TitleIntegrating AFSIM as an internal predictor
Publication TypeReport
Year of Publication2014
AuthorsJensen, BAW, Karneeb, J, Borck, H, Aha, DW
Series TitleNCARAI Technical Note
Document NumberAIC-14-172
InstitutionNaval Research Laboratory, Navy Center for Applied Research in Artificial Intelligence
CityWashington, DC

My Summer 2014 internship consisted of contributing to the Autonomy for Air Combat Missions (ATACM) project at the Knexus Research Corporation, a largely collaborative effort involving the Naval Research Lab (NRL), the Naval Air Systems Command (NAVAIR) and the Air Force Research Lab (AFRL). The ATACM project's main purpose is to create an autonomous Unmanned Aerial Vehicle (UAV) capable of lying alongside a human pilot in BVR (Beyond Visual Range) air combat missions. My job as part of Knexus, subcontracted under NRL, consisted of working on the reasoning systems for the UAV, with the label the Tactical Battle Manager (TBM). NAVAIR and AFRL were responsible for interfacing the TBM with their respective simulators and providing domain specific knowledge in the realm of air combat. My contribution to the ATACM project took the form of a predictor, under the name PEPR (the Planned Expected State Predictor), for use in the goal reasoning components of the TBM.

Refereed DesignationNon-Refereed
Full Text
NRL Publication Release Number: