Autonomica: Ontological Modeling and Analysis of Autonomous Behavior

Model-based system autonomy is a complex integration of planning of low-level command sequences from high-level goals, and execution of such command sequences to control a system. The need for autonomy has accelerated in recent years to meet the demand for more complex missions in domains like automotive, space, and defense. During system development, an understanding of the complex relationship between system autonomy and the physical environment (including hardware) is critical to support trade studies, developing concepts of operations, characterizing risk, and performing testing. This paper describes the initial results of developing Autonomica, an ontology-based method, and framework for autonomous behavior modeling and analysis. This method formalizes the State Analysis (SA) [Ingham et al, 2005] architectural pattern as a vocabulary with description logic semantics, and provides authoring viewpoints and analysis capabilities (reasoning, querying, simulation) for the SA-based architectures. The framework implements the method in an integrated workbench. Ideas are illustrated using a running example of a hypothetical mission to a small space body using an autonomous spacecraft.