A Bayesian approach to model change propagation mechanisms

Abstract

Engineering Changes (EC) are often the answers the designers think of while dealing with new performance targets or customers needs and expectations, regarding functionality, aesthetics, security, etc. Engineering change management (ECM) techniques look to predict or control these consequences within an existing system or product, to limit the generated cost and required efforts to integrate such changes. Engineering change management (ECM) techniques look to predict or control these consequences within an existing system or product, to limit the generated cost and required efforts to integrate such changes. This paper addresses the issue of enhancing the ability of the ”change evaluation” through the suggested technique. We propose a methodology that covers change analysis and evaluation from change’s objective initialization to formulation of recommendation based on system engineering framework (ANSI/EIA-649 1998). We use a modeling technique based on influence diagrams able to integrate different uncertainty levels (on dependencies) in a unique model. It will be shown that the models offer the ability to analyze change impacts but also allow to synthesize the system. Finally, these results can be obtained in a very efficient way which gives the possibility of their use during a design meeting in a practical way at the very beginning of a change project and mainly to support go/no-go decision.