A study of consistency in design selection and the rank ordering of alternatives using a value driven design approach

Subramanian, Tenkasi
Major Professor
Christina L. Bloebaum
Committee Member
Journal Title
Journal ISSN
Volume Title
Research Projects
Organizational Units
Aerospace Engineering
Organizational Unit
Journal Issue
Aerospace Engineering

In the current day, with the rapid advancement in technology, engineering design is growing in complexity. Nowadays, engineers have to deal with design problems that are large, complex and involving multi-level decision analyses. With the increase in complexity and size of systems, the production and development cost tend to overshoot the allocated budget and resources. This often results in project delays and project cancellation. This is particularly true for aerospace systems. Value Driven Design proves to be means to strengthen the design process and help counter such trends. Value Driven is a novel framework for optimization which puts stakeholder preferences at the forefront of the design process to capture their true preferences to present system alternatives that are consistent the stakeholder’s expectations.

Traditional systems engineering techniques promote communication of stakeholder preferences in the form of requirements which confines the design space by imposing additional constraints on it. This results in a design that does not capture the true preferences of the stakeholder. Value Driven Design provides an alternate approach to design wherein a value function is created that corresponds to the true preferences of the stakeholder. The applicability of VDD broad, but it is imperative to first explore its feasibility to ensure the development of an efficient, robust and elegant system design. The key to understanding the usability of VDD is to investigate the formation, propagation and use of a value function.

This research investigates the use of rank correlation metrics to ensure consistent rank ordering of design alternatives, while investigating the fidelity of the value function. The impact of design uncertainties on rank ordering. A satellite design system consisting of a satellite, ground station and launch vehicle is used to demonstrate the use of the metrics to aid in decision support during the design process.