system engineering problem of determining what open interfaces will provide the most benefit, for example determining which technology’s upgrade cycles would produce the greatest capability. The second dimension involves sending a clear message that the move to MOSA is happening and cannot simply be waited out or undermined. One of the most powerful ways for the government to communicate this is through budgets, which shows how messaging translates into resources dedicated to MOSA-compliant systems and processes.

The second category, Enabling Environment, provides the foundation for turning intentions and project-specific investments into successful results. This is the infrastructure and expertise for MOSA. Although often government-led, it is also dependent on industry participation and initiative. There are three dimensions under this category:

  1. DOD and service enterprise investments in MOSA enablers
  2. Government engagement with key enablers and cross-platform standards development that builds and sustains consensus with industry
  3. Acquisition and sustainment workforces’ business and technical expertise

The third dimension reflects that many of the advantages of MOSA are not achievable on a project-by-project basis. DOD as a whole and the services must look at the bigger picture, coordinate within portfolios, and share knowledge and approaches. The fourth dimension requires partnering closely with industry—and not just a single prime contractor, but a range of competitors at multiple tiers. This process can be slow and often involves tracking developments in the larger commercial industry, where DOD is not in the lead. The fifth dimension is the workforces, from both government and industry. The workforce will face technical challenges, including designing and maintaining open architectures, mastering new and evolving standards, and making a larger shift toward more iterative and software-based approaches. On the business side, approaches like other transaction authority offer more flexibility, and licensing arrangements can help close the business case. However, this flexibility can demand more from the defense workforce, as they involve tailoring but also ensuring core openness requirements are met.

The third category, Business Models, underlines the change from a traditional acquisition environment to an uncertain new one that draws on and reflects the successes of commercial open standard approaches, but also incorporates the unique considerations of defense acquisition. There are two dimensions in this category:

  1. Business models that incentivize defense-industrial base transition
  2. Expansion of the supplier base and inclusion of commercial technology

The sixth dimension relates to defense-industrial base transition to MOSA. These models will need to reward investment in ways that do not rely on backloaded profits made during a vendor-locked sustainment phase. They will also need to effectively incorporate new advantages, such as wider markets opened by standardization and the possibility of more opportunities to compete throughout the lifecycle of the program. The seventh dimension looks beyond traditional primes to greater participation by nontraditional and lower-tier companies. In parallel, MOSA holds the potential to incorporate more commercial technology, as long as cybersecurity concerns can be addressed. In both cases, adapting acquisition approaches may be necessary to ensure that the goal of a more inclusive industrial base translates into reality.

Why Is Achieving Readiness Hard?

The overarching issue that makes implementing MOSA in defense acquisition so difficult is that achieving a critical mass for transition involves coordinating more stakeholders than that of similar efforts in the private sector, either led by a single company or driven by a handful of pivotal players. MOSA has a coordination problem because achieving the benefits requires multiple actors, both inside government and industry, managing conflicting interests to act in tandem. Settling on standards is a common example of a coordination problem because many of their benefits depend on the extent of adoption. From a vendor perspective, as open standards and interfaces become more widely adopted, they can take their products and ideas and offer them to a wider market of potential users while reducing the reengineering and investments to master the eccentricities of each system.

A relevant classic model of a coordination problem is a stag hunt. This model imagines a group of hunters who collectively can patiently play their role to have a chance of trapping a large stag and all eat well. However, rather than vigilantly working as a team, a hunter could instead chase after a rabbit, guaranteeing themselves a smaller meal but potentially allowing the greater prize to escape. In the context of MOSA, catching the stag can mean collaboration within and between government and industry to achieve relevant MOSA goals. Hunting the rabbit can mean abandoning MOSA goals for short-term priorities or a vendor undercutting openness to enhance their position as an incumbent.

