Semantic BIM Design Methodology for Energy Efficient Building

The purpose of this thesis is to improve the early stage design process, within computer aided design, allowing more design options to be compared and anticipating feedback on choices that influence the building's energy consumption. To reach this purpose it analyses the role of the human designer and the encountered limits of the computer aided design process. The research opens on the state of the art by investigating relevant open standards in Building Information Modelling (BIM) and semantic web, design automation in layout generation, early stage energy simulation and evidence based design. As a practical application, the author took a layout generating tool developed within Streamer, the European applied research project he collaborated on. A semantic BIM design method using open standards for early stage design is described, the main building typology referenced is hospitals, suitable for their strict functional requirements and high energy consumption. Within the method’s workflow, the design process starts from a classification of Programme of Requirement data, which is then used to generate alternative design layouts. The generated proposals are evaluated by an evolutionary algorithm, based on pre-defined prioritised design rules. Several alternative designs can be exported as Industry Foundation Class (IFC) files, in which properties can be retained for downstream use. Optionally, to improve placement of building envelope elements & building services, an expert can create zones manually within the file. An automated energy calculation uses the semantic data contained in the IFC file. A decision support system is used to compare and evaluate the generated options. Both designs for new buildings and the use in a retrofitting scenario are considered. To close the inquiry a speculative investigation of other building typologies attempts to understands the limits of the automated process and tools, and the creative contribution of the human designer. It seeks to describe which adaptations are required to design other building typologies based on concrete examples of contemporary design. The philosophy of the method and its strengths and limitations are discussed and topics for future research are outlined. In conclusion, the author proposes an iterated interaction between designer and brief to combine the strengths of the human designer and the semantic design method.