Engineering change management in product development : dynamic modeling and applications

產品開發中的工程變更管理 : 動態建模及其應用

Student thesis: Doctoral Thesis

View graph of relations


  • Jianbin ZHAO


Awarding Institution
Award date15 Jul 2008


Engineering Change Management (ECM) is a crucial part of any product development project. It is a dynamic process of knowledge generation and reuse, for products, projects, processes and resources in a company's development and business. However in the literature, the product-oriented change analyses typically omit the time-related dynamic features. Investigations in industrial practice show that project managers demand a systematic way to manage changes in a project's life cycle and development processes. This research models ECM in a dynamic way at two levels. At the project level, a framework is developed to predict changes and understand change distribution in product development projects. At the process level, a genetic algorithm and a simulation tool are developed to generate optimal process plans. This research covers the following: First, three Research and Development (R&D) companies were investigated to find out the industrial needs for ECM and to generate a knowledge-base. These companies vary in their business nature, management style, organizational structure and product type. One company was selected for further investigations on their product characteristics, project features, employees' perspectives and historical change databases. Sixteen crucial factors for ECM, including organizational, strategic, project, market and product categories, were identified and discussed in detail. Second, based on the aforementioned factors and historical data, an exploratory framework for change prediction and distribution in projects was proposed. Detailed procedures were developed to demonstrate and implement the framework, with examples. Third, a Genetic Algorithm-based approach for process planning was developed to facilitate various planning strategies using the an extended Design Structure Matrix (DSM) model. A chromosome of the Genetic Algorithm (GA) was used to represent the sequence and cluster information of tasks. Fitness functions were used to describe the planning strategies, including parallel alignment and phased alignment, to reduce changes and to minimize design effort. Fourth, a simulation tool was developed to evaluate the plans generated by the GA. A fitness function formula was chosen, from among twelve options, to generate optimal plans for both parallel and sequential alignment strategies. The performance of a plan was shown using measures including the mean effort, mean durations, efficiency, change loss, communication loss, review loss, average engineer's workload, etc. The industrial investigations, model building, applications, algorithm and tool development contribute to both the research literature and industrial project management. Using the change prediction and distribution framework, project managers will be able to identify the core factors affecting ECM and make optimal resource allocations and project scheduling. The GA-based DSM analysis makes up for the flaws and insufficiencies of the existing approaches. The simulation has shown that there is a limit of concurrency and phase division, the number of which is subject to product and project features. This approach will act as the foundation of a knowledge-driven ECM approach for industries and will also serve as a good process planning tool for professional solution providers.

    Research areas

  • Production engineering, Organizational change, Product management, New products, Management