Government/Military Trends

FEA Reference Models: Into the Next Phase of Development

Issue Table of Contents

Outsourcing MRO Sourcing of Parts for Aging Aircraft Wiring: A Critical Issue in Aging Aircraft

Federal Enterprise Architecture (FEA), the President’s Management Agenda initiative to transform currently cobbled-together e-Government and IT structures into a smoothly interwoven collaborative network, is primarily made up of five reference models. They are essentially analysis tools that allow the FEA to establish a framework. The first four to have complete iterations were the Business Reference Model (BRM), the Service Component Reference Model (SRM), the Performance Reference Model (PRM), and the Technical Reference Model (TRM). The fifth, the Data Reference Model (DRM), was much anticipated and many months delayed, but it was ultimately released in four volumes in October 2004.

Each of the models represents one member of a component-based architecture and has a precisely defined purpose.

• The BRM, focusing on the business-driven core philosophy of FEA, seeks to clarify a functional (as opposed to organizational) definition of the government’s lines of business. By finding shared business functions and subfunctions across the federal government and listing them in a tiered hierarchy, rather than in an agency-by-agency organizational chart, the BRM attempts to find common operations and enable collaboration among them.
• The SRM is a vertically and horizontally constructed hierarchy of service components classified in accordance with the way they support business and performance objectives. The SRM is broken down into domains including customer services, process automation, business management, digital asset services, business analytical services, back office services, and support services. Its purpose is to discover or create avenues for leveraging IT investments and assets.
• The purpose of the PRM is to measure the performance of IT assets and investments and gauge their contributions to program success. It has a three-part aim: to improve decision-making capabilities through expanded performance information; to align investments and outcomes, creating a clear “line of sight” to intended results; and to identify opportunities to improve performance.
• The TRM offers a scaffold for assembly and categorization of the standards, specifications, and technologies on which the government relies to enable the delivery of services. Its intent is to advance the reuse and standardization of technology and service components across the government. Identifying agency capital investments encourages a standard vocabulary and unifies existing agency models and guidance. These advantages in turn enable collaboration and interoperability.
• The DRM, the Data Reference Model, considered by many experts to be the hinge on which the project turns, will allow agencies to share electronic information. In order for this goal to be realized, the DRM uses three approaches to data—categorization, structure, and exchange. This process will enable agencies to give data uniform descriptions, so that each unit can be easily shared, sorted, used, stored, and retrieved. The category label addresses the subject and type of the data; the structure label lists its form (i.e., integers, plain text, PDF, and so on), and the exchange label describes how the information is transported among agency officials.

Still a work in progress, the reference model suite represents a good starting point, but it also has weaknesses that must be solved in future iterations. Because the models were built one by one, they stand alone but not together, though analytics and data mining to create linkages are underway. Critics also point out that some of the language is too vague to offer useful guidance, which, if left unaddressed, would leave agency officials guessing about their next steps. In its 2006 Action Plan the FEA Program Management Office promises to publish use cases, workshops, and an official registry for data-related services as more specific guidance. A vital but not yet tackled chore for the data reference model is security. Consensus among architecture specialists is that resolution of this vulnerability should not be left for later. A Security and Privacy Profile, when its second phase is complete, will offer help for agencies to incorporate security and privacy requirements into EA development. The structural ideals embodied by the reference model set are sound, but much remains to be done if they are to succeed.

Subscribe to eNewsletter