Spending patterns show little movement toward fully shared enterprise services.
Network-centric operations as a doctrine based on a shared U.S. Defense Department infrastructure is now 10 years old. The concept of networked warfare has been under active consideration by the Office of the Secretary of Defense since 2001, but no evidence suggests that significant infrastructure consolidations are taking place. As yet, no new major systems have been built on a shared infrastructure platform. The fiscal year 2009 budgets show no significant reallocations of funds to propel network centricity into reality any time soon.
The proposed fiscal year 2009 information technology budget does not convey a sense of urgency to fund more rapid migration to network-centric systems. Commercial information technologies are now progressing faster than the Defense Department's capabilities. To keep the services and agencies from falling further behind the accelerating pace of technology, the department's components must proceed with expedited conversions to deliver lower-cost, interoperable, secure, real-time networks in the next three to seven years.
The Defense Department's fiscal year information technology budget is projected at $33 billion, which represents a spending reduction of 3.9 percent compared to fiscal year 2007. The Air Force, Army and Navy budgets have been reduced by 9.5 percent, whereas the agencies' information technology budgets have increased by 9 percent. The money for network-centric projects is now almost entirely within the agencies and particularly within the Defense Information Systems Agency (DISA). An examination of fiscal year 2009 plans can yield clues as to whether the Defense Department is moving closer to its proclaimed goal of network-centric systems.
The dominant share of fiscal year 2009 information technology spending is with agencies accounting for one-third of total Defense Department spending (34 percent). The other two-thirds are with the Air Force (21 percent); Army (23 percent); and the Navy (21 percent). For delivering results, what matters most is the allocation of projects across the entire Defense Department.
The Defense Department's total information technology budget is spent mostly on its infrastructure (46 percent) and not on warfighters (35 percent). Except for logistics (7 percent), the business and medical systems account for only a small share of total spending. The largest cost component is for the communication infrastructure (28 percent), involving 172 communications projects that account for $9.14 billion of spending.
The communications infrastructure accounts for the largest share of spending. Whether or not the Defense Department will be network centric will require interoperability among the 172 communication projects. The 2009 budget gives no indication that a merger in communications is taking place among the various projects underway.
Communication projects continue as independent engagements. The total infrastructure consists of 578 diverse applications costing $14.94 billion. Most of these applications do not share a common data dictionary and therefore are not interoperable. This diversity is the consequence of networks that are contractor defined and not integrated, such as the largest single infrastructure investment, the Navy/Marine Corps Intranet (NMCI) project. The top 20 infrastructure projects, accounting for about half of total infrastructure costs, are operating separately by services or agencies.
The Defense Department's infrastructure is diversified and not shared. There may be department enterprise-oriented investments already in place as byproducts of fiscal year 2009 projects. However, it is not clear how this will materially diminish the diversity of 578 infrastructure projects slated for convergence into a shared environment. There is no indication how fast the services and agencies will install replacement systems that are interoperable and that utilize shared services. Next year's budgets do not reveal how rapidly the Defense Department will be able to transform its $14.9 billion worth of incompatible infrastructure-at a speed that matches current progress in commercial technologies-into a more coherent environment.
To better understand future directions, the term "shared infrastructure" requires an explanation. The fiscal year 2009 Defense Department budget submission to the Office of Management and Budget identified 2,193 information technology projects (Air Force, 466; Army, 216; Navy, 973; agencies, 538). Of the total of $33 billion, $11.9 billion is for development and $21 billion is for operations and maintenance. An estimated seven-year lifespan for Defense Department software reveals the software's cumulative value to be near $70 billion, which keeps consuming double that amount in maintenance. What would be the business case for the department to migrate toward a shared information infrastructure?
Each of the 2,193 Defense Department projects contains lines of application code, data management instructions, telecommunications interfaces, security routines and maintenance procedures. A shared infrastructure could deliver a common support service that would only require writing the application code. Such an infrastructure permits the pooling of operations and maintenance costs, instead of spending separately on each project.
A stand-alone information technology project, especially one that is contractor managed (as is true in two-thirds of Defense Department cases), involves numerous elements. These include requirements, prototyping, architecture, project planning, initial design, detail design, design reviews, coding, code inspections, independent verification and validation, configuration management, integration, user documentation, unit testing, function testing, integration testing, system testing, field testing, acceptance testing, quality assurance, installation/training and project management. According to Estimating Software Costs by Capers Jones, if a shared infrastructure can relieve a project manager of data management, telecommunications, security and most ongoing maintenance, the project manager would be left with less than 20 percent of costs. More important, providing a service-oriented architecture (SOA) infrastructure would simplify introduction of innovative new applications while relieving the Defense Department budget of operations and maintenance expenses.
How much the SOA infrastructure would reduce project costs currently is unknown. With huge costs associated with 2,193 projects, a shared infrastructure could offer significant price reductions while accelerating progress to an environment that looks more like Google than NMCI.
DISA's Net-Centric Enterprise Services (NCES) project has a development budget of $37.2 million and an operating budget of $89.9 million for fiscal year 2009. The objective of the NCES is to "... provide a secure, collaborative information sharing environment and to deliver enterprise capabilities in all DoD [Defense Department] domains." The declared purpose of the NCES is to enable information sharing by "...connecting people and systems that have information for those who need information."
The scope of the NCES includes mediation (across computer services), messaging, content delivery, collaboration, people discovery and security. All of these services are clustered around a central core of portal and metadata services. The NCES is, in effect, an intermediary between hundreds of existing agency and service systems that continue to remain in place. The NCES is a SOA approach to negotiate exchanges of transactions between individual projects. It is not a shared infrastructure on which complete applications could be built. For example, the NCES adopted Army Knowledge Online and made it available as Defense Knowledge Online. From the standpoint of advancing the Defense Department toward net-centric capabilities, the 2009 development budget of $37.2 million represents a modest investment to advance progress toward an integrated future.
The Net-Enabled Command Capability (NECC) is the Defense Department's principal command and control capability of the future. It will be accessible in a net-centric environment and will provide commanders with data and information to make informed decisions. The NECC provides a fully integrated, interoperable, collaborative solution. It will be founded on a single, net-centric, services-based architecture.
The greatest challenge in migrating to the NECC is converting the existing Global Command and Control System (GCCS) to a common architecture. So far, much of the development money ($203.6 million in fiscal year 2009) is still with service-specific projects and requires large operations and maintenance costs ($261.3 million).
The fiscal year 2009 project budgets do not indicate how fast the services and agencies currently operating their GCCS versions will migrate to a joint environment. No funds are allocated for a project designated as "NECC" in the proposed 2009 budget.
Proliferation of Defense Department infrastructure investments persists as a legacy of Cold War project acquisition practices that favored isolated and completely self-sufficient investments. The fiscal year 2009 budgets do not reveal a material redirection of this approach.
Paul A. Strassmann is a distinguished professor of information sciences at George Mason University's School of Information Technology and Engineering. He is the former director of defense information at the Office of the Secretary of Defense.
U.S. government budget for fiscal year 2009: http://www.whitehouse.gov/omb/budget/fy2009/