Informed Mission Management via PAiring Resources to Tasks (IMPART)
Agency / Branch:
DOD / USAF
Within Air Operations Centers (AOCs), teams of highly-experienced personnel rely heavily on their expertise to match available resources to requested tasks. However, as platforms and sensors begin to emerge with multi-role capabilities, new challenges arise for mission managers in finding the optimal allocation of resources to tasks. Automated resource allocation tools can help; however, current tools often arrive at solutions that do not "make sense" according to experienced personnel, failing to take into account the subtleties of expert planning. To improve the quality of planning systems, we need to capture planning experts' experiential knowledge, encode it in a machine-useable form, and bootstrap automated algorithms to derive more suitable solutions to large-scale resource allocation problems. We propose to design and demonstrate a framework for Informed Mission Management via PAiring Resources to Tasks (IMPART), an intelligent, extensible mapping framework that supports AOC functional teams in the collaborative planning across the full-spectrum of air operations for superior resource utilization. Our approach includes the development of (1) a rich, extensible data representation to express air operations mission management elements; (2) a learning framework to capture experiential knowledge from expert planners; and (3) planning services to recommend appropriate resource-task pairings that leverage the captured knowledge. BENEFIT: The research performed under this effort will have immediate benefit to a number of military planning and execution systems including the AOC Weapon System, TBMCS, and DCGS-AF. Additionally, the standard data representation refined under IMPART will help provide a common representation for the Air Operations Community of Interest (AO COI). We seek to transition the data representation, learning framework, and collaborative planning tools to the C2IS and C2AOS programs currently spearheaded by the 350 ELSG at ESC. This research will also have direct application to enhance our commercial EAToolkitT product, a software development kit for optimization using evolutionary algorithms.
Small Business Information at Submission:
Ninos E. Hanna
Charles River Analytics Inc.
625 Mount Auburn Street Cambridge, MA 02138
Number of Employees: