Personalized assistant for conceptual structural design

At the conceptual design stage, design process is not always logical because of lack of design criterion, procedure and knowledge clearly established. This process is dependent on the individual designer with specific design experience. It is important to consider the designer-dependent subjective aspects. In this article, we discuss an artificial design assistant for conceptual structural design, and its personalization to individual human designer. In this framework, the rule-based inference is used to generate and to evaluate the design candidates from multiple criteria to be considered by the human designer. The genetic operations combined with the genetic case-base are used to suggest the candidate to the human designer. The progress of the design process is managed by the human designer through the interaction with the artificial design assistant. The preference of the artificial design assistant has been adjusted to meet that of the human designer in terms of the parameters for the genetic operations. Proposed idea is implemented and examined for the conceptual design of skeletal bridge structure. Case studies demonstrate the effectiveness of the artificial design assistant with the personalization capability.