An experimental study on the effects of a computational design tool on concept generation

We have developed a computational design tool to help designers create conceptual solutions to detailed functional specifications. The computational method extracts design knowledge from an expanding online design library in the form of procedural rules, and provides these rules as the building blocks for solving new problems. In this paper, we study how this automated approach would benefit designers during concept generation. Accordingly, we test the effects of using our computational tool as an aid for concept generation in an experiment mimicking real design scenarios. Three metrics (completeness, novelty and variety) are used to evaluate the solutions generated to two separate design problems in order to determine how effective the computational method outputs are in improving conceptual design generation.