| Product development is concerned with the design and Concurrent Engineering of products to reduce costs, lead-times and improve quality. By contrast, product deployment is concerned with the management of the supply chain to ensure a rapid and low-cost response to customer demands. The importance of these areas has been raised by the development of communication, primarily Internet-based, that offers the promise of a seamless Product Development and Deployment System (PDDS), where both the products and the supply chain for delivering the products can be simultaneously designed. However, PDDS currently lacks a formal methodology for system design. This thesis aims to address this deficiency by proposing a network-based approach, called Trans-Nets, that provides a unified framework applicable to model a PDDS and gives some insight into it behavior. Furthermore, Trans-Nets can be extended to address some important research issues in PDDS.; The basis of Trans-Nets for PDDS model and analysis is detailed. By applying a Constraint-Based Genetic Algorithm (CBGA), Extended Trans-Nets is proposed to improve PDDS design. By applying Interval Arithmetic, I-Trans-Nets is proposed to analyze PDDS design problems with imprecise data. By applying Extended Kalman Filter, Augmented Trans-Nets is proposed to achieve better PDDS estimation with incomplete or error data. A test bed system in an industry case study is described.; There are several contributions in this thesis. First, a formalism for removing many of the complexities, while retaining the ability to analyze the important characteristics of the system; its graphical approach allows users to examine the interaction between different considerations and it applies a mathematical mechanism to the problem, with the potential of being relatively straightforward to computerize. Second, Extended Trans-Nets allows PDDS to be modeled with alternatives presented and system improvements can be automatically identified. Third, I-Trans-Nets allows PDDS design problems with imprecise data to be addressed. Fourth, Augmented Trans-Nets provides a better estimation of the state of PDDS. Fifth, a test bed implementation system can validate the modeling approach for real world problems. |
