Agent-based simulation framework and consensus algorithm for observing systems with adaptive modularity

In this era of the big data revolution, the desired capabilities of Earth Observing Systems are growing fast: we need ever more frequent data sets, covering a larger part of the frequency spectrum, with lower latency, and higher spatial resolution. To better address these needs, the space systems community has been exploring the value of shifting from highly monolithic architectures, in which large and isolated spacecraft carry multiple instruments with synergistic and complementary goals, toward more distributed architectures, where the functions of these large systems are partitioned into a larger number of smaller satellites. In this paper, we present an agent-based simulation framework that can help systems engineers assess whether or not it makes sense to be able to change system modularity during operations by means of temporary coalitions. Systems of observing autonomous vehicles work together to perform a set of observational tasks. The vehicles can decide to form physical coalitions with other vehicles for collective sensing of a target, when no agent alone can carry out the task, or individual observation results in degraded satisfaction. The framework extends the well-known decentralized Coupled-Constraint Consensus-Based Bundle Algorithm to multivehicle single-task allocation and introduces constraints on the formation of coalitions, so that agents can create or split a coalition depending on the benefits and costs associated with these actions. The framework is described in detail and demonstrated on a case study.