Incumbent advantage is one example of a second aspect to the coordination problem: the choices of what interfaces to open and the characteristics of standards have distributional implications.xi Standards choices involve trade-offs; different military services optimize for different missions, so operating environments and standardization can mean either sacrificing some specialization or increasing complexity in order to serve multiple purposes. Worse yet, while all vendors have some potential gains from MOSA, reduced IP and data rights ownership may lead them to fear that their share of the benefits in a MOSA model may no longer justify investment, or even bidding.

MOSA has a coordination problem because achieving the benefits requires multiple actors, both inside government and industry, managing conflicting interests to act in tandem.

Looking at the transition to MOSA as a coordination problem underlines why MOSA readiness matters. This paper’s readiness framework was presented to a working group of experts from government, industry, and academia to explore these dimensions and the present state of readiness. Expert feedback provided examples and nuances. One common thread was that the incentives for MOSA transition are often poor when considered at the individual program level. But the experts had their disagreements, most notably regarding the importance of the degree of modularity and testing approaches.

Communication and Commitment

The first category, Communication and Commitment, is critical for multiple reasons: communication is central to coordination problems, definitions of MOSA are varied, and many key decisions need to be made up front.

Dimension 1: Government and broad industry communication about achievable first iterations and future roadmaps

The first dimension is about finding suitable stag hunting grounds: sharing enough information to engineer systems that will meet critical mission needs, effectively mobilizing industry’s capacities and investment, and acting as adaptable systems in the field by improving with iteration. Government failure to achieve readiness in this dimension would be shown by setting requirements that have MOSA requirements added on top of a traditional system approach, rather than integrated into it. For industry, lower readiness may mean simply applying the MOSA buzzword to existing plans or participating half-heartedly if at all in working groups and coordination bodies. Greater readiness is indicated by the government knowing what it wants. This could look like leadership specifying what MOSA objectives they are prioritizing or laying out cases that demonstrate how they expect results will be improved.

Workshop participants raised cases from different services and underlined the importance of feedback for this dimension. The Army MOSA implementation guide was praised as making good suggestions on decisions the government should be making and communicating. The B-21 bomber was cited as a positive example of employing an open mission architecture, where past Northrup Grumman investments and clear government priorities have kept requirements stable during development. As the project enters the prototype stage, the Air Force has acquired key IP and data rights and expects that the open architecture will allow them to integrate new systems quicker and employ continuous, rather than block, upgrades.xii Even though key choices are made at the front end of long-term project, both before and after these choices are made, successful communication is a feedback loop; approaches evolve and feedback from the end user community, as well as a range of industry, helps guide future choices.

Dimension 2: Sustained government commitment to MOSA objectives and a credible MOSA funding model

The second dimension is about reducing the rewards for chasing a rabbit and increasing confidence that the other hunters will help bring down the stag. Recall that the rabbit in this analogy is to defect from the MOSA goals sought in the stag version of the program and instead advance a program that does not incorporate commonly used open standards or the other MOSA-specific objectives. Given legislative mandates and leadership mandates, both vendors and acquisition officials have reason to describe their programs and proposals as MOSA, regardless of whether they are genuinely open or truly meet standards. As shown in Figure 1, a survey conducted by CSIS for a forthcoming project focused on the defense vertical lift industrial base found hopeful news regarding readiness in this dimension, with a majority of respondents believing that the government had at least a conditional long-term commitment to MOSA.

Some workshop participants had a clear answer for how to lower the benefits of defection for the vendor: “If you don’t conform, we don’t buy.” This is harder than it might sound; discerning whether a proposal or existing program is meeting the goals requires a mix of technical tests and metrics. This may mean making lab facilities available to make verifying module compliance simple or conducting “shoot-offs” that challenge presenting integrators to swiftly incorporate third-party technology. The tougher requirement is that the government must be willing to reject a proposal or modification that is putatively cheaper or faster while offering “good enough” openness or adherence to standard. Trade-offs at the margins are an important part of competition and project evolution, but to achieve readiness in this dimension, MOSA goals must be required or heavily weighted in best-value competition criteria, and thus not easily bypassed. Vendor proposals offering a short-term perk that undermines the long-term benefits of MOSA must be rejected.

