Utility applications should be integrated with an interface based on a canonical data model, not directly with each other

Utilities are successfully using various middleware technologies, including enterprise application integration (EAI), to integrate a variety of systems. However, typical integration efforts have concentrated on the mechanics, leaving semantics to be driven by projects focused on their individual application needs. Without the use of an intermediate canonical data model as a basis for integration, efforts to automate and manage business processes are debilitated by incongruent data from disparate applications. Integration remains complex and slow because data resides in thousands of incompatible formats and cannot be systematically managed, integrated, or cleansed. This paper describes how a model driven integration (MDI) process leverages mainstream technologies and standards like EAI and XML; the only difference is that now applications interface with a semantic layer rather than directly with each other. This decoupling of applications reduces interdependencies among projects, which reduces risk and sets the stage for scalable inter-application business process automation.