Price, Robert Matthew (2016-08). A Capability-based Framework for Applying Value-driven Design to Systems with Multiple Value-producing Scenarios. Master's Thesis. | Thesis individual record

Value-driven design is an emerging paradigm in systems engineering and design that leverages insights from decision theory to inform decision-making in engineering design and improve systems engineering outcomes. A basic premise of value-driven design is that systems engineers should formalize preferences for system engineering project outcomes using a single overall scalar measure. To date, most value modeling work reported in the literature addresses the value of a system for a single primary usage scenario. The capabilities of the system are typically modeled with the assumptions and needs of this single scenario in mind, with the model of system capability integrated into the value model for this single scenario. However, many systems support multiple usage scenarios. Optimizing these systems solely for any one scenario may diminish their value on another. Additionally, maintaining multiple models of the system, each tailored for a single value scenario, is effort-intensive and introduces the risk that inconsistencies will develop. This thesis presents a framework for modeling these types of systems in a way that avoids the pitfalls and reflects a more accurate picture of the true system value. A key aspect of the framework is the clear distinction between modeling the technical capabilities of a system and modeling the value generation of usage scenarios that leverage those capabilities. System capabilities are modeled in such a way as to be compatible with any individual usage scenario while remaining broad enough to address all usage scenarios. This may necessitate expanding the top-level representation of system capabilities beyond a vector of attribute values to include relationships between attributes. Constructing the system capability model separate from the value models of individual scenarios clarifies the modeling tasks, simplifies model traceability, and grants decision makers flexibility in exploring portfolios of value-producing scenarios. The framework is demonstrated in a notional example involving the selection of system architectures for a non-commercial space launch vehicle that must support a variety of value-generating missions. In order to illustrate the usefulness of the approach, this example is then expanded upon to investigate the concept of robustness and how it might be valued in such systems.

etd chair
publication date