MANERVA: Mission Assessment of Non-Manned Entities for Rating the Validation of Autonomy

The dependability, persistence, and versatility of unmanned systems have made them indispensable assets in today’s warfighting. As they are tasked to fulfill new missions in unpredictable, dynamic environments, they are transitioning from remote control into the realm of autonomy. Here they perform rapid and semi-autonomous behaviors, such as re-planning and re-tasking, to accomplish mission objectives in mixed-initiative fashion along with skilled human operators. For operators and their commanders to develop confidence in these autonomous control algorithms, they must demonstrate that they can reliably complete mission tasks, and provide operators with a better understanding of the capabilities and limitations of their systems during varying mission scenarios. We propose to develop a Mission Assessment of Non-Manned Entities for Rating the Validation of Autonomy (MANERVA) analytical verification and validation workbench. MANERVA will allow the user to configure verification experiments for particular mission scenarios, select autonomous control algorithms to run, choose metrics (and add new ones) to promote objective evaluation, execute simulations to generate vehicle paths and behavioral actions, and analyze the results and rate the validity of the vehicle’s determined course of action by representing its effectiveness using novel visualization techniques.