Classical programmatic risk analysis focuses exclusively on budget and schedule. Yet, in the development of a critical system, the reliability also matters. The purpose of this dissertation is to develop a qualitative approach to tradeoff technical and management risks in interdependent projects within a program. Technical risks concern engineering failures. Management risks refer to schedule and budget overruns.; This research develops a probabilistic program risk management (PPRM) model involving a sequence of three optimization steps. The first step optimizes feasible technical design alternatives over the range of potential budgets to minimize each alternative's probability of technical failure. The second step considers the potential management risks associated with each design alternative and optimizes the risk mitigation strategy as a function of the budget reserve. The third step determines the optimal technical design alternative and budget reserves based on the lowest overall expected failure costs considering both technical and management failures.; The presentation of the PPRM model is structured around a set of assumptions regarding problem detection, partial failures, and project dependencies. First, the model analyzes, for one project, the optimal selection of the design configuration, the choice of components, and the optimal reserve level. Second, the model considers the same decisions in conjunction with the optimal level of testing and reviews. Third, the model considers the same decisions for one project, but includes partial failures. Finally, the model examines the management of one project when the outcome of this project affects other projects in the program. Illustrations of the model are based on a hypothetical case from NASA's unmanned space missions, which provides rich examples of dependent projects involving limited resources and multiple tradeoffs within programs.; The contribution of this dissertation is an analytical framework for (1) quantifying program risks (technical failures and management failures) to support management decisions about system design and financial reserves as a function of the budget, (2) explicitly comparing and trading off technical and management risks, and (3) modeling the effects of dependencies among projects in a program.
展开▼
