A Multi-agent Approach to Collaborative Design of Modular Products

Abstract

Concurrent Engineering (CE) has become a very popular term while complex product markets are demanding the ability to deliver a high variety of customized products quickly and globally. One of the CE issues has been centered on the way in which the knowledge of the functions such as manufacturing, logistics, marketing, and sales, is captured and managed. To be easily communicated and shared, tacit knowledge needs to be externalized as explicit knowledge, e.g., design rules, and this explicit knowledge incorporated with information technology supports the design decision-making. This paper presents an agent-based system to knowledge management, specifically in design knowledge externalization, and focuses on decision support of modular product collaborative design. The solution approach involves (1) modeling/representing modular products and design knowledge explicitly and (2) supporting collaborative decision-making with agent techniques intelligently. The agent-based system allows the designer to develop a product with the assistance of agents characterizing intelligence, autonomy, and cooperation. The novelty of this paper is that the design operations are formulated and stored in data mart. Therefore, implicit knowledge can be externalized as design rules using the data mining technology. With these rules, design decision-making is supported effectively.