Comparative cost and utility analysis of monolith and fractionated spacecraft using failure and replacement Markov models

Abstract Failure of a single component on-board a spacecraft can compromise the integrity of the whole system and put its entire capability and value at risk. Part of this fragility is intrinsic to the current dominant design of space systems, which is mainly a single, large, monolithic system. The space industry has therefore recently proposed a new architectural concept termed fractionation, or co-located space-based network (SBN). By physically distributing functions in multiple orbiting modules wirelessly connected, this architecture allows the sharing of resources on-orbit (e.g., data processing, downlinks). It has been argued that SBNs could offer significant advantages over the traditional monolithic architecture as a result of the network structure and the separation of sources of risk in the spacecraft. Careful quantitative analyses are still required to identify the conditions under which SBNs can “outperform” monolithic spacecraft. In this work, we develop Markov models of module failures and replacement to quantitatively compare the lifecycle cost and utility of both architectures. We run Monte-Carlo simulations of the models, and discuss important trends and invariants. We then investigate the impact of our model parameters on the existence of regions in the design space in which SBNs “outperform” the monolith spacecraft on a cost, utility, and utility per unit cost basis. Beyond the life of one single spacecraft, this paper compares the cost and utility implications of maintaining each architecture type through successive replacements.