An empirical foundation for design patterns in parametric design

This paper presents the results from exploring the impact of using a parametric design tool on designers' behaviour in terms of using design patterns in the early conceptual development stage. It is based on an empirical cognitive study in which eight architectural designers were asked to complete two architectural design tasks with similar complexity respectively in a parametric design environment (PDE) and a Geometric modelling environment (GME). Protocol analysis was employed to study the designers' behaviour. To explore the development of design patterns during the design process, we utilise the technique of Markov model analysis. Through Markov models analysis of the PDE and GME results, we found that there are significantly more Function to Structure transitions in PDE than in GME. During this transition process, designers select an existing structure/ solution for the particular function/design problem based on their experience or knowledge, which is a process of applying an existing design pattern to the problem. From this result we can infer that when architects apply programming and scripting in their design, such as in a PDE, they exhibit the characteristic of using design patterns.