While the way vendors are selected shapes incentives, budgets are a grounded signal of government priorities and, thus, of what reward is available for capturing the stag and what opportunities may exist for future hunts. MOSA would not eliminate major development programs, though it should reduce their current winner-take-all character, where a diminishing number of new-start programs are feared to be make or break, with opportunities for subsequent competitions often mediated through the prime. Opportunities could also involve waiting more than a decade for a recapitalization program or other block upgrade. As Figure 2 shows, an increasing number of budget lines explicitly discuss open systems or architectures in their justifications. There are a range of funding models for open systems, including existing funding lines for cross-platform efforts, such as development funding for the Air Force’s Battlefield Airborne Communication Nodes and procurement funding for the Army’s Universal Ground Control Equipment for remotely crewed vehicles. Other examples include purchasing modules for the Littoral Combat Ship; a range of faster turn options under the adaptive acquisition framework; and experimental new approaches such as research, development, test, and evaluation (RDT&E) budget activity for buying software. For this dimension, MOSA readiness means that there are budget and funding models that can support iteration within a program and investments that can bring benefits to multiple platforms.

Enabling Environment

The second category, Enabling Environment, is less focused on the choices of individual programs and vendors and more focused on the larger infrastructure that can enable the coordinated choices. In terms of the stag hunt model, this would cover reducing the difficulty of capturing the stag and, if widespread standards are achieved, increasing the value of each success. The first dimension under this category can address government incentives

Dimension 3: DOD and service enterprise investments in MOSA enablers

This dimension includes a key problem identified by workshop participants: choosing either bespoke standards and interfaces that are not widely used or highly detailed requirements that lock in a particular solution. Even selecting the same standard across multiple programs is only part of the solution, as the many options for implementation undercut cross‑compatibility. One participant cited the analogy that two programs can both use metric bolts, but one turns them clockwise and the other counterclockwise.xiii

The incentives can be wrong at the program level because tying programs together is seen as a source of risk. Achieving commonality can also involve the government taking on the work of leading with key choices for standards and interface, rather than leaving that selection to industry. Such specifications should still leave room for a variety of technical approaches. Budget savings may be achieved in development by doing this right, but the costs may disproportionately fall on a lead program rather than being easily split.xiv To address this challenge, one workshop participant proposed that choosing not to adopt an existing standard or open interface implementation should require a justification. At present, DOD-wide efforts have a more advisory character, creating open interface libraries and trying to capture success stories.xv The U.S. Army Program Executive Office - Aviation goes further: MOSA-related spending within Army aviation must be approved at the enterprise level.xvi

A lack of readiness in this dimension is a shown by different programs facing MOSA challenges on their own and reinventing the wheel. Success is shown by compatible standards and interfaces being adopted across programs and services, as well as resources and expertise from higher levels being used by individual programs. An even more challenging step would be reforming the requirements process to make aligning programs with one another easier.

Dimension 4: Government engagement with key enablers and cross-platform standards development that builds and sustains consensus with industry

This dimension is hard because industry is diverse with often conflicting interests. Building common agreement takes time and standards are never truly “done,” but instead continue to evolve over time as new technologies are introduced and implementation and users generate feedback. From a coordination problem standpoint, industry consensus pertains to the distribution of rewards for reaching agreement. Readiness in this aspect does not refer to total agreement, as even a fair-minded process can generate winners and losers. Instead, readiness means that a wide range of industry stakeholders are contributing to standards development. They help standards evolve in ways that allow them to bring in future innovation and they raise the alarm when a change would mean a system was no longer effectively open to them.

