Value-centric methodology for requirement verification under uncertainty.

摘要

The development of a Department of Defense (DoD) Product, no matter how simple or complex, will ultimately undergo the execution of a requirement verification plan to make sure that the product being developed meets the requirements that it was designed to meet. In DoD organizations, these plans are usually developed and executed with cost and schedule as two significant variables. Technical risk is also addressed but from the context of acknowledging that certain attributes have the potential to impact cost, schedule, or mission. While other influential pressure points are also critical (e.g., expected results, benefits, and uncertainty) these do not typically play a major role in the verification plan development. The execution of these plans can be very expensive and decision makers in an effort to meet funding levels or schedule constraints are tempted to utilize verification approaches that are less costly or less time consuming (e.g., less testing). Choosing to verify a product by analysis rather than testing may initially be cost and schedule favorable, but will it be the most appropriate decision when other factors are taken into account, such as the probability that testing the article may yield a more complete verification than by analysis. How about if the product passes the analysis but still has a probability of failing in the field? What are the consequences associated with a failure in the field? These questions, elements, and other variables (e.g., political pressure, environmental concerns, expected profit, deterrence, and uncertainty) play a less influential role in the decision but can have a substantial impact on the program. While a holistic view of all of the pressure points that would impact the decision should be made, more often than not, decisions are driven by cost or schedule pressures. How can the decision maker adequately choose between a test or an analysis within the context of addressing cost, schedule, verification compliance assuredness, known conditions, and uncertainty? The research outlined in this dissertation develops a value-centric methodology to address the latter question. This value-centric methodology can assist DoD decision makers in choosing the most appropriate requirement verification strategy for those requirements that can be verified by either test or analysis. The methodology will incorporate cost and schedule impacts, as well as other influential variables such as uncertainty, benefit, and value. While this methodology is focused on DoD requirement verification decisions addressing test or analysis approaches, it can be applied to a wide range of decisions.