CUED Publications database

Integrating value assessment into the computational engineering design cycle

Kipouros, T and Isaksson, O (2014) Integrating value assessment into the computational engineering design cycle. In: UNSPECIFIED pp. 2446-2455..

Full text not available from this repository.


In modern engineering, the application of computational design techniques is common practice at all phases of the overall design process of a product. Most often, these design systems are configured to identify the appropriate values of settings, or design parameters, that improve the performance of the product. However, very often, there are important aspects that can characterize the product in a qualitative way, for example, the comfort or the experience of the passengers in an aircraft. These 'metrics' can express the value of the product which is equally important as the technical performance. The development teams wish to include such metrics in their design process in order to develop a successful product as perceived by stakeholders such as the end customer. Evidently, there is a tradeoff to be made between quantitative performance characteristics and other - less quantifiable value assessment criteria. In this paper, we explore and expand an established Value Driven Design methodology for an aerospace problem. First Value Dimensions and Value Drivers are identified and somehow quantified and then we proceed with the visualization in Parallel Coordinates that enables the analysis and understanding of the conflicting trends between the Value metrics and the technical performance metrics, as well as hidden correlations between these metrics. Ultimately, we want to identify and formalize these relationships between the Value Dimensions and the Value Drivers for a given scenario. Hence, we can then proceed with the formulation of the design problem as a multi-objective, or multi-disciplinary optimization problem that will allow the designers to identify new solutions for a product that synthesize and balance desirable characteristics not only for a range of different disciplines, but also between quantitative and qualitative valuable properties. The steps of the proposed method are presented in detail that lead to the development of Value Driven Optimization capabilities. This is an iterative process that requires the interaction with the human designer which has a positive effect on understanding the design problem, establishing balanced requirements and allowing value for a broader range of stakeholders to be introduced in a computational engineering process.

Item Type: Conference or Workshop Item (UNSPECIFIED)
Divisions: Div C > Engineering Design
Depositing User: Cron Job
Date Deposited: 17 Jul 2017 19:40
Last Modified: 16 Aug 2018 02:03