There are a multitude of such efforts ongoing, with the Future Airborne Capability Environment being one prominent cross-service example that has spent years building up an approach to software. Standards are slow to develop, and even as with other forms of advanced military developments, years of efforts will not necessarily be sufficient to achieve proven results. Intriguingly, the workshop participants did not identify areas of key gaps where new efforts needed to be undertaken. Instead, some argued that the multitude of often overlapping standards were sufficient to the task and that instead the priority should be better understanding boundaries and achieving wide use. Another participant cited the Air Force, which has adopted a comparatively government-led approach, for pushing different standard consortiums and turning to rapid capability demonstrations to put concepts to the test.xvii The Air Force chief software officer’s encouragement of infrastructure reuse for Development Security Operations was also raised as a positive example, though the subsequent departure of the that officer underscores the difficulties of the enabling environment for software.xviii The Air Force’s approach was praised for bringing in different platforms, subsystems within those platforms, and vendors, all within a window of six to eight months. Taken together, an absence of readiness in this dimension would be industry and the government talking past one another or limited numbers of prime contractors dominating the development. High readiness would involve demonstrated participation by a mix of vendors, especially those that have a harder time breaking through under traditional acquisition.

Dimension 5: Acquisition and sustainment workforces’ business and technical expertise

While the enabling environment has institutional aspects and builds out technical approaches, it has people as its foundation, as covered by the last dimension in this category. The workforce for both industry and government are relevant here, as are issues of both expertise and culture. Expertise questions are not limited to technical issues; in business terms, MOSA approaches deal heavily with IP questions and often employ mechanisms like other transaction authority and licensing arrangements. This allows for highly customized approaches that have the potential to bring in nontraditional vendors, but in turn means that existing regulation and templates answer far fewer questions. As with other forms of adaptable acquisition, cultural shifts may be necessary to effectively employ these approaches.

Culture shifts may also be necessary on both sides. Some in the government and many in industry see risks of MOSA disrupting relationships that they rely on and, thus, they are unlikely to be forward leaning in ensuring that openness is achieved. For others in government, the relationship with industry can be more adversarial, and breadth of IP and data rights—even within modules—is seen as a necessity. As Figure 3 shows, industrial confidence in government expertise is mixed and lags behind perceptions of government commitment shown in Figure 1.

One workshop participant argued that the key to culture change is sustained persistent messaging across administrations. The Army’s IP roadshow was raised as an example of seeking to bend culture on IP, which required sustained attention and outreach. This can easily be disrupted, even when a new administration agrees with the message, as waiting out a change is a longtime bureaucratic approach. Further, if the message is not reaffirmed, then momentum can be lost.

Technical and business expertise can require new hiring, training, and making experts available to assist with challenges. A lack of readiness in this dimension means that experts are rare, isolated, or overwhelmed and that—by reason of training or culture—familiar approaches are applied even when they are contrary to leadership direction. By contrast, readiness means that experts are effectively deployed and that the wider workforce understands MOSA concepts and is capable of using tools in its domain to make progress toward MOSA goals.

Business Models

The third and final category, Business Models, focuses on the models for industry.

Dimension 6: Business models that incentivize defense-industrial base transition

 The sixth dimension covers how traditional defense industry participants—major primes in particular—are encouraged to compete and invest. Ongoing CSIS research has found that IP and data rights are the first or second concern for the majority of those surveyed. Commercial business models for open systems have been employed for decades and often emphasize licensing approaches. As workshop participants noted, the broad outlines of a MOSA business model are well established: focus investments and business strategies more on modules than interfaces. Original equipment manufacturers can also look to commercial open system analogs and see a potential for enhancing the value and longevity of their product and generating potentially greater sales. The business models for system integrators may be the most complex, as their role is potentially most changed, both in terms of having to integrate government-furnished modules at greater speed and in terms of experiencing a reduction in leverage that comes with the loss of control. Another challenge is the purchase of software, which has been difficult for the government beyond MOSA, as modern software approaches often have fast time horizons and shifting requirements, and do not fit cleanly under traditional government purchasing colors of money.xix MOSA can accentuate these challenges, as a system may be segmented in such a way that a vendor who previously combined their hardware and software sales may now find themselves in a new business model, selling software in support of multifunction hardware.

