SUPPORTING THE MANAGEMENT OF THE ENGINEERING CHANGE PROCESS THROUGH A CROSS-DOMAIN TRACEABILITY MODEL
Author: Ahmad, Naveed
Supervisor: Clarkson, P. John
Institution: University of Cambridge
Engineering change originating from rework or from changed requirements is a significant part of any product development programme. These changes can arise at many points throughout the product life-cycle, resulting in change processes which can ripple through different stages of the design process. Managing these processes is thus a critical aspect of design projects. Through a literature review and industrial perspective, this dissertation will first show that there is currently little support to help designers and managers to evaluate proposed changes in terms of the total process cost of implementing them and executing any knock-on rework which may occur. It also shows the diversity of information models, from different domains of design process, used by different change management methods. The influence of different domains on managing change processes is also discussed and requirements for an approach to manage changes throughout the design process are established through a case study. I argue that such support would be useful for change process management since there are often multiple ways to implement a requirement change, and it is thus important to identify the most suitable option prior to beginning its implementation. The thesis then presents a framework to create a model which captures the four domains of requirements, functions, components/subsystems and detailed design process and subsequently shows how the resulting models could be used to make cross-domain models by creating the AutoBell (product of Digital Research Labs) model. Elements are linked within and across these four domains via a systematic approach, which allows representation of key aspects of designers’ knowledge regarding the change process. A laboratory experiment is also conducted to evaluate the framework for structuring information where modellers were asked to model using this framework. This experiment provides interesting insights into the thinking of designers and the effectiveness of the framework for modelling products. In particular, the dissertation shows how changes in requirements can be viewed as propagating through these four layers to cause rework in the design process, and how a traceability approach based on the model can be used to help reason about the cost of implementing a given requirement change. This technique is implemented in a software tool which is applied to the management change cases of the AutoBell. The change management support approach is separately evaluated through a series of student experiments which were divided into two groups to test the usability and utility of the approach and the tool. This dissertation also reflects on the importance of using change management methods with the project management methods. A simulation approach is developed which is used with the Gantt charts to devise a schedule for managing changes through the project. The initial results of the simulation experiments show that a schedule can be made where by packaging change requests and executing them after a fixed interval will reduce the delay in project lead time. Finally the main findings of the research, limitations of the proposed approach, and opportunities for further work are highlighted.