Description: This work presents research on the use of multi-agent systems (MAS) in architectural design with the goal of both improving and automating the design process as well as the design outcomes in terms of geometry and environmental performance. The work is based upon the research on agent based modelling and simulation techniques (ABMS), and the combination of analytical and optimization methods with geometric modelling and user centered design. The motivation is manifold and includes the exploration of the generative aspect of bottom up design techniques coupled with optimization and performance criteria. The objective of the work is to develop and test a design methodology, which enables designers to explore larger sets of informed solutions by coupling design requirements and generative design techniques with validated simulations (i.e., environmental analysis) and user light preferences early in the design process.

The development of the proposed methodology and the implementation of the design system for a building façade are described, while the robustness and validity of the framework is initially tested by applying it to the generation and optimization of façade panel configurations of an office building. The results show that the MAS for design system can generate unique design configurations that perform environmentally (daylight factor analysis) better than a façade with normative façade shading and can be further enhanced through integrating user preferences.