An absence of readiness would be shown in contracting approaches that implicitly rely on profits during sustainment or that fail to reinforce MOSA objectives. Greater readiness means applying a diverse mix of acquisition approaches that use the adaptable tools available to incentivize MOSA goals.

Dimension 7: Expansion of the supplier base and inclusion of commercial technology

The final dimension is challenging because of the barriers to entry for the defense market, including military specifications, cybersecurity concerns, regulatory requirements such as cost and accounting systems, and export controls. The DOD has attempted a range of approaches to address these barriers, such as the Defense Innovation Unit, small business innovation research, pitch days where contracts can be rapidly written, other transaction authority agreements not bound by many contracting regulations, and even the development of in-house expertise. However, challenges remain: the supplier base shrunk during the budget cap period and the rate of vendors entering the defense-industrial base has declined.xx A workshop participant cited the challenges of bringing commercial software through federal certification processes, as well as access to advanced simulators for testing, as key barriers to entry, especially for small business.

Published open standards and more completed modules are not sufficient to ensure that the supplier base will expand. However, this dimension not only covers nontraditional firms, but also means that lower-tier firms are able to grow—for example, a transponder component maker becoming a competitor for the entire radio. Likewise, bringing in commercial technology can mean incorporating commercial standards, such as with ethernet, or government standards that hew closer to commercial ones.

In the stag hunt model, for nontraditional vendors the choice is one of participating in the hunt or not contracting directly with the government. Organizations like the Defense Innovation Unit exist to help find promising investment opportunities and encourage participation from high technology companies with limited history as prime federal vendors. Microsoft was willing to wade into new territory with its Integrated Visual Augmentation System (IVAS). While Microsoft is not entirely new to defense contracting, IVAS marks a significant endorsement of MOSA, as Microsoft was willing to leverage commercial off-the-shelf technology for a rapidly fielded military application. For lower-tier providers, they are already in the defense-industrial base, but they may need to make additional investments and take on risks to become a competitor on their own. Making more room for commercial approaches or a greater depth of modularization can lower the barriers to entry for these firms, making participation more attractive. Even when participating, these firms could choose to defect and not contribute to the development of open standards, but the traditional acquisition and sustainment system is a less tempting rabbit for these organizations.

An absence of readiness in this dimension is shown by low levels of engagement in DOD-led standards-creating bodies and consortia; a lack of interest in or knowledge of competition opportunities; and ultimately, a stagnant or shrinking industrial base. A higher level of readiness would be shown by greater engagement, new and growing vendors competing for modules, and reduced bifurcation between defense and commercial technology markets.

Key Disagreements

The expert workshop also highlighted a crosscutting area of debate: How deeply should open interfaces be pushed into a system or to what extent should it be modularized? Greater modularization can mean going from subsystems, like a communication system, to a component, like an antenna. Greater modularization can also mean splitting software from hardware. The case for modularization leans on multifunction hardware, with a classic example being the Navy Program Executive Office for Submarines early open system success story that sacrificed legacy approaches to allow for the use of commercial processors and greatly sped technology insertion.xxi The Army’s C5ISR/EW Modular Open Suite of Standards (CMOSS) was also raised by a participant as embracing multifunction components. For supporters, segmentation would qualify as an eighth dimension and enables expanding the supplier base by bringing in larger markets for components like sensors or processors.

This view was opposed by some, who argued that modularity works best at a higher level and that greater segmentation of the system undermines the business case for industry investments because there is less room for IP. This viewpoint argues that commercial subsystems, often produced by companies that manufacture both defense and civilian goods, would be bisected by deeper standards in ways that would make their sale to the government unappealing. For other participants, the extent of modularity is orthogonal to MOSA readiness, with the suitability of deeper divisions dependent on the domain and the MOSA goals.

