Search

This study proposes a new set-based design approach for preliminary engineering design that intrinsically contains various sources of uncertainties. The goal is to achieve design flexibility and robustness while capturing designer's preference, The proposed design approach includes three computational methods: (1) set representation method to specify the varying degree of desirability of a ranged set of design solutions and performance requirements, thereby enabling the manipulation of uncertain design solutions and requirements based on designer's preference structure; (2) set propagation method to obtain performance possibilities achievable by uncertain design solutions, thus exploding a broader design possibilities; (3) set narrowing method to generate a ranged set of feasible solutions (i.e., robust and flexible solution set) instead of single point solution that satisfies changing sets of performance requirements by eliminating infeasible and inferior subsets of solutions, thus allowing designs to be readily adapted to changing conditions. Finally, the proposed design approach is illustrated with a successful implementation of real industrial design problem (i.e., vehicle side-door structure design) in the simulation-based design environment. [ABSTRACT FROM AUTHOR]

The early phase of design intrinsically contains multiple sources of uncertainties in describing the design, and nevertheless the decision-making process at this phase exerts a critical effect upon drawing a successful design. Recently, set-based concurrent engineering (SBCE) has been attracting public attention as an emerging CE paradigm. Representing and communicating set-valued information are useful in handling engineering uncertainties and efficient for the collaboration between designers in the CE environment. Even if we can readily know that SBCE is efficient and powerful enough for large-scale design and management, the existing literature does not give the details of methods to achieve SBCE principles. We have developed a computer-based framework, called 'preference set-based design' (PSD), for implementing SBCE principles, which consists of set representation, propagation, modification, narrowing, and aggregation methods. This article focuses on the set representation and aggregation methods that play an important role in the CE environment. This article proposes new engineering quantities that can capture the designer's preference structure on both the continuous and the discrete sets. A new aggregation method is presented to aggregate the proposed engineering quantities. A comparative analysis with the existing representation and aggregation techniques is also provided. [ABSTRACT FROM AUTHOR]

Concurrent engineering (CE) has been widely accepted as an effective engineering practice for decreasing product development time, improving quality, decreasing manufacturing costs, and improving customer satisfaction. Since CE philosophy is intended to consider all elements of product life cycle from the outset, CE approach increases the complexity of design problem and makes it more difficult to manage. In addition, the advances in computer networks and information technology have brought engineering design into a new era. Although a considerable number of studies have been made on design decomposition (e.g., product and process (or perspective) decomposition) as a means to reduce the complexity of a large scale design problem, relatively little attention has been given to the computational framework for dynamic and systematic design integration in a computer network-oriented design environment. This paper describes an integrated product and process modeling (IPPM) framework for collaborative product design through the Internet. Here, the hierarchical and heterarchical dependencies between the decomposed smaller design problems are addressed. An agent-based approach is proposed for the computer support of hierarchical and heterarchical design topologies. In our framework, a smaller design problem is dynamically modeled in terms of interacting and pluggable design agents that represent a specific aspect of the design problem and provide the interoperability between heterogeneous design agents or software applications, and the results of these smaller problems are recursively integrated to formulate solutions to the larger problems in a systematic manner. Finally, the proposed framework forms a global design network through which a number of design possibilities are explored. [ABSTRACT FROM AUTHOR]