Framework for the reduction of programmatic risk on complex systems projects

Abstract

"In 2008, the cumulative cost growth in the Department of Defense's (DoD) portfolio of 96 major defense acquisition programs was $296 billion and the average delay in delivering promised capabilities to the war fighter was 22 months." This statement from the Director of Acquisition and Sourcing Management of the Government Accountability Office (GAO) before a House of Representatives panel is in reference to an alarming, decades-long trend in the Defense Industry of budget and schedule overruns. Defense projects are complex systems of humans, software, and hardware interacting in unpredictable and often-uncontrolled ways. The research presented in this thesis demonstrates that component and systemic failures in DoD systems have much in common with the overruns that their executing organizations experience. Complex systems accidents occur when their control mechanisms do not sufficiently enforce constraints on system components and their interactions. Similarly, project losses, in terms of budget and schedule overruns, occur when the control mechanisms of the executing organization do not sufficiently enforce constraints on project teams and their interactions. This thesis proposes a framework based on the principles of Control Theory, Systems Safety Analysis, and Earned Value Management, which project managers can apply in order to reduce programmatic risk on complex systems projects. The objectives of the thesis are: to provide project managers with a mechanism to control risk within the scope of the work they oversee, to provide individual contributors with a mechanism to control risk within the scope of the work they execute, to clearly demonstrate how poorly designed organization structures facilitate program losses, and to clearly demonstrate how well-designed organization structures can prevent or at the minimum mitigate program losses. At the completion of this thesis, it was found that complex systems programs have many tools at their disposal for defining relationships between elements of project scope and between teams in the executing organization. But few tools are available to specify how exactly a manager can accurately monitor and safely affect the scope elements under their control. The control structure specification and design presented within this thesis will address the primary causes of risk that lead to program losses.