5.2.4 The CAESAR Framework

The CAESAR framework (Christiansen and Thomsen, 1994) was developed as a part of the CAESAR Offshore project, focusing on more cost effective development of small petroleum fields located at the Norwegian continental shelf. Enterprise modeling was considered an important prerequisite for succeeding with new, improved organizations and work processes.

The purpose of enterprise modeling according to the CAESAR framework is to (ibid.:4)

"enhance understanding of probable effects of proposed or anticipated changes in project requirements and execution."
Understanding is mainly assumed to be gained through the use of simulation. The type of work is routine engineering work. Hence, even if the work being modeled is design, the focus is on design as routine work, and the framework focuses on work associated with replication risk, and not design risk.

The CAESAR framework is illustrated in figure 5.4. The figure describes dimensions or aspects that are assumed to be present in all engineering projects and thus have to be modeled in an integrated enterprise model. The objectives of a project is decomposed through functional decomposition into a set of requirements to a product (desired solution). The product is assumed to be produced through a work process, a sequence of activities that are ordered and controlled according to subactivity dependencies. The activities are performed by an organization, comprised of actors that are responsible for the various activities. Actors must communicate in various ways to be able to fulfill their responsibilities. The outcome of actual project execution is a realized solution, hopefully matching the desired solution. The aspects in this model are denoted Objective, Product, Process, and Organization, giving rise to the acronym OPPO.

Figure 5.4: The CAESAR enterprise modeling framework (Christiansen and Thomsen, 1994:7)

An interesting extension to the basic CAESAR framework is the work of Lie (1994), extending the organizational aspect with concepts based on different perspectives (e.g., taking a structural, political, anarchist, ecological or strategic perspective on the organization). Each perspective included an ontology of concepts, properties and relationships. Lie used OMT (Rumbaugh et al., 1991) as a language and method for enterprise modeling within the CAESAR framework.

The enterprise modeling process is not discussed in the same detail as model features, but an overview of the process can be found in figure 5.5. Given a set of objectives and a desired general model, one decomposes this model into constituents (being in effect a classification process). A type structure is then instantiated to create models of the real world phenomenon in question. Instances are then recomposed to form an aggregate model for further analysis.

Figure 5.5: The Caesar enterprise modeling methodology (Christiansen and Thomsen, 1994:14)

The CAESAR modeling method is too abstract to be of immediate practical use in enterprise modeling. The steps you go through in modeling must be specialized to be practically applicable and tailored to the language of the enterprise that is to be modeled. It is rather to be seen as a framework.

The CAESAR enterprise modeling approach does not discuss philosophical foundations and it is difficult to assess their bias. However, the modeling process is not considered problematic (e.g., decomposition of an overall objective is definitely not straight forward if you do not know the solutions). Sense-making is not discussed. Hence, the CAESAR framework is assessed to be dominantly objectivistic.