An Information Integration Theory and Reference Model

An information and decision theoretic model of integration identifying information requirements for the major manufacturing functions answers these questions . Such a model is developed using a paradigm of parallel formulation. The model unites previous results in several major disciplines to uncover basic tasks, logical interactions, and required data and knowledge classes.

The new paradigm guides a new critical review of existing process planning, production planning and shop floor control functions. This review reveals the highly interdependent nature of the tasks whose consolidation characterizes integration and uncovers information requirements expressed or implied by the unit tasks in these domains.

Based on an analysis of the logic and the information flows among the unit tasks, a model using parallel reformulation is prescribed to consolidate interdependency, effect a global decision space, and reduce uncertainty for integrated manufacturing planning and control. The characterizing information requirements of the model are determined accordingly.

These results are unified into a theory comprised of definitions, assertions and arguments. The definitions and some of the assertions are derived from existing production, organization, and information theory and from the analyses in the thesis; some assertions are built up from other assertions and specify the nature of integration; and requirements for information classes and knowledge classes are given as assertions.

Finally, information requirements derived from the theory are expressed formally using the Two-Stage Entity Relationship (TSER) methodology to provide a reference model for applications. This model includes the tasks, interactions, data classes, and knowledge classes required for achieving integration in practice.

By using a formal representation method to develop the information requirements model, the resulting reference model serves as a starting point for enterprise-specific information modeling projects. The reference model can also be used to evaluate the extent to which other models satisfy the information requirements for integrated planning and control.