Metrics for Intelligent Autonomy

Abstract : Intelligent Autonomy (IA) is a multi-year program within the Office of Naval Research (ONR) Autonomous Operations (AO) Future Naval Capabilities (FNC) program. The primary goal of the effort is to develop and demonstrate technologies for highly automated and fully autonomous mission planning and dynamic re-tasking of multiple classes of Naval unmanned systems and minimization of human intervention in unmanned vehicle operations. This technology is being applied to both individual and teams of unmanned air, surface, ground, and undersea vehicles for a variety of mission areas including reconnaissance/search, persistent surveillance, tracking, and some limited application to strike. Autonomy technologies will be matured through a series of phased demonstrations to allow low risk transition to current and future Navy and Marine Corps systems. Demonstrations will be done using both real vehicles and simulation. Some of the major simulation demonstrations will be done within the context of a simulated warfare environment at the Naval Air Systems Command based around the Air Combat Environment Test and Evaluation Facility (ACETEF) and the Unmanned System Research and Development Lab (USRDL). The demonstrations at NAVAIR will utilize much of the architecture and many of the assets from the NCW4.0X Virtual Laboratory (V-LAB) project. Metrics for testing of IA software in this environment are currently being developed. This paper will discuss some candidate performance metrics that are currently being considered for evaluation of the Intelligent Autonomy technologies.