TR077: Functional Analysis for Facility Engineering Data Modeling using the PArtitioned eNgineering DAta flow model (PANDA)

Our research focuses on developing a general approach to the conceptual modeling of facility engineering data and to the data integration support of many participants, phases, and computer applications. Functional analysis is highly relevant to the research. It is particularly important for representing data and developing databases in facility engineering. The development work described here was motivated by the need for a reference model for functional analysis that supports our research goal. Traditionally, the Data Flow model has been the most popular choice for doing general functional analysis of a variety of processes. It provides the basic concepts of data repository, data flow, activity, and data source or sink. However, analyzing complex facility engineering processes places additional requirements on that model.

In this paper, we propose the PArtitioned eNgineering DAta flow model (abbreviated as PANDA), an extension of the Data Flow model developed to meet those requirements. PANDA supports the concepts needed to analyze facility engineering processes, while adhering as much as possible to the simplicity and ease of use of the Data Flow model. PANDA adds to the original model the concepts of participant, participation, precedence relationship, decision, alternative, interference, subprocess, boundary, data item, data generation, material or product, physical flow, mixed flow and flow network. In addition to its concepts. PANDA has graphical representations that are consistent with those of the original Data Flow model. Moreover, PANDA has a unique architecture that includes three partitions: (1) Participants, (2) Process and (3) Data-Material-Products. With PANDA, the analyst can functionally decompose a process into many hierarchical levels of description. At the detailed level, the data flow diagram is structured according to the three major partitions. Therefore, the diagram is also called die Partitioned Data How diagram, or P-diagram. This architecture helps the analyst organize his or her thinking about a complicated engineering process. It also enables the analyst to produce functional schemata that are highly readable, both conceptually and graphically. PANDA also provides syntactic and semantic rules that govern the way in which the concepts should be used. Several basic schema transformation operations are provided to enable the analyst to develop a functional schema incrementally. To assist us users, PANDA offers a customized methodology that benefits from the model's partitioned architecture and guides the analyst in applying the model to his or her domain. PANDA also provides guidelines for using specific concepts of the model, validating the resulting functional schemata and drawing the Partitioned Data Flow diagrams of those schemata. In the future, we plan to apply PANDA to other facility engineering domains. The experience gained will help us further enhance the model.