Manufacturing companies are confronted with short product life cycles, more variety of products and short cycles of leap innovations. This results in a higher frequency of changes in factory structures and an increasing importance of factory planning processes. Factory planning processes are characterized by participative and interdisciplinary processes due to various actors dealing in different domains and working in distributed environments. The result is a heterogeneous IT-landscape based on increasing use of multiple isolated and domain-specific IT tools and systems and hereby an increasing redundant, inhomogeneous and inconsistent data-holding. The control of these factory planning processes can be reached by holistic approaches and consistent system integration. The mean of system integration is the consideration of all domains involved in planning processes, used IT tools and systems and business processes. In this paper, the approach of a Federative Factory Data Management (FFDM) based on Service Oriented Architecture (SOA) and Semantic Model funded by the DFG (Germany) and CAPES (Brazil) will be described, which faces up the described challenges of factory planning processes. The focus of this approach is on the integration of isolated used IT tools for the dimensioning and structuring of factory systems, the generated domain-specific partial models as well as the coordination and synchronization of engineering workflows. In order of the control of factory planning processes the integration and coupling of the views of products, processes and resources on metadata level is required for a communication between different isolated and domain-specific IT tools of the various involved domains without losses or redundancies. The integration and coupling of these three views is based on a document independent factory structure description linked with factory defining metadata. In order to integrate and couple these different views, the relevant information and independencies are identified. Current reference process models for production and factory planning as well as the current methods to describe domain-specific models are analyzed. This is the basis for the development of the FFDM to build up a semantically coherent information model as a common communication and integration framework to represent the factory and to define and to access factory data. The goal of the presented approach is the increasing of planning harmonization, certainty, quality and frequency by a consistent information flow as well as the reduction of time of product development and factory planning processes.