Automated design knowledge capture as an aid for improved decision making and product cost reduction activities

Abstract

This thesis presents the development and evaluation of the automated design knowledge capture methodology that aids design decisions and cost reduction activities.

While a number of methods to capture and represent the design knowledge and also methods to aid the decision making are addressed by existing research, all of them focus on separate product development domains or partially linked domains. Furthermore, the existing methodologies require human intervention in one way or another that interrupts the engineers’ work-flow. Therefore, this research addresses the mentioned gaps by establishing the relational links between the product development domains and capturing the design knowledge in an automated way without interrupting the work-flow.

The proposed methodology is based on the network interdependency model where links between the nodes serve as information carriers allowing the data to be fed in a forward or backward direction. The links between the requirements, design and cost domains capture the relational dependency information as well as the domain data which is then used for calculation and visualisation purposes. Better understanding of the relational complexities and interactions between the domains leads to better decision making. The framework follows the work-flow of the product development life cycle and captures targeted design knowledge in an automated way.

The three case studies, varying in size and complexity, demonstrate a successful application of the methodology. In addition, the calculated requirements impact highlights the requirements that are linked to the high cost features as well as the number of interdependencies between the requirements. The design knowledge map helps to quickly identify the potential cost reduction areas and highlights which design features and requirements might be affected by any changes in the system. Having a holistic view of the relational dependencies between the requirements, design and cost domains together with the associated data enables the cost reduction scenarios to be simulated in a shorter time resulting in the significantly reduced overall cost and trade off analysis time. The evidence from the second and third case studies suggests that the same level of effort is required applying the framework on one component and on the mini sub-system of three components. However, more empirical studies are necessary to confirm that the proposed methodology is scalable when applied to the large sub-systems, such as IP Turbine.

More information

Identifiers

Catalogue record

Export record