Using a matrix-based approach to model change propagation

Abstract

It is common to develop new products through engineering changes (Jarratt et al., 2005). However, such changes can propagate undesirably (Eckert et al., 2004) especially for complex technical systems (e.g. Giffin et al., 2007). It is thus important to model and predict how engineering changes can propagate in order to better manage the design process of complex products. To address this, various authors have developed matrix-based approaches to model the dependencies between different engineering domains. For instance, Mocko et al. (2007) describe the use of a matrix-based modelling scheme which analyses the inter-relationships between engineering domains such as system requirements, functions, components, engineering characteristics, and test. In this paper, we extend the analysis by presenting a method which explicitly considers the effect of engineering change propagation. The goal of the method is to direct engineers to critical change propagation paths early and support engineers in the assessment of solution alternatives during preliminary design. Such assessment is analogous to the concept selection phase as described by Ulrich and Eppinger (1995) except the evaluation criteria are broader and considers both the product and the organisation.