Modeling Consulting Services

Principled Design for Complex Enterprises

Expansion of

A proposal to the National Science Foundation
Edward Robertson - PI and Richard A. Martin - co-PI


Executive Summary

Government Involvement

Intellectual Merit

Formalism Factors

Broader Impacts


Executive Summary

This proposal addresses issues of design at the broadest scope - the scope of enterprises encompassing multiple organizations, and of systems where software provides both components for those systems and means for management of enterprise complexity. In particular, it seeks to complete and apply a formalization of enterprise architectural frameworks. To exercise and motivate our work, we have begun to apply it in a most significant venue, the development of international standards for enterprise architectures.

An Enterprise Architecture Framework (EAF) is a conceptual and representational mechanism that organizes a set of artifacts which describes an enterprise at various levels of abstraction and (ideally) links those artifacts in a coherent manner. An EAF should facilitate principled, model-driven specification, design, and implementation of systems whose scope is broad and whose realization is multifaceted.

Return to Top

Government Involvement

The adoption of enterprise architectures is strongly evident in the federal government with the enactment of the Information Technology Management Reform Act (also known as the Clinger-Cohen Act), extended in the E-Government Act of 2002. Correspondingly, the Department of Defense has defined "a coordinated approach (i.e. a framework) for Command, Control, Communications, Computers, Intelligence, Surveillance, and Reconnaissance (C4ISR) architecture development, presentation, and integration."

Addressing discontinuities that occur between, rather than within, organizations, the ISO (International Standards Organization) and the CEN (European Committee for Standardization) are promulgating enterprise architecture standards. These standards, ISO 15704 (Requirements for Enterprise Reference Architectures and Methodologies) and ISO FDIS 19439 (CIM Systems Architecture -- Framework for Enterprise Modeling) are prompted by the growing dependence of industry on electronic data interchange and the realization that interchange standards must cover much more than just how data is encoded. They provide "a common conceptual high-level framework within which key concepts of the (distributed, extended, virtual, etc.) enterprise can be identified, documented, and shared with partners."

Return to Top

Intellectual Merit

Our goal extends beyond merely developing architectures: our goal emphasizes formalization of frameworks for architectures. In order to advance enterprise architecture beyond the heuristics of "best practice" into the realm of engineered complexity management, we identify mechanizable conceptual principles, transform those concepts into usable forms, and apply those forms in support of enterprise operations. Perhaps the most unique aspect of our proposal is the use of the ISO process to gather input and vet contributions.

We are already engaged in a formalization of enterprise architectures, guided by principles derived from "best practice" and also from the mistakes of "poor practice." These principles identify characteristics that enable enterprise-wide management of model artifacts in a manner that endures as the enterprise context changes. Adherence to these principles across many domains will further our ability to understand inter-domain interactions and relationships which are becoming ever more critical characteristics of global economies and societies.

Return to Top

Formalism Factors

There are four factors to our formalism: structure, connections, views, and constraints. Tasks to be performed as part of this project include: (1) continued investigation of fundamental issues and development of formal mechanisms addressing these issues, (2) validation of concepts and formalism through case studies -- studies also targeted as expositional organs, (3) developing "wrappers" that facilitate practitioners' application of the formalisms, and (4) dissemination of the result of our efforts. Furthermore, an important and unusual task under this project, cross-cutting case studies, usability considerations, and dissemination, is (5) participation in international standards bodies.

To aid in the examination and presentation of the architectural formalism, we are creating a standard scenario that has just enough complexity to demonstrate the methods that our formalism enables. We have chosen a supply chain procurement process that nests purchases along the chain in a recursive manner. In this way we can present in one scenario all the architectural principles we ascribe and a corresponding formal description. This exemplar is critical to the conveyance of both the principles and formalism to others. Beyond intellectual accessibility, it is essential that the framework formalism be operationally accessible.

Return to Top

Broader Impacts

In the applications realm, another significant audience is professionals designing systems beyond strictly software systems. Participation in standards development activities and international standards bodies thus comprises an essential part of this proposal. The PIs have most recently participated with standards efforts in meetings and exchanges with WG1 of ISO TC184 SC5. In particular, Richard Martin is convener of WG1. Our participation as invited experts enabled us to discuss our framework formalization with educators and practitioners in the automated manufacturing domain which is the charter for ISO TC184. In addition, Edward Robertson uses an enterprise framework approach to facilitate a year-long course "Software Engineering for Information Systems."

Return to Top