This division also appeared in a discussion of the testing approaches, with one participant arguing for the importance of end-to-end testing in addition to the testing of individual components. This also came up with a criticism of the IVAS program that argued that IVAS met leadership goals but was judged a poor fit by military end users.xxii Early testing—and sometimes making major changes in response to operational user concerns—is critical to adaptable system MOSA success.xxiii Advocates of greater modularity argued for the adoption of an open integrated modular avionics model and incremental certification—both approaches demonstrated in civil aviation—as vital for achieving rapid upgrade cycles.xxiv These integration considerations lead back to the workforce dimension, as under MOSA, the government workforce does have an increased integration role.


MOSA readiness is multifaceted and can also vary between different technology groups and DOD customers. The three categories and seven dimensions provide a systematic way of thinking about the problem, which matters because both success and failure can be mutually reinforcing. Success in one area can demonstrate best practices and mature standards and offer opportunities for commonality. On the other hand, while the MOSA priorities and technical challenges will vary, MOSA transition is consistently a coordination problem with distributional implications. A key trait of coordination problems is that participants make choices based on what they expect others to do. Failure in one area thus risks breeding skepticism and making success harder elsewhere. While positive or negative results will flow over naturally, deliberate effort is needed to identify examples of high readiness to promulgate. Similarly, as many new programs set MOSA goals, identifying areas where readiness is sorely lacking is best done while time and attention can still be applied to prevent problems.

Coordination problems grow more challenging as the number of actors needed for success grows, thus it is little surprise that the expert workshop discussion repeatedly returned to the point that individual programs often have perverse incentives when it comes to MOSA. Happily, the larger enterprise has multiple ways to act as a coach for the team of program managers. This point cut across multiple dimensions and while addressing it is primarily a matter of the enabling environment, achieving common open standards is a boon to providing appealing business models for the traditional industrial base and potential new prime supplies.

Perhaps the most important aspect of readiness not resolved in this paper is how segmented modules should be. While greater commonality is a sign of more readiness, there is benefit in some parts of the government pushing further with this approach, as this gives an opportunity to put proponents’ views to the test. There is massive variation in the DOD enterprise—that said, this framework looks for common factors, in part because of the risk that MOSA transition could be defeated in detail. Debate should continue and varied approaches may at times allow experimentations or address different circumstances.

Achieving common open standards is a boon to providing appealing business models for the traditional industrial base and potential new prime supplies.

Even if successful, MOSA transition is likely to be unevenly distributed and will need to iterate itself to keep up with evolving technologies, reinforce commonality, and otherwise build on successes and address failures. Many metrics are difficult to apply across the larger enterprise; nonetheless, big picture monitoring, such as these readiness dimensions, is part of addressing the coordination problem. The problems discussed in this piece are real and there are gaps in readiness. However, the time is right for MOSA; it’s worth sustained leadership engagement to build momentum.

Gregory Sanders is deputy director and fellow with the Defense-Industrial Initiatives Group at the Center for Strategic and International Studies in Washington, D.C. Alexander Holderness is a research intern with the CSIS Defense-Industrial Initiatives Group.

The authors are grateful to Leah Matchett, who in her time as a contract researcher with the CSIS Defense-Industrial Initiatives Group greatly contributed to the theoretical framework employed in this paper.

This brief is made possible thanks to expertise, engagement, and support from the Boston Consulting Group.

CSIS Briefs are produced by the Center for Strategic and International Studies (CSIS), a private, tax-exempt institution focusing on international public policy issues. Its research is nonpartisan and nonproprietary. CSIS does not take specific policy positions. Accordingly, all views, positions, and conclusions expressed in this publication should be understood to be solely those of the author(s). 

© 2021 by the Center for Strategic and International Studies. All rights reserved. 

Please consult the PDF for references.

Gregory Sanders
Deputy Director and Fellow, Defense-Industrial Initiatives Group
Alexander Holderness

Alexander Holderness

Former Associate Fellow, Defense-Industrial Initiatives Group