An agent-based approach for collecting and utilizing design information throughout the product life cycle

Abstract

The study proposes a novel approach to systematically collecting and utilizing product feedback information throughout the product life cycle to support life cycle design methodologies. The approach has a solid theoretical basis that is the theory of complex system evolution and, at the same time, is technologically sound and does not require significant investments to be realized. The natural evolution of a design as an information object is discussed in the context of the product life cycle, and the importance of evaluation of design requirements in post-design life cycle stages is argued. The notion of design expectations is introduced. A prognosis technique based on a parametric model is outlined that allows one to estimate product characteristics and predict violations of design expectations in dynamic environments. An agent based approach is proposed to evaluate the production and product environments, assess design expectations, and supply the designer with relevant information to improve the product design. A case study of the development of an expectation agent for a refrigerator is described. Summarizing the experience accumulated through the research, some conclusions and implications for future work are given.