Preference-Centric Design Methodology: Explicit Quantitative Measures for Engineering Design Evaluation and Selection

Abstract

Large-Scale Complex Engineered System (LSCES), such as those within the space industry, often have significant challenges to overcome. These include consequential cost and risk, substantial design cycles, extensive operational timelines, complex management structures, and large geographically distributed organisations. National Aeronautics and Space Administration (NASA) developed Systems Engineering to overcome these challenges. However, the exercise of Systems Engineering has consistently resulted in severe consequences for the stakeholders in a market. These have included reducing the overall timely benefits or usefulness of the LSCES due to increasing development time and cost. This research indentifies and traces these consequences back to the initial stages of the Systems Life Cycle (SLC), where Systems Engineers decide on the ‘best’ engineering design. Systems Engineering commonly utilises Requirements-Based Design (RBD) as the engineering design methodology within the SLC. Systems Engineers use requirements as a proxy for stakeholder preferences over the attributes. However, requirements tend to deviate from the preference ranking of attributes due to a translation process involved. As a result, Decision-Based Design (DBD) has been advocated as a replacement engineering design methodology, as it can exactly represent stakeholder preferences over attributes. Two prominent and potential DBD engineering design methodologies are Utility-Centric Design (UCD) and Value-Centric Design (VCD). VCD creates an objective ordinal ranking for plausible outcomes of engineering design alternatives under worth preferences. UCD creates a subjective ordinal ranking for plausible outcomes under risk preferences. However, neither VCD or UCD considers the ordinal ranking of plausible outcomes under time preferences. The distinct treatment of the worth, time, and risk preferences have not been explored in any engineering design context. This research developed a rigorous engineering design methodology and a framework for evaluating and selecting the best engineering design in the concept stage of the SLC. The researcher introduces the notion of Engineering Design Environment (EDE) to significantly increase the effectiveness and efficiency of the design methodology involved in the framework. This required the development of a new engineering design paradigm called Information Thinking. This paradigm regulates and governs the process of obtaining information on engineering designs and the resulting preferences of all the key stakeholders. This includes obtaining information under uncertainty, worth, time, and risk to create a new definition for engineering design. The new engineering design paradigm enables the development of a new engineering design methodology, which combines the strengths of UCD and VCD. A case study from the space industry demonstrates the use of Preference-Centric Design (PCD) and the engineering design framework. The objective function formulated for this method and framework was Expected Worth-Time Utility, which captures the uncertainty, worth, time, and risk. Hence, the best engineering design alternative is the one with the maximum Expected Worth-Time Utility. Ultimately, the PCD demonstrates how Systems Engineers can perform informed, objective, repeatable, transparent, and traceable engineering design under uncertainty, worth, time, and risk.

Date of Award	1 Aug 2022
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Peter Hollingsworth (Supervisor) & Katharine Smith (Supervisor)