Stochastic modeling of responsiveness, schedule risk and obsolescence of space systems, and implications for design choices
Dubos, Gregory Florent
MetadataShow full item record
The U.S Department of Defense and the National Aeronautics and Space Administration continue to face common challenges in the development and acquisition of their space systems. In particular, space programs repeatedly experience significant schedule slippages, and spacecraft are often delivered on-orbit several months, sometimes years, after the initially planned delivery date. The repeated pattern of these schedule slippages suggests deep-seated flaws in managing spacecraft delivery and schedule risk, and an inadequate understanding of the drivers of schedule slippages. Furthermore, due to their long development time and physical inaccessibility after launch, space systems are exposed to a particular and acute risk of obsolescence, resulting in loss of value or competitive advantage over time. The perception of this particular risk has driven some government agencies to promote design choices that may ultimately be contributing to these schedule slippages, and jeopardizing what is increasingly recognized as critical, namely space responsiveness. The overall research objective of this work is twofold: (1) to identify and develop a thorough understanding of the fundamental causes of the risk of schedule slippage and obsolescence of space systems; and in so doing, (2) to guide spacecraft design choices that would result in better control of spacecraft delivery schedule and mitigate the impact of these "temporal risks" (schedule and obsolescence risks). To lay the groundwork for this thesis, first, the levers of responsiveness, or means to influence schedule slippage and impact space responsiveness are identified and analyzed, including design, organizational, and launch levers. Second, a multidisciplinary review of obsolescence is conducted, and main drivers of system obsolescence are identified. This thesis then adapts the concept of a technology portfolio from the macro- or company level to the micro-level of a single complex engineering system, and it analyzes a space system as a portfolio of technologies and instruments, each technology with its distinct stochastic maturation path and exposure to obsolescence. The selection of the spacecraft portfolio is captured by parameters such as the number of instruments, the initial technology maturity of each technology/instrument, the resulting heterogeneity of the technology maturity of the whole system, and the spacecraft design lifetime. Building on the abstraction of a spacecraft as a portfolio of technologies, this thesis then develops a stochastic framework that provides a powerful capability to simultaneously explore the impact of design decisions on spacecraft schedule, on-orbit obsolescence, and cumulative utility delivered by the spacecraft. Specifically, this thesis shows how the choice of the portfolio size and the instruments Technology Readiness Levels (TRLs) impact the Mean-Time-To-Delivery (MTTD) of the spacecraft and mitigate (or exacerbate) schedule risk. This work also demonstrates that specific combinations/choices of the spacecraft design lifetime and the TRLs can reduce the risk of on-orbit obsolescence. This thesis then advocates for a paradigm shift towards a calendar-based design mindset, in which the delivery time of the spacecraft is accounted for, as opposed to the traditional clock-based design mindset. The calendar-based paradigm is shown to lead to different design choices, which are more likely to prevent schedule slippage and/or enhance responsiveness and ultimately result in a larger cumulative utility delivered. Finally, missions scenarios are presented to illustrate how the framework and analyses here proposed can help identify system design choices that satisfy various mission objectives and constraints (temporal as well as utility-based).