This thesis addresses an important issue of the common building delivery and optimization process: The generation of building components. While this process is regularly considered to be of high importance for the final quality of a building, it is regarded as a tiring repetitive routine by most planners. Toward this end this contribution addresses the possibility of (semi)automation of (layer wise) building part composition. Thereby, the relationships and interdependencies between different layers are examined to identify typical structures and derive rules for the building component composition. While this task can be performed by human planners given a certain domain knowledge rather easily, the formulation of building composition rules for IT-based automation is far from trivial. A major objective is to (i) collect and to (ii) formalize the required knowledge of building component composition in a way that it can easily be transformed to programmed routines. There are a multitude of interdependencies and relations between the different layers of composition, and the properties of the different layers, as well as the overall composition strongly influence the final performance of a building component. After definition, collection, and structuring of all required information, this knowledge is formalized in a close-to computer-readable format. In the present case, the overall process of component generation is depicted as 'Pseudo-Code', which offers three major advantages: (i) Pseudo-Code is a vendor and platform allowing programmers, who are regularly non-professionals in the building domain, to implement the rules independent, in process able software code; (ii) potential mistakes and issues can be easily identified; (iii) Flexibility, Extensibility and easy editing are assured. The building component generation process was fully modelled for a large number of constructions. Furthermore, certain normative performance indicators of the created constructions were calculated. These indicators, including the U-Values, a condensate evaluation, the thermal storage mass and an environmental footprint indicator, act as decision support to evaluate the appropriateness of a construction for a given, specific purpose. These methods are not taxative; Indeed, later enhancements of the environment would allow to integrate further performance indicators, such as acoustical performance parameters and cost calculation. The results of this work - a formalized modelling language for building component generation - are intended for use in applications, such as building performance simulation. There is a major need in practice for such developments, as can be seen as proven in different related efforts, such as the SEMERGY project or the BAU-WEB project. The opportunity to integrate the developments of this thesis in these or other efforts can be seen as a large (side)benefit of this work.