Analyzing the Systems Underlying an Enterprise

Abstract

Enterprises are complex systems. As such, they should lend themselves to improvement via the application of systems engineering principles and approaches. To do so, we must first identify the system or systems underlying an enterprise. Four key systems that tend to receive attention are the product, process, organization, and information systems. Product systems are the valued outputs or results produced by the enterprise, such as aircraft, buildings, computers, consulting, etc. Today, the design of complex product systems is the focus of most systems engineering applications. Process systems are the networks of work packages (activities) and decisions that produce a product system. Process systems are also referred to as value streams, and when the frame of reference becomes multiple companies, the terms supply chain and value chain are often heard. Organization systems are the structures of companies, teams, departments, and people that execute process systems. Information systems provide the infrastructure in which much of the data regarding the product and process systems are managed. Surely there are other important systems or systems views within an enterprise, but most enterprises do not have a satisfactory understanding of these four, basic, critical ones. Enterprises are faced with the problems of describing, connecting, organizing, visualizing, understanding, managing, and improving these four systems. Usually, an enterprise will have multiple instances of each of these systems, compounding the challenge. Furthermore, the elements of each of these systems are related to elements of the others. An intelligent enterprise will address, learn about, understand, and leverage these systems and their relationships for competitive advantage. Of the four systems, organizations and processes each provide a view of the system underlying an enterprise, because the structure of an enterprise can be described based on its organizations or its processes. The organizational view is more traditional, relying upon organization charts, descriptions of roles and responsibilities, team structures, boards and committees, etc. However, these descriptions do not adequately describe the organization system architecture, since relationships between organizational elements are often ignored. Recently, many enterprises are discovering the advantages of the process view, which tends to align better with the product system, and thus with customers. The process view, since it describes work packages that must be done and work products that must be produced, also ties to the product, organization, and information systems in a relatively straightforward way. Thus, the process view provides a point of integration for the four systems: activities generate results while requiring people and tools. In accounting circles, the process view is akin to the activity-based view (e.g., activity-based costing). Activities (process elements) and their results seem to be powerful kernels for modeling and integrating a baseline system view. The concept of a system architecture has proved helpful in describing systems. A system architecture includes the elements of the system and their relationships. Architecture modeling (where a model can be thought of as a systematic and disciplined, albeit abstract, description) has helped with visualizing and understanding systems. A classic approach to modeling a system architecture includes: (1) decomposing the system into subsystems, (2) defining the attributes of each subsystem, (3) noting and verifying the relationships between (the integration of) the subsystems that give rise to the system’s behavior, and (4) noting and verifying the relationships with external systems and their impact on the system. The resulting model improves understanding of the system and provides the basis for insightful analyses. However, the size and complexity of enterprise systems makes modeling them especially challenging. This article presents a technique, the design structure matrix (DSM), for modeling, representing, and analyzing at least three of the aforementioned enterprise systems—products, processes, and organizations. This article will focus on applying the DSM to modeling and analyzing a process system architecture, referring the reader to the references for information